3GPP TR 38.811

3GPP TR 38.811 V15.1.
0 (2019-06)
Technical Report
3rd Generation Partnership Project;

Technical Specification Group Radio Access Network;
Study on New Radio (NR) to support non-terrestrial networks
(Release 15)
The present document has been developed within the 3rd Generation Partnership Project (3GPP TM) and may be further elaborated for the purposes of 3GPP.
The present document has not been subject to any approval process by the 3GPP Organizational Partners and shall not be implemented.
This Report is provided for future development work within 3GPP only. The Organizational Partners accept no liability for any use of this Specification.
Specifications and Reports for implementation of the 3GPP TM system should be obtained via the 3GPP Organizational Partners' Publications Offices.
Release 15 2 3GPP TR 38.811 V15.1.0 (2019-06)
Keywords
Satellite, Aerial, 5G
3GPP
Postal address
3GPP support office address

650 Route des Lucioles - Sophia Antipolis
Valbonne - FRANCE
Tel.: +33 4 92 94 42 00 Fax: +33 4 93 65 47 16
Internet
http://www.3gpp.org
Copyright Notification
No part may be reproduced except as authorized by written permission.

The copyright and the foregoing restriction extend to reproduction in all media.
© 2019, 3GPP Organizational Partners (ARIB, ATIS, CCSA, ETSI, TSDSI, TTA, TTC).
All rights reserved.
UMTS™ is a Trade Mark of ETSI registered for the benefit of its members
3GPP™ is a Trade Mark of ETSI registered for the benefit of its Members and of the 3GPP Organizational Partners
LTE™ is a Trade Mark of ETSI registered for the benefit of its Members and of the 3GPP Organizational Partners
GSM® and the GSM logo are registered and owned by the GSM Association
3GPP
Release 15 3 3GPP TR 38.811 V15.1.0 (2019-06)
Contents
Foreword............................................................................................................................................................. 6
1 Scope ........................................................................................................................................................ 7
2 References ................................................................................................................................................ 7
3 Definitions, symbols and abbreviations ................................................................................................. 10
3.1 Definitions ....................................................................................................................................................... 10
3.2 Symbols ........................................................................................................................................................... 11
3.3 Abbreviations ................................................................................................................................................... 11
4 Non-Terrestrial Networks Overview – background information ........................................................... 13
4.1 Roles for Non-Terrestrial Networks in 5G system .......................................................................................... 13
4.2 5G Use Cases wherein Non-Terrestrial Network components have a role ...................................................... 13
4.2.1 5G use cases introduction ........................................................................................................................... 13
4.3 Satellite and aerial access network architecture principles .............................................................................. 17
4.4 Characteristics of NTN Terminals for satellite / aerial access network ........................................................... 20
4.5 Air/Space borne vehicle characteristics ........................................................................................................... 21
4.6 Coverage pattern of NTN ................................................................................................................................ 21
4.7 Non-Terrestrial Network architecture options ................................................................................................. 22
4.8 Spectrum .......................................................................................................................................................... 23
5 Non-Terrestrial Networks deployment scenarios ................................................................................... 24
5.1 Scenarios overview .......................................................................................................................................... 24
5.2 Attributes ......................................................................................................................................................... 25
5.3 Doppler and Propagation delay characterisation.............................................................................................. 26
5.3.1 Methodology .............................................................................................................................................. 26
5.3.1.1 Propagation delay ................................................................................................................................. 26
5.3.1.2 Differential delay .................................................................................................................................. 26
5.3.1.3 Doppler shift/variation .......................................................................................................................... 26
5.3.2 Geo-stationary platforms ............................................................................................................................ 27
5.3.2.3 Doppler shift ......................................................................................................................................... 29
5.3.3 Aerial vehicle ............................................................................................................................................. 32
5.3.4 Non geostationary satellites ....................................................................................................................... 33
5.3.4.3 Doppler Shift and variation rate ........................................................................................................... 34
5.3.4.3.1 Case at 2 GHz ................................................................................................................................. 36
5.3.4.3.2 Case in Ka band .............................................................................................................................. 37
5.3.4.4 Doppler Shift and variation rate ........................................................................................................... 41
5.3.5 Synthesis for each scenarios ....................................................................................................................... 42
6 Non-Terrestrial Networks channel models............................................................................................. 43
6.1. Status/expectation of existing information for satellite/HAPS channels ......................................................... 43
6.1.1 Channel modeling works outside of 3GPP ................................................................................................. 43
6.1.2 Targeted user environment ......................................................................................................................... 43
6.1.3 Modeling objectives ................................................................................................................................... 43
6.2 Differences between satellite/HAPS and cellular channel modelling .............................................................. 43
6.3 Coordinate system ........................................................................................................................................... 45
6.4 Antenna modelling ........................................................................................................................................... 45
6.4.1 HAPS/Satellite antenna .............................................................................................................................. 45
6.4.2 UE antenna pattern ..................................................................................................................................... 46
6.5 Methodology to define channel models ........................................................................................................... 46
6.5.1 System-level methodology ......................................................................................................................... 46
6.5.2 Link-level methodology ............................................................................................................................. 47
6.6 Large scale model ............................................................................................................................................ 48
6.6.1 LOS probability .......................................................................................................................................... 48
3GPP
Release 15 4 3GPP TR 38.811 V15.1.0 (2019-06)
6.6.2 Path loss and Shadow fading ...................................................................................................................... 48

6.6.3 O2I penetration loss ................................................................................................................................... 50
6.6.4 Atmospheric absorption ............................................................................................................................. 51
6.6.5 Rain and cloud attenuation ......................................................................................................................... 52
6.6.6 Scintillation ................................................................................................................................................ 52
6.6.6.1 Ionospheric scintillation ....................................................................................................................... 52
6.6.6.1.1 Ionospheric scintillation indices ...................................................................................................... 53
6.6.6.1.2 Ionospheric scintillation location dependence ................................................................................ 53
6.6.6.1.3 Frequency scaling ........................................................................................................................... 54
6.6.6.1.4 Model for Ionospheric scintillation loss .......................................................................................... 54
6.6.6.2 Tropospheric scintillation ..................................................................................................................... 56
6.6.6.2.1 Model for Tropospheric scintillation loss ....................................................................................... 57
6.7 Fast fading model ............................................................................................................................................ 58
6.7.1 Flat fading .................................................................................................................................................. 58
6.7.2 Frequency selective fading ......................................................................................................................... 61
6.8. Additional modelling components ................................................................................................................... 85
6.8.1 Time-varying Doppler shift ........................................................................................................................ 85
6.8.2 Faraday rotation ......................................................................................................................................... 85
6.9 Channel models for link level simulations ....................................................................................................... 87
6.9.1 CDL models ............................................................................................................................................... 87
6.9.2 TDL models ............................................................................................................................................... 90
6.10 Channel model calibration ............................................................................................................................... 93
6.10.1 NTN channel model features per deployment scenarios ............................................................................ 93
7 Potential key impact areas on NR to support NTN ................................................................................ 95
7.1 Specific constraints associated to NTN ........................................................................................................... 95
7.2 NR features/protocols potentially affected ...................................................................................................... 97
7.3 NR modifications to support the Non-Terrestrial Network deployment scenarios ........................................ 100
7.3.1 Methodology ............................................................................................................................................ 100
7.3.2 Motion of the space/aerial vehicles .......................................................................................................... 100
7.3.2.1 Hand-Over and paging........................................................................................................................ 100
7.3.2.1.1 Problem statement ......................................................................................................................... 100
7.3.2.1.2 Assessment of conditions for NR operation in Non-Terrestrial networks .................................... 101
7.3.2.1.3 NR impact considerations ............................................................................................................. 101
7.3.2.2 TA adjustment .................................................................................................................................... 101
7.3.2.3 Initial synchronization in downlink .................................................................................................... 103
7.3.2.4 DMRS time density ............................................................................................................................ 104
7.3.3 Altitude of the space/aerial vehicles ......................................................................................................... 105
7.3.3.1 HARQ ................................................................................................................................................. 105
7.3.3.1.2 Assessment of conditions for NR operation in non-terrestrial networks ...................................... 107
7.3.3.2 MAC/RLC procedures ........................................................................................................................ 107
7.3.3.3 Physical layer procedures (ACM, power control) .............................................................................. 108
7.3.4 Cell size (Beam foot print) ....................................................................................................................... 108
7.3.4.1 PRACH and Random access .............................................................................................................. 108
3GPP
Release 15 5 3GPP TR 38.811 V15.1.0 (2019-06)

7.3.4.2 TA in Random access response message ............................................................................................ 110
7.3.5 Propagation channel ................................................................................................................................. 111
7.3.5.1 DMRS frequency density ................................................................................................................... 111
7.3.5.2 Cyclic prefix ....................................................................................................................................... 112
7.3.6 Duplex mode ............................................................................................................................................ 113
7.3.6.1 FDD/TDD Duplexing mode ............................................................................................................... 113
7.3.7 Satellite or aerial Payload performance .................................................................................................... 114
7.3.7.1 PT-RS ................................................................................................................................................. 114
7.3.7.2 PAPR .................................................................................................................................................. 115
7.3.8 Network architecture ................................................................................................................................ 116
7.3.8.1 Protocols ............................................................................................................................................. 116
7.3.8.1.2 Assessment of conditions for NR operation in Non-Terrestrial Networks ................................... 117
8 Recommendations on the way forward ................................................................................................ 121
8.1 General outcomes .......................................................................................................................................... 121
8.2 Reference deployment scenarios.................................................................................................................... 121
8.3 Non-Terrestrial network channel modelling .................................................................................................. 121
8.4 NR impacts to support Non-Terrestrial networks .......................................................................................... 121
8.4.1 Type of potential NR impacts .................................................................................................................. 121
8.4.2 Assessment of potential NR impacts to support Non-Terrestrial networks .............................................. 121
Annex A: Example of reference scenario for calibration of large scale parameters...................... 124
Annex B: Non Terrestrial network characteristics ........................................................................... 125
B.1 NTN Phase noise masks ....................................................................................................................... 125
Annex C: Change History ................................................................................................................... 126
3GPP
Release 15 6 3GPP TR 38.811 V15.1.0 (2019-06)
Foreword
This Technical Report has been produced by the 3rd Generation Partnership Project (3GPP).
The contents of the present document are subject to continuing work within the TSG and may change following formal
TSG approval. Should the TSG modify the contents of the present document, it will be re-released by the TSG with an
identifying change of release date and an increase in version number as follows:
Version x.y.z
where:
x the first digit:
1 presented to TSG for information;
2 presented to TSG for approval;
3 or greater indicates TSG approved document under change control.
y the second digit is incremented for all changes of substance, i.e. technical enhancements, corrections, updates,
etc.
z the third digit is incremented when editorial only changes have been incorporated in the document.
3GPP
Release 15 7 3GPP TR 38.811 V15.1.0 (2019-06)
1 Scope
This technical report is related to a study item New Radio to support Non-Terrestrial Networks. The purpose of this TR
is to collect the TSG RAN and RAN WG1 findings related to the study.
The objectives for the study are the following
- Definition of the Non-Terrestrial Networks deployment scenarios and related system parameters such as
architecture, altitude, orbit etc.
- Adaptation of the 3GPP channel models for non-terrestrial networks (propagation conditions, mobility, …).
- For the described deployment scenarios, identification of any key impact areas on the New Radio interface that
may need further evaluations.
2 References
The following documents contain provisions which, through reference in this text, constitute provisions of the present
document.
- References are either specific (identified by date of publication, edition number, version number, etc.) or
non-specific.
- For a specific reference, subsequent revisions do not apply.
- For a non-specific reference, the latest version applies. In the case of a reference to a 3GPP document (including
a GSM document), a non-specific reference implicitly refers to the latest version of that document in the same
Release as the present document.
[1] 3GPP TR 21.905: "Vocabulary for 3GPP Specifications".
[2] 3GPP TS 36.101: "Technical Specification Group Radio Access Network; Evolved Universal
Terrestrial Radio Access (E-UTRA); User Equipment (UE) radio transmission and reception
(Release 14)".
[3] 3GPP TR 38.801: "Technical Specification Group Radio Access Network; Study on new radio
access technology: Radio access architecture and interfaces (Release 14)".
[4] 3GPP TR 38.804: "Technical Specification Group Radio Access Network; Study on New Radio
Access Technology; Radio Interface Protocol Aspects (Release 14)".
[5] 3GPP TR 38.913: "Study on Scenarios and Requirements for Next Generation Access
Technologies (Release 14)".
[6] 3GPP TS 22.261: "Service requirements for next generation new services and markets".
[7] A. Guidotti et. al, "Satellite-enabled LTE systems in LEO Constellations", 2017 IEEE
International Conference on Communications (ICC), Paris, May 2017, pp. 876-881, doi:
10.1109/ICCW.2017.7962769
[8] 3GPP TR 38.802 v14.1.0, "Study on New Radio Access Technology Physical Layer Aspects
(Release 14)".
[9] Void
[10] Recommendation ITU-R P.681-10, "Propagation data required for the design of earth-space land
mobile telecommunication systems", Dec. 2017.
[11] Recommendation ITU-R P.618-13, "Propagation data and prediction methods required for the
design of Earth-space telecommunication systems", Dec. 2017.
3GPP
Release 15 8 3GPP TR 38.811 V15.1.0 (2019-06)
[12] 3GPP TR 38.901: "Study on channel model for frequencies from 0.5 to 100 GHz (Release 14)".
[13] Recommendation ITU-R P.531-13, "Ionospheric propagation data and prediction methods required
for the design of satellite services and systems", September 2016.
[14] Wheelon, A.D., Electromagnetic Scintillation II. Weak Scattering, Cambridge Univ. Press, pp.
110-111, Cambridge, U.K., 2005.
[15] F. P. Fontan, M. Vazquez-Castro, C. E. Cabado, J. P. Garcia and E. Kubista, "Statistical modeling

of the LMS channel," in IEEE Transactions on Vehicular Technology, vol. 50, no. 6, pp. 1549-
1567, Nov 2001
[16] A. Jahn and E. Lutz, "Propagation Data and channel model for LMS systems", Final Rep. ESA PO
141 742. DLR, 1995.
[17] Prieto-Cerdeira, R., Perez-Fontan, F., Burzigotti, P., Bolea-Alamañac, A. and Sanchez-Lago, I.,
"Versatile two-state land mobile satellite channel model with first application to DVB-SH
analysis", Int. J. Satell. Commun. Network., 28: 291–315., June 2010.
[18] Void
[19] Void
[20] 3GPP TS 38.321 "Technical Specification Group Radio Access Network; NR; Medium Access
Control (MAC) protocol specification (Release 15)".
[21] 3GPP TS 38.211, "Technical Specification Group Radio Access Network; NR; Physical channels
and modulation (Release 15)".
[22] 3GPP TS 38.133 "Technical Specification Group Radio Access Network; NR; Requirements for
support of radio resource management (Release 15)".
[23] 3GPP TS 38.214, "Technical Specification Group Radio Access Network; NR; Physical layer
procedures for data (Release 15)".
[24] 3GPP TR 38.802 v14.1.0, "Study on New Radio Access Technology Physical Layer Aspects
(Release 14)", June 2016.
[25] 3GPP TS 36.101: "Technical Specification Group Radio Access Network; Evolved Universal
Terrestrial Radio Access (E-UTRA); User Equipment (UE) radio transmission and reception
(Release 14)".
[26] 3GPP TS 38.101, "Technical Specification Group Radio Access Network; NR; User Equipment
(UE) radio transmission and reception".
[27] 3GPP TS 36.104, "Technical Specification Group Radio Access Network; Evolved Universal
Terrestrial Radio Access (E-UTRA); Base Station (BS) radio transmission and reception".
[28] O. Kodheli, A. Guidotti, and A. Vanelli-Coralli. "Integration of Satellites in 5G through LEO

Constellations", CoRR abs/1706.06013 (2017): pp. 1-6
[29] 3GPP TS 36.321, "Evolved Universal Terrestrial Radio Access (E-UTRA); Medium Access
Control (MAC) protocol specification (Release 15)", v1.0.0 (2017-12).
[30] R1-1802632, "Considerations on random access for non-terrestrial networks", Interdigital Inc.,
3GPP TSG RAN WG1 Meeting #93, Busan, Korea, May 21st – 25th, 2018.
[31] R1-1804236, "Discussion on the NR impacts on random access for NTN", ZTE, 3GPP TSG RAN
WG1 Meeting #93, Busan, Korea, May 21st – 25th, 2018.
[32] RP-180664, "NR-NTN: solution principles for NR to support non-terrestrial networks", Thales et
al, 3GPP TSG RAN Meeting #80, La Jolla, USA, June 11th – 14th, 2018.
[33] R1-1806768 "Considerations on timing advance and random access for NTN", Nokia, Nokia
Shanghai Bell, 3GPP TSG RAN WG1 Meeting #93, Busan, Korea, May 21st – 25th, 2018.
3GPP
Release 15 9 3GPP TR 38.811 V15.1.0 (2019-06)
[34] ITU-R M.1225, "Guidelines for evaluation of radio transmission technologies for IMT-2000",
1997
[35] E. Lutz, M. Werner, and A. Jahn, "Satellite Systems for Personal and Broadband
Communications", Berlin, Germany, Springer-Verlag, 2000.
[36] T. Rappaport, "Wireless Communications", New Jersey: Prentice Hall, 1996
[37] DVB Document A171-2, Digital Video Broadcasting (DVB) Implementation guidelines for the
second generation system for Broadcasting, Interactive Services, News Gathering and other
broadband satellite applications; Part 2 - S2 Extensions (DVB-S2X), March 2015.
[38] 3GPP TR 38.803 v14.2.0, "Study on new radio access technology: Radio Frequency (RF) and co-
existence aspects".
[39] R1-1806750, "Considerations on random access for NTN", Samsung, 3GPP TSG RAN WG1
Meeting #93, Busan, Korea, May 21st – 25th, 2018.
[40] 3GPP TS 38.104, "Technical Specification Group Radio Access Network; NR; Base Station (BS)
radio transmission and reception".
[41] R. Ali Ahmad, J. Lacan, F. Arnal, M. Gineste and L. Clarac, "Enhancing satellite system
throughput using adaptive HARQ for delay tolerant services in mobile communications", 2015
Wireless Telecommunications Symposium (WTS), New York, NY, 2015, pp. 1-7.
[42] Tae Chul Hong, Kunseok Kang, Bon-Jun Ku, Dae-Ig Chang, "Receiver memory management
method for HARQ in LTE-based satellite communication system", International Journal of
Satellite Communications and Networking, 2017, 35, 1, 3.
[43] R1-1802613, "NTN NR impacts on the HARQ Operation", Fraunhofer, 3GPP TSG RAN1 meeting
#92, Athens, Greece, Feb. 26 - Mar. 2, 2018.
[44] R1-1804857, "Deactivating HARQ for Non-Terrestrial Networks", InterDigital Inc, 3GPP TSG
RAN WG1 Meeting #93, Busan, Korea, May 21st – 25th, 2018.
[45] R1-1805848, "Consideration on HARQ Impact for NTN", Nokia, Nokia Shanghai Bell, 3GPP
TSG RAN WG1 Meeting #93, Busan, Korea, May 21st – 25th, 2018.
[46] O. Kodheli, A. Guidotti, and A. Vanelli-Coralli. "Integration of Satellites in 5G through LEO

Constellations", CoRR abs/1706.06013 (2017): pp. 1-6.
[47] R1-1807164, "NR-NTN Channel Modeling – Flat fading criteria", THALES, 3GPP TSG RAN
WG1 Meeting #93, Busan, Korea, May 21st – 25th, 2018.
[48] R1-1802551, "UE antenna assumption and beam modelling for NTN", Nokia, 3GPP TSG RAN
WG1 Meeting #92, Athens, Greece, February 20th – March 2nd, 2018.
3GPP
Release 15 10 3GPP TR 38.811 V15.1.0 (2019-06)
3 Definitions, symbols and abbreviations
3.1 Definitions
For the purposes of the present document, the terms and definitions given in 3GPP TR 21.905 [1] and the following
apply. A term defined in the present document takes precedence over the definition of the same term, if any, in 3GPP
TR 21.905 [1].
Aerial: an airborne vehicle embarking a bent pipe payload or a regenerative payload telecommunication transmitter,
typically at an altitude between 8 to 50 km.
Airborne vehicles: Unmanned Aircraft Systems (UAS) encompassing tethered UAS (TUA), Lighter than Air UAS
(LTA), Heavier than Air UAS (HTA), all operating in altitudes typically between 8 and 50 km including High Altitude
Platforms (HAPs)
Availability: % of time during which the RAN is available for the targeted communication. Unavailable
communication for shorter period than [Y] ms shall not be counted. The RAN may contain several access network
components among which an NTN to achieve multi-connectivity or link aggregation.
Beam throughput: data rate provided in a beam
Bentpipe payload: payload that changes the frequency carrier of the uplink RF signal, filters and amplifies it before
transmitting it on the downlink
Connectivity: capability to establish and maintain data / voice / video transfer between networks and parts thereof
Geostationary Earth orbit: Circular orbit at 35,786 kilometres above the Earth's equator and following the direction of
the Earth's rotation. An object in such an orbit has an orbital period equal to the Earth's rotational period and thus
appears motionless, at a fixed position in the sky, to ground observers.
Low Earth Orbit: Orbit around the around Earth with an altitude between 500 kilometres (orbital period of about 88
minutes), and 2,000 kilometres (orbital period of about 127 minutes).
Medium Earth Orbit: region of space around the Earth above low Earth orbit and below geostationary Earth Orbit.
Mobile Services: a radiocommunication service between mobile and land stations, or between mobile stations
Mobile Satellite Services: A radiocommunication service between mobile earth stations and one or more space stations,
or between space stations used by this service; or between mobile earth stations by means of one or more space stations
Non Geostationary Satellites: Satellites (LEO and MEO) orbiting around the Earth with a period that varies
approximately between 1.5 hour and 10 hours. It is necessary to have a constellation of several Non Geostationary
satellites associated with handover mechanisms to ensure a service continuity.
Non-terrestrial networks: Networks, or segments of networks, using an airborne or space-borne vehicle to embark a
transmission equipment relay node or base station.
On Board processing: digital processing carried out on uplink RF signals aboard a satellite or an aerial.
One way latency: time required to propagate through the RAN from a terminal to the gateway or from the gateway to
the terminal. This is especially used for voice and video conference applications.
Regenerative payload: payload that transforms and amplifies an uplink RF signal before transmitting it on the
downlink. The transformation of the signal refers to digital processing that may include demodulation, decoding, re-
encoding, re-modulation and/or filtering.
Relay node: Relay of Uu radio interface. The relay function can take place at Layer 1, 2 or 3.
Reliability: probability that the RAN performs in a satisfactory manner for a given period of time when used under
specific operating conditions. The RAN may contain several access network components including an NTN to achieve
multi-connectivity or link aggregation.
3GPP
Release 15 11 3GPP TR 38.811 V15.1.0 (2019-06)
Round Trip Delay: time required for a network communication to travel from a terminal to the gateway or from the
gateway to the terminal and back. This is especially used for web based applications.
Satellite: a space-borne vehicle embarking a bent pipe payload or a regenerative payload telecommunication transmitter,
placed into Low-Earth Orbit (LEO) typically at an altitude between 500 km to 2000 km, Medium-Earth Orbit (MEO)
typically at an altitude between 8000 to 20000 km, or Geostationary-satellite Earth Orbit (GEO) at 35 786 km altitude.
Space-borne vehicles: Satellites including Low Earth Orbiting (LEO) satellites, Medium Earth Orbiting (MEO)
satellites, Geostationary Earth Orbiting (GEO) satellites as well as Highly Elliptical Orbiting (HEO) satellites
User Connectivity: capability to establish and maintain data / voice / video transfer between networks and Terminals
User Throughput: data rate provided to a terminal
3.2 Symbols
For the purposes of the present document, the following symbols apply:
<symbol> <Explanation>
3.3 Abbreviations
For the purposes of the present document, the abbreviations given in 3GPP TR 21.905 [1] and the following apply. An
abbreviation defined in the present document takes precedence over the definition of the same abbreviation, if any, in
3GPP TR 21.905 [1].
2D two-dimensional
3D three-dimensional
ACM Adaptive Modulation and Coding
AMF Access and Mobility Management Function
AOA Azimuth angle Of Arrival
AOD Azimuth angle Of Departure
ARQ Automatic Repeat Request
AS Angular Spread
ASA Azimuth angle Spread of Arrival
ASD Azimuth angle Spread of Departure
AWGN Additive White Gaussian Noise
BLOS Beyond Line of Sight
BS Base Station
BW Bandwidth
C2 Control and Command
CDF Cumulative Distribution Function
CDL Cluster Delay Line
CFO Carrier Frequency Offset
CP Cyclic Prefix
CPE Common Phase Error
DS Delay Spread
EIRP Equivalent Isotropically Radiated Power
eMBB enhanced Mobile Broadband
FDM Frequency-Division Multiplexed
FOTA Firmware Over The Air services
FSS Fixed Satellite Services
GEO Geostationary Earth Orbiting
gNB next Generation Node B
GNSS Global Navigation Satellite System
GSO Geo Synchronous Orbit
GW Gateway
HAPS High Altitude Platform Station
HARQ Hybrid Automatic Repeat Request
HD High Definition
HEO Highly Elliptical Orbiting
3GPP
Release 15 12 3GPP TR 38.811 V15.1.0 (2019-06)
ICI Inter-Carrier Interference

IoT Internet of Things
K Ricean K factor
KPI Key Performance Indicator
IMUX Input MUltipleXer
ISL Inter-Satellite Links
LEO Low Earth Orbiting
LMS Land Mobile Satellite
LOS Line of Sight
Mbps Mega bit per second
MEO Medium Earth Orbiting
mMTC massive Machine Type Communications
MS Mobile Services
MSS Mobile Satellite Services
NGSO Non Geostationary Satellite Orbit
NLOS Non-LOS
NT Non Terrestrial
NTN Non Terrestrial Network
O2I Outdoor-to-Indoor
OBO Output Backoff
OMUX Output MUltipleXer
PA Power Amplifier
PAPR Peak-to-Average Power Ratio
PL Path Loss
POI Point of Interest
PRACH Physical Random Access Channel
PSS Primary Synchronization Signal
RACH Random Access Channel
RAN Radio Access Network
RAR Random Access Response
RAT Radio Access Technology
RTT Round Trip Time
RRH Remote Radio Head
SCS Subcarrier Spacing
SIB System Information Block
SNR Signal-to-Noise Ratio
RMa Rural Macro
RMS Root Mean Square
Rx Receiver
SF Shadow Fading
SOTA Software Over The Air services
SSPA Solid-State Power Amplifier
SSS Secondary Synchronization Signal
TA Timing Advance
TAG Timing Advance Group
TDL Tapped Delay Line
TOA Time Of Arrival
Tx Transmitter
TV Television
UAS Unmanned Aerial System
UE User Equipment
UMa Urban Macro
UMi Urban Micro
URLLC Ultra-Reliable Low Latency Communications
VSAT Very Small Aperture Terminal
XPR Cross-Polarization Ratio
ZOA Zenith angle Of Arrival
ZOD Zenith angle Of Departure
ZSA Zenith angle Spread of Arrival
ZSD Zenith angle Spread of Departure
3GPP
Release 15 13 3GPP TR 38.811 V15.1.0 (2019-06)
4 Non-Terrestrial Networks Overview – background

information
4.1 Roles for Non-Terrestrial Networks in 5G system

Thanks to the wide service coverage capabilities and reduced vulnerability of space/airborne vehicles to physical attacks
and natural disasters, Non-Terrestrial Networks are expected to
- foster the roll out of 5G service in un-served areas that cannot be covered by terrestrial 5G network
(isolated/remote areas, on board aircrafts or vessels) and underserved areas (e.g. sub-urban/rural areas) to
upgrade the performance of limited terrestrial networks in cost effective manner,
- reinforce the 5G service reliability by providing service continuity for M2M/IoT devices or for passengers on
board moving platforms (e.g. passenger vehicles-aircraft, ships, high speed trains, bus) or ensuring service
availability anywhere especially for critical communications, future railway/maritime/aeronautical
communications, and to
- enable 5G network scalability by providing efficient multicast/broadcast resources for data delivery towards the
network edges or even user terminal.
The benefits relate to either Non-Terrestrial networks operating alone or to integrated terrestrial and Non-Terrestrial
networks. They will impact coverage, user bandwidth, system capacity, service reliability or service availability, energy
consumption, connection density (See [5]).
A role for Non-Terrestrial Network components in the 5G system is expected for the following verticals: transport,
Public Safety, Media and Entertainment, eHealth, Energy, Agriculture, Finance, Automotive.
4.2 5G Use Cases wherein Non-Terrestrial Network

components have a role
4.2.1 5G use cases introduction
A use case, typically refers to the interactions between a role and a system, to achieve a specific goal. Hence, it is
necessary to identify the goal of the service enabled by a Non-Terrestrial network component integrated in the 5G
system.
The tables in the clauses after respectively identify for each of the 5G service enablers, the use cases wherein Non-
Terrestrial Network components have a role to play.
- 5G service enablers refer to eMBB (enhanced Mobile Broadband), URLLC (Ultra-Reliable Low Latency
Communications) and mMTC (massive Machine Type Communications).
- 5G use cases correspond to the interactions between a stakeholder (user, operator, service provider) and the 5G
system, to achieve a specific goal.
- The role of the Non-Terrestrial Network refers to services enabled by the Non-Terrestrial Network component in
the 5G system to support the use case.
- 3GPP reference documents are provided in which the use cases are mentioned.
NOTE: While the propagation delay of satellite systems may be an issue for certain applications requiring ultra
low latency, the importance of satellite for Critical Communications including public safety
communications due to their dependability and large coverage area is well known.
3GPP
Release 15 14 3GPP TR 38.811 V15.1.0 (2019-06)
Table 4.2.1-1: 5G use cases for Satellite access networks
5G 5G Use case 5G Use case description Satellite service 3GPP

Service References
enabler
eMBB Multi Users in underserved areas (home or in Small Broadband connectivity to TR 22.864,
connectivity Offices, big events in ad-hoc built-up facilities) cells or relay node in §5.5:
are connected to the 5G network via multiple underserved areas in Backhauling
combination with TR 22.863,
network technologies and benefit from 50
terrestrial wireless/cellular §5.6: Fixed
Mbps+. Delay sensitive traffic may be routed or wire line access Mobile
over short latency links while less delay featuring limited user Convergence
sensitive traffic can be routed over the long throughput.
TR 22.863,
latency links.
§5.7: Femto
cell
TR 22.863,
§5.4: Higher
user mobility
TS 22.261
(related to
§6.3)
eMBB Fixed cell Users in isolated villages or industry premises Broadband connectivity TR 22.863,
connectivity (Mining, off shore platform) access 5G between the core network §5.3:
services and benefit from 50 Mbps+. and the cells in un-served Deployment
areas (isolated areas). and coverage
eMBB Mobile cell Passengers on board vessels or aircrafts Broadband connectivity TR 22.863,
connectivity access 5G services and benefit from 50 between the core network §5.3:
Mbps+. and the cells on board a Deployment
moving platform (e.g. and coverage
aircraft or vessels).
TS 22.261
(related to
§7.1)
eMBB Network Some critical network links requires high Secondary/backup TR 22.862,
resilience availability which can be achieved through the connection (although §5.5: Higher
aggregation of two or several network potentially limited in availability
connections in parallel. capability compared to the
primary network TS 22.261
The intent is to prevent complete network connection). (related to
connection outage. §6.3)
eMBB Trunking A network operator may want to deploy or Broadband connectivity TR 22.863,
restore (disaster relief) 5G service in an between the public data §5.3:
isolated area (not connected to public data network and a mobile Deployment
network). network anchor point or and coverage
between the anchor points
A network operator may want to interconnect of two mobile networks.
various 5G local access network islands not
otherwise connected
eMBB Edge Media and entertainment content such as live Broadcast channel to TR 22.864,
network broadcasts, ad-hoc broadcast/multicast support Multicast delivery §5.4: Efficient
Delivery streams, group communications, Mobile Edge to 5G network edges. content
Computing's Virtual Network Function delivery
updates are transmitted in multicast mode to
a RAN equipment at the network edge where TS 22.261
it may be stored in a local cache or further (related to
distributed to the User Equipment. §6.6)
The intent is to off load popular content from

the mobile network infrastructure (especially
at backhaul level).
3GPP
Release 15 15 3GPP TR 38.811 V15.1.0 (2019-06)

Service References
enabler
eMBB Mobile cell Passengers on board public transport Broadband connectivity TR 22.863,
hybrid vehicles (e.g. high speed/regular trains, combined with terrestrial §5.3:
connectivity buses, river boats) access reliable 5G cellular access to connect Deployment
services. They are served by a base station a cell/group of cells or and coverage
which is connected by a hybrid relay node(s) on board
cellular/satellite connection. The cellular moving platforms. TR 22.862,
connectivity may be intermittent and/or §5.5: Higher
support limited user throughput. availability
TS 22.261
(related to
§7.1)
eMBB Direct To TV or multimedia service delivery to home Broadcast/Multicast TR 22.864,
Node premises or on board a moving platform service to access points in §5.4: Efficient
broadcast homes or on board content
moving platforms. delivery
TS 22.261
(related to
§6.6)
mMTC Wide area Global continuity of service for telematic Connectivity between IoT TR 22.861,
IoT service applications based on a group of devices (battery activated §5.2:
sensors/actuators (IoT devices, battery sensors/actuators or not) connectivity
activated or not) scattered over or moving and spaceborne platform. aspects
around a wide area and reporting information Continuity of service
to or controlled by a central server. across spaceborne TR 22.864,
platforms and terrestrial §5.6: Access
These sensors and/or actuators may be used base stations is needed.
for example the following telematics TR 22.862,
applications: §5.1: Higher
reliability and
- Automotive and road transport: high lower latency
density platooning, HD map updates, Traffic
flow optimisation, Vehicle software updates, TS 22.261
automotive diagnostic reporting, user base (related to
insurance information (e.g. speed limit, §6.2.3 Service
driving behaviour), safety status reporting continuity
(e.g. air-bag deployment reporting), across
advertising based revenue, Context different
awareness information (e.g. neighbouring access
bargain opportunities based on revenue), technologies)
remote access functions (e.g. remote door
unlocking).
- Energy: Critical surveillance of oil/gas
infrastructures (e.g. pipeline status)
- Transport: Fleet management, asset
tracking, digital signage, remote road alerts
- Agriculture: Livestock management,
farming
mMTC Local area Group of sensors that collect local Connectivity between TR 22.863,
IoT service information, connect to each other and report mobile core network and §5.3:
to a central point. The central point may also base station serving IoT Deployment
command a set of actuators to take local devices in a cell or a and coverage
actions such as on-off activities or far more group of cells.
complex actions. TS 22.261
(related to
The sensors/actuators served by a local area §7.1)
network may be located in a smart grid sub-
system (Advanced Metering) or on board a
moving platform (e.g. container on board a
vessel, a truck or a train).
3GPP
Release 15 16 3GPP TR 38.811 V15.1.0 (2019-06)

Service References
enabler
eMBB Direct to Public safety authorities want to be able to Broadcast/Multicast TR 22.864,

mobile instantaneously alert/warn the public (or service directly to User §5.4: Efficient
broadcast specific subsets thereof) of catastrophic Equipment whether content
events and provide guidance to them during handheld or vehicle delivery
the disaster relief while the terrestrial network mounted.
might be down. TR 22.862,
§5.6: Mission
Automotive industry players, are interested to critical
provide instantaneously Firmware/Software services
Over The Air services (FOTA/SOTA) to their
customers wherever they are. This will include TS 22.261
information updates such as map information (related to)
including points of interest (POI), real-time
traffic, weather, and early warning broadcasts
(e.g. floods, earthquakes and other extreme
weather situations, as well as terror attacks),
parking availability, infotainment, etc.
Media and entertainment industry can provide
entertainment services in vehicles (cars,
buses, trucks).
eMBB Wide area Emergency responders, such as police, fire Access to User TR 22.862,
public safety brigade and medical personnel can exchange Equipment (handset or §5.6: Mission
messaging and voice services in outdoor vehicle mounted). critical
conditions anywhere they are and achieve services
continuity of service whatever mobility
scenarios. TS 22.261
(related to)
eMBB Local area Emergency responders, such as police, fire Broadband connectivity TR 22.862,
public safety brigade, and medical personnel can set up a between the core network §5.6: Mission
tactical cell wherever they need to operate. and the tactical cells. critical
This cell can be connected to the 5G system services
via satellite to exchange data, voice and video
based services between the public safety TS 22.261
users within a tactical cell or with the remote (related to)
coordination centre.
3GPP
Release 15 17 3GPP TR 38.811 V15.1.0 (2019-06)
Table 4.2.1-2: 5G use cases for Aerial access networks
5G 5G Use 5G Use case description Aerial access service 3GPP

Service case References
enabler
eMBB Hot spot Users in un/underserved areas (big events) are Broadband connectivity to TR 22.863,
on connected to the 5G network and benefit from cells or relay node in §5.3:
demand 50 Mbps+. un/underserved areas. Deployment
and coverage
TS 22.261
(related to)
eMBB Regional Emergency responders, such as police, fire Access to User Equipment TR 22.862,
area brigades, and medical personnel can exchange (handset or vehicle §5.6: Mission
public messaging, voice and video services in mounted). critical services
safety
indoor/outdoor conditions anywhere they are Adhoc connectivity between TS 22.261
and whatever mobility scenarios. two cells (related to)
eMBB Fixed Users in isolated villages or industry premises Broadband connectivity TS 22.261
cell (Mining, off shore platform) access 5G services between the core network (related to
connecti and benefit from 50 Mbps+. and the cells in un-served §7.1)
vity areas (isolated areas).
TR 22.863,
§5.3:
Deployment
and coverage
4.3 Satellite and aerial access network architecture principles

Non-Terrestrial Network access typically features the following system elements:
- NTN Terminal: It may refer to directly the 3GPP UE or a terminal specific to the satellite system in case the
satellite doesn't serve directly 3GPP UEs.
- A service link which refer to the radio link between the user equipment and the space/airborne platform. In
addition the UE may also support a radio link with terrestrial based RAN.
- A space or an airborne platform embarking a payload which may implement either a bent-pipe or a regenerative
payload configuration:
- A bent pipe payload: Radio Frequency filtering, Frequency conversion and amplification:
- A regenerative payload: Radio Frequency filtering, Frequency conversion and amplification as well as
demodulation/decoding, switch and/or routing, coding/modulation. This is effectively equivalent to having
base station functions (e.g. gNB) on board the space/airborne vehicle
- Inter satellite/aerial links in case of regenerative payload and a constellation of satellites. ISL may operate in RF
frequency or optical bands
- Gateways that connect the satellite or aerial access network to the core network
- Feeder links which refer to the radio links between the Gateways and the space/airborne platform
We shall distinguish between two types of Satellite and Aerial access network
- Broadband access network serving Very Small Aperture Terminals that can be fixed or mounted on a moving
platform (e.g. bus, train, vessel, aircraft, etc.). In this context, Broadband refers to at least 50 Mbps data rate and
even up to several hundred Mbps (satellite) or even up to several Gbps (aerial) on the downlink. The service
links operate in frequency bands allocated to satellite and aerial services (Fixed, Mobile) above 6 GHz.
- Narrow or wide band access network serving terminals equipped with Omni or semi directional antenna (e.g.
handheld terminal). In this context, Narrowband refers to less than 1 or 2 Mbps data rate on the downlink. The
service links operate typically in frequency bands allocated to mobile satellite or aerial services below 6 GHz.
3GPP
Release 15 18 3GPP TR 38.811 V15.1.0 (2019-06)
It is also helpful to distinguish between satellite and aerial systems with inter-satellite links (ISL) or inter-aerial links
(IAL) and those without ISL/IAL.
For Aerial networks, we consider a configuration where base station functions are on board the airborne vehicle. The
purpose of this network component is to provide the 5G service enablers to handheld devices.
Based on these principles, the figures below illustrate possible satellite and aerial access network architectures.
Very Small
Aperture Terminal Spaceborne
Plateform Gateway
Public
Core Data
Service link Feeder link network network
Figure 4.3-1: Satellite access network (without ISL) with a service link operating in frequency bands
above 6 GHz allocated to Fixed and Mobile Satellite Services (FSS and MSS)
Very Small Spaceborne Spaceborne

Aperture Terminal Platform Platform Gateway
Public
Core Data
Service Inter Feeder network network
link Satellite link
link
Figure 4.3-2: Satellite access network (with ISL) with a service link operating in frequency bands
above the 6 GHz allocated to Fixed and Mobile Satellite Services (FSS and MSS)
Handheld Spaceborne
or IoT device Plateform Gateway
Public
Core Data
Figure 4.3-3A: Satellite access network with a service link operating in frequency bands below 6 GHz
allocated to Mobile Satellite Services (MSS)
3GPP
Release 15 19 3GPP TR 38.811 V15.1.0 (2019-06)
Figure 4.3-3B: Satellite access network which service link operates below 6 GHz frequency bands
allocated to Mobile Satellite Services (MSS) and complemented with the terrestrial access network
served by the same or independent core networks.
Handheld or IoT device Airborne

Platform Gateway
Public
Core Data
Figure 4.3-4: Aerial access network (without IAL) with a service link operating in frequency bands
below or above 6 GHz
Handheld Airborne Airborne

or IoT device Platform Platform Gateway
Public
Core Data
link Aerial link link
Figure 4.3-4B: Aerial access network (with IAL) with a service link operating in frequency bands
below or above 6 GHz
Very Small Airborne Airborne

Aperture Terminal Platform Platform Gateway
Public
Core Data
link Aerial link link
Figure 4.3-4C: Aerial access network (with IAL) with a service link operating in frequency bands
above 6 GHz
3GPP
Release 15 20 3GPP TR 38.811 V15.1.0 (2019-06)
It is recommended to select a range of deployment scenarios with either bent pipe or regenerative payloads. Note that
the technical details of implementing the ISL interface is beyond the scope of this study.
4.4 Characteristics of NTN Terminals for satellite / aerial access

network
Table 4.4-1 gives the minimum RF characteristics of the terminals operating respectively in Ka band (e.g. very small
aperture terminals) and in S band (e.g. handheld terminals).
Table 4.4-1: Typical minimum RF characteristics of UE in satellite and aerial access networks
Very Small Aperture Terminal Handheld or IoT devices (3GPP

(fixed or mounted on moving class 3, see [2])
platforms)
Transmit Power 2 W (33 dBm) 200 mW (23 dBm)
Antenna type 60 cm equivalent aperture diameter Omnidirectional antenna (linear
(circular polarisation) polarisation)
Antenna gain Tx: 43.2 dBi Tx and Rx: 0 dBi
Rx: 39.7 dB
Noise figure 1.2 dB 9 dB
EIRP 45.75 dBW -7 dBW
G/T (NOTE 1) 18.5 dB/K -33.6 dB/K
Polarisation (NOTE 2) Circular Linear
NOTE 1: For the computation of G/T or figure of merit, following formula applies in dB:
G/T = Ga – NF – 10*LOG (To+(Ta-To)/(100.1*NF))
Where:
- Antenna Gain : Ga in dBi

- Ambient Temperature : T0 (usually 290 K)
- Antenna temperature : Ta (typically 290 K with 0 dBi gain and 150 K with >30 dBi gain)
- Noise Figure: NF in dB
NOTE 2: For S band, we assume that the User Equipment has an omni-directional antenna of linear polarization,
while the antenna on board space-borne or airborne platforms features typically employs circular
polarization. Hence a polarization mismatch of 3 dB has to be taken into account for the radio link
budget computation. This will impact the UE RF characteristics as below:
- Equivalent EIRP of 20 dBm (-10 dBW) under satellite coverage.

- Equivalent G/T of -36,6 dB/K under satellite coverage.
Note that other performance may be considered.
3GPP
Release 15 21 3GPP TR 38.811 V15.1.0 (2019-06)
4.5 Air/Space borne vehicle characteristics

Table 4.5-1 provides characteristics of aerial and satellite vehicles that are relevant for the purpose of the study item:
Table 4.5-1: Typical characteristics of Airborne or Space-borne vehicles
Characteristics Geostationary satellites Non-Geostationary Airborne platforms

satellites
Altitude 35 786 km Low Earth Orbiting Typically from 8 to 50 km
satellites: From 600 km up
to 1500 km
Medium Earth Orbiting
satellites: From 7000 up to
20000 km
Motion Typically within a cube of 50-100 We shall assume here only Typically in motion within
km side around the theoretical circular orbits around the TBD km from the notional
orbital position fixed in terms of earth station keeping position
elevation/azimuth with respect to a fixed in terms of
given earth point elevation/azimuth with
respect to a given earth
point
Elevation angle Typically more than 10° for user terminal and more than 5° for gateways
(NOTE 1)
NOTE 1: The minimum Elevation angle refers to the minimum angle under which the airborne/spaceborne platform
can be seen by a terminal. Below is a summary table of minimum elevation angles for different types of
satellite and aerial based systems applications.
The characteristics of the air/space-borne vehicles create specific Doppler and propagation delay conditions that NR has
to cope with.
Table 4.5-2: Typical elevation angles in aerial and satellite based systems
Satellite & aerial Typical minimum Elevation Rationale/remarks

Systems Angle for terminals
International GEO 5 degrees Serving earth stations equipped with very large antennas
(Trunking)
Regional GEO 10 degrees Addressing regions in lower and medium latitude
International (GEO) 5 degrees Addressing large ships
Maritime
Aeronautical 20 degrees Taking into account aero-dynamic constraints prevents
operation at lower angles
Vehicles 15 degrees Taking into account road conditions, terrain, and vehicle
mechanics
Non GSO 10 to 30 degrees Ensuring service continuity optimising the number of
satellites
Aerial In the range of 10 degrees Maximising the service area
4.6 Coverage pattern of NTN

Satellite or aerial vehicles typically generate several beams over a given area. The foot print of the beams are typically
elliptic shape.
The beam footprint may be moving over the earth with the satellite or the aerial vehicle motion on its orbit.
Alternatively, the beam foot print may be earth fixed, in such case some beam pointing mechanisms (mechanical or
electronic steering feature) will compensate for the satellite or the aerial vehicle motion.
Table 4.6-1: Typical beam foot print size
Attributes GEO Non-GEO Aerial

Beam foot print 200 – 1000 km 100 – 500 km 5 - 200 km
size in diameter
3GPP
Release 15 22 3GPP TR 38.811 V15.1.0 (2019-06)
Typical beam patterns of various NTN access networks are depicted below:
Transparent (bent-pipe) Non Transparent (On Board processor)

satellite/HAPS satellite/HAPS
5G
NR via RAN NGc/NGu over
frequency f2 frequency f2
NR via NR via
frequency f1 frequency f1
Gateway
Gateway
5G
RAN
5G 5G
CN CN
Figure 4.6-1: NTN Beam patterns
4.7 Non-Terrestrial Network architecture options

The possible options of NTN architecture in 5G context based on the RAN architecture principles described in [3] are
shown below:
NGc &
NGu NG6 Data
UE gNB NGC network
Uu
Figure 4.7-1: NTN featuring an access network serving UEs and based on a satellite/aerial with bent
pipe payload and gNB on the ground (Satellite hub or gateway level)
In figure 4.7-1, the satellite or the aerial will relay a "Satellite friendly" NR signal between the gNB and the UEs in a
transparent manner.
NGc &
NGu NG6 Data
UE gNB NGC network
Uu
Figure 4.7-2: NTN featuring an access network serving UEs and based on a satellite/aerial with gNB
on board
In figure 4.7-2, the satellite or the aerial includes full or part of a gNB to generate/receive a "Satellite friendly" NR
signal to/from the UEs. This requires sufficient on board processing capabilities to be able to deploy gNB or Relay
Node functions.
3GPP
Release 15 23 3GPP TR 38.811 V15.1.0 (2019-06)
NGc &
Relay NGu NG6 Data
UE gNB NGC network
Uu Node Un
Figure 4.7-3: NTN featuring an access network serving Relay Nodes and based on a satellite/aerial
with bent pipe payload
In figure 4.7-3, the satellite or the aerial will relay a "Satellite friendly" NR signal between the gNB and the Relay
Nodes in a transparent manner.
NGc &
Relay NGu NG6 Data
UE gNB NGC network
Uu Node Un
Figure 4.7-4: NTN featuring an access network serving Relay Nodes and based on a satellite/aerial
with gNB
In figure 4.7-4, the satellite or the aerial includes full or part of a gNB to generate/receive a "Satellite friendly" NR
signal to/from the Relay Nodes. This requires sufficient on board processing capabilities to be able to deploy gNB or a
Relay Node functionality.
NOTE: In the figures above a satellite represents both satellite and aerial platforms.
Table 4.7-1: 5G system elements mapping in NTN architecture
5G elements - NTN elements mapping

NTN architecture options NTN Terminal Space or HAPS NTN Gateway
A1: access network serving UE Remote Radio Head gNB
UEs via bentpipe (Bent pipe relay of Uu radio
satellite/aerial interface signals)
A2: access network serving UE gNB or Relay Node Router interfacing to Core
UEs with gNB on board functions network
satellite/aerial
A3: access network serving Relay Node Remote Radio Head gNB
Relay Nodes via bent pipe (Bent pipe relay of Uu radio
satellite/aerial interface signals)
A4: access network serving Relay Node gNB or Relay Node Router interfacing to Core
Relay Nodes with gNB on functions network
board satellite/aerial
4.8 Spectrum
Satellite and Aerial systems operate in allocated frequency bands as per ITU-R/national allocation regime.
3GPP
Release 15 24 3GPP TR 38.811 V15.1.0 (2019-06)
5 Non-Terrestrial Networks deployment scenarios
5.1 Scenarios overview

Based on the NTN overview background information presented in clause 4 and in examples of deployment scenarios,
defined in the TR 38.913 [5] under clause 6.1.12 "Satellite extension to Terrestrial", the following Non-Terrestrial
Network (NTN) reference deployment scenarios are down selected and will be further detailed in the study and used for
the evaluations:
Table 5.1-1: Reference Non-Terrestrial Network Deployment scenarios to be considered in the NR-
NTN study
Main attributes Deployment-D1 Deployment-D2 Deployment-D3 Deployment-D4 Deployment-D5

Platform orbit and GEO at 35 786 km GEO at 35 786 km Non-GEO down to Non-GEO down to UAS between 8
altitude 600 km 600 km km and 50 km
including HAPS
Carrier Frequency Around 20 GHz for Around 2 GHz for Around 2 GHz for Around 20 GHz for Below and above
on the link between DL both DL and UL (S both DL and UL (S DL 6 GHz
Air / space-borne Around 30 GHz for band) band) Around 30 GHz for
platform and UE UL (Ka band) UL (Ka band)
Beam pattern Earth fixed beams Earth fixed beams Moving beams Earth fixed beams Earth fixed beams
Duplexing FDD FDD FDD FDD FDD
Channel Bandwidth Up to 2 * 800 MHz Up to 2 * 20 MHz Up to 2 * 20MHz Up to 2 * 800 MHz Up to 2 * 80 MHz
(DL + UL) in mobile use and
2 * 1800 MHz in
fixed use
NTN architecture A3 A1 A2 A4 A2
options (See clause
4)
NTN Terminal type Very Small Aperture Up to 3GPP class Up to 3GPP class Very Small Up to 3GPP class
Terminal (fixed or 3 UE [2] 3 UE [2] Aperture Terminal 3 UE [2]
mounted on Moving (fixed or mounted Also Very Small
Platforms) on Moving Aperture Terminal
implementing a relay Platforms)
node implementing a
Relay node
NTN terminal 100% Outdoors 100% Outdoors 100% Outdoors 100% Outdoors Indoor and
Distribution Outdoor
NTN terminal up to 1000 km/h (e.g. up to 1000 km/h up to 1000 km/h up to 1000 km/h up to 500 km/h
Speed aircraft) (e.g. aircraft) (e.g. aircraft) (e.g. aircraft) (e.g. high speed
trains)
Main rationales GEO based indirect GEO based direct Non-GEO based Non-GEO based Support of low
access via relay access direct access indirect access via latency services
node relay node for 3GPP mobile
UEs, both indoors
and outdoors
Supported Uses 1/ eMBB: multi- 1/eMBB: Regional 1/eMBB: Regional 1/ eMBB: multi- 1/ eMBB: Hot spot
cases, see clause 4 connectivity, fixed area public safety, area public safety, homing, fixed cell on demand
cell connectivity, Wide area public Wide area public connectivity,
mobile cell safety, Direct to safety, Wide area mobile cell
connectivity, network mobile broadcast, IoT service connectivity,
resilience, Trunking, Wide area IoT network resilience,
edge network service Trunking, Mobile
delivery, Mobile cell cell hybrid
hybrid connectivity, connectivity
Direct To Node
multicast/ broadcast
The scenarios attributes used in the Table 5.1-1 are described below.
Only the main attributes are discussed in this clause, for simplification purposes. Complementary attributes that should
be set up for each scenario are described in the clause 7.
The use cases mentioned in the Table 5.1-1, are described in clause 4.2.1.
3GPP
Release 15 25 3GPP TR 38.811 V15.1.0 (2019-06)
5.2 Attributes
Platform orbit and altitude
This attribute stands for the Platform orbit type (GEO, Non-GEO) and its altitude.
A platform is either a satellite (alias space-borne vehicle), or a HAPS (airborne vehicle).
See clause 4.5 for further characteristics of Air / space borne vehicles.
Carrier frequency between air / space-borne platform and UE
The study addresses the whole frequency range between 0.5 – 100 GHz. For channel modelling and for the
identification of areas of impact on the NR , the following frequency bands will be in particular considered:
- For VSAT, Ka band: Downlink: 19.7 - 21.2 GHz, Uplink: 29.5 – 30.0 GHz
- For UE, S band: Downlink: 2170 - 2200 MHz, Uplink: 1980 - 2010 MHz
The UE characteristics are described in clause 4.4.
Beam pattern
Beam pattern stands for "beam coverage pattern". It is described in clause 4.6.
Access scheme
For these scenarios, the FDD (Frequency Division Duplexing) is selected, versus TDD (Time Division Duplexing).
FDD means that the transmitter and the receiver operate at different carrier frequencies. Uplink and downlink sub-bands
are separated by the named frequency offset.
Channel Bandwidth (DL + UL)
This scenario attribute stands for the available bandwidth for channels, for DL and for UL. It depends on the used
carrier frequencies. For evaluation purposes, we will consider:
- For Satellite and aerial networks operating in frequency bands above 6 GHz, the bandwidth is up to 800MHz on
both Downlink and Uplink
- For Satellite and aerial networks operating in frequency bands below 6 GHz, the bandwidth is up to 80MHz on
both Downlink and Uplink
NTN architecture options
See clause 4.7.
NTN terminal type
For evaluation purposes:
- The VSAT transmit power will be set to 33dBm (2W), with a 60 cm equivalent aperture diameter (circular
polarisation).
- For each 3GPP FDD power class (PC), the maximum output Power is: 33dBm (2W) for PC 1, 27dBm (0.5W)
for PC 2 and 23dBm (0.20W) for the PC 3, with an omnidirectional antenna. In a 1st approach and for
evaluation purposes, the PC 3 UE will have a Transmit Power set to 23dBm (0.20W).
Relay node is defined in clause 3.1.
See clause 4.4 for further characteristics of UE of satellite / aerial access network.
NTN terminal distribution
This attribute is set to either:

- 100% outdoors UEs,
- 100% indoors UEs,
- Or mixed indoors & outdoors UE distribution
3GPP
Release 15 26 3GPP TR 38.811 V15.1.0 (2019-06)
NTN terminal Speed
This attribute is a generic term relative to the transmitter/receiver which is on board the satellite or aerial platform. It
stands for:
- High speed / low speed UE
- High speed / low speed platform (such as Trains, Boats embedding base stations )
For evaluation purposes, the selected maximum values are:

- 1000 km/h (e.g. aircraft)
- 500 km/h (e.g. high speed trains)
The effect of the maximum NTN terminal speed as well as satellite or aerial motion are considered in each deployment
scenario.
See clause 4.5 for further characteristics on air / space borne platform.
5.3 Doppler and Propagation delay characterisation

5.3.1 Methodology
We shall distinguish between geostationary satellite, non geo stationary satellite and HAPS platforms.
5.3.1.1 Propagation delay

We consider the one way propagation delay as the delay:
- from the Gateway to the UE via the space/airborne platform (bent pipe payload)
- from the space/airborne platform to the UE (regenerative payload)
The Round Trip Time corresponds to the two way propagation delay:
- from the Gateway to the UE via the space/airborne platform and return (bent pipe payload)
- from the space/airborne platform to the UE and return (regenerative payload)
For the propagation delay analysis, we consider a minimum gateway elevation angle of 5° (the elevation angle of the
space/air borne platform from the Gateway). While the minimum terminal elevation angle is typically 10°.
The actual propagation delay depends on the space/airborne platform altitude and respective position of the gateway
and terminal.
5.3.1.2 Differential delay

The differential delay corresponds to the difference of propagation delays between two chosen points that are at some
specific positions within the beam foot print: for example the points can be selected at nadir and Edge of Coverage.
The path to gateway is likely to be the same for all terminals, but this just to simplify the computation.
5.3.1.3 Doppler shift/variation

Doppler shift: Shift of the signal frequency due to the motion of the receiver, the transmitter or both.
Doppler variation rate: During time, the Doppler shift is evolving. This is called the Doppler variation rate or simply
Doppler rate.
The Doppler shift and Doppler variation depend on the relative speed of the space/airborne platforms, the speed of the
UE, and the carrier frequency.
The figure below recalls the basic geometry of the system. The carrier frequency of the signal received at satellite is
affected by the motion of the transmitter. The Doppler on the signal received by the UE from the satellite is also
3GPP
Release 15 27 3GPP TR 38.811 V15.1.0 (2019-06)
impacted by the motion of the UE in addition to the motion of the satellite or HAPS. Since both the air/spaceborne
vehicle and the UE are moving relatively to the Earth, their respective effects can be added algebraically.
In the present document, we can follow a non-relativistic approach to compute the Doppler shift and variation rate given
that the ratio between the relative speed of the transmitter or receiver and the light velocity is negligible.
- For a terminal at a speed of 1000 km/h (or 0.277 km/s): this ratio is 0.277/300000 = 0.00009
- For a relative speed of Non-Geo satellite of 7.5 km/s (orbital speed): the ratio is 0.000025
The signal received by the satellite at a nominal carrier frequency Fo is affected by a Doppler shift.
The Doppler shift is computed with the formula:
Doppler shift formula: ∆F = Fo*V* cos (θ)/c
Where
- Fo: nominal carrier frequency
- V = UE velocity
- Θ is the angle between the velocity vector V of the mobile (Transmitter or receiver ) and the direction of
propagation of the signal between the UE and the space/airborne platform.
Angle θ
Figure 5.3.1.3-1: Definition of the theta angle between space/airborne platform and the direction of
the UE in motion
When a transmitter is moving away from a receiver, ∆F is negative.
When a transmitter is moving towards a receiver, ∆F is positive.
The Doppler shift variation corresponds to the variation of the Doppler shift over time. In other words, it refers to the
derivative of the Doppler shift function of time.
5.3.2 Geo-stationary platforms

Geostationary platform are orbiting at 35786 km altitude in the equatorial plan, and is fixed with respect to the earth.
Nevertheless, due to some imperfection of the terrestrial potential, the satellite has some motion around its orbital
position, as described in a further section.
3GPP
Release 15 28 3GPP TR 38.811 V15.1.0 (2019-06)

One should distinguish between
- Bent pipe payloads
- One way propagation delay is the sum of feeder link propagation delay and user link propagation delay, thus
the propagation delay between Gateway and UE via the satellite
- Round Trip Time is the delay over the path, Gateway-Satellite-UE-Satellite-Gateway. It corresponds to twice
the one way propagation delay
- Regenerative payloads (decoding/coding on board)
- One way propagation delay is the propagation delay between the satellite and the UE
- Round Trip Time is the delay over path: satellite-UE-satellite
In both cases the transit time and/or processing time are not taken into account.
For the propagation delay computation, the min elevation angle is set at 5° for the Gateway, and is set at 10°for the
terminal can be set at various elevation angles, but we consider that the worst case is 10° elevation angle.
The following table summarises the different situation and the different distances in km and the different propagation
delays in ms.
Table 5.3.2.1-1: Propagation delays for GEO satellite at 35786 km
GEO at 35786 km
Elevation angle Path D (km) Time (ms)
UE :10° satellite - UE 40586 135.286
GW : 5° satellite - gateway 41126.6 137.088
90° satellite - UE 35786 119.286
Bent Pipe satellite
One way delay Gateway-satellite_UE 81712.6 272.375
Round trip Time Twice 163425.3 544.751
Regenerative Satellite
One way delay Satellite -UE 40586 135.286
Round Trip Time Satellite-UE-Satellite 81172 270.572

In this clause, we compute the differential delays between specific positions: for instance at nadir and Edge of Coverage.
The path to gateway is supposed to be the same for all UEs.
Table 5.3.2.2-1: Differential Delay for GEO satellite
GEO at 35786 km
Delta D (km) Delta Time (ms)
Differential One way delay between nadir and EOC paths 4800 16
Percentage of the difference compared to maximum delay
5.9 %
(bent pipe)
Percentage of the difference compared to maximum delay
11.9 %
(regenerative satellite)
For Geostationary satellites we have also taken a satellite located at 10 ° E and we have computed different differential
delays between some points, provided all points were linked to the same Gateway.
The table is valid for both bent pipe satellite and regenerative satellite.
3GPP
Release 15 29 3GPP TR 38.811 V15.1.0 (2019-06)
Table 5.3.2.2-2: Differential delay examples
Duo of Cities Delta Time(ms)

Paris-Marseille 1.722
Lille-Toulouse 2.029
Brest-Strasbourg 0.426
Oslo-Tromsoe -3.545
Oslo-Svalbard -6.555
Oslo-Paris 3.487
5.3.2.3 Doppler shift

In principle, the Geostationary satellite is fixed and therefore no Doppler shift is induced except that due to possible UE
motion.
In reality the satellite is moving around its nominal orbital position, due to perturbations (e.g. sun, moon) and to non-
spherical component earth attraction which impact the earth gravitational force.
The satellite must be kept to inside a box described here below. The satellite is typically maintained inside a box that
has the following dimensions, by thrust or plasmic propulsion.
Figure 5.3.2.3-1: Trajectory box for a Geostationary satellite
Without maintaining the satellite inside the box, the motion could be a higher value like inclination up to +/6 °.
We take the hypothesis that the satellite is kept in the limited box. The trajectory that the satellite follows is an "8" as
shown in next figure. The plane is seen from the centre of the Earth. The blue arrows in the next figure indicate the
sense of motion of the satellite around its orbital position So.
3GPP
Release 15 30 3GPP TR 38.811 V15.1.0 (2019-06)
S1 S2
N
So
E W Equatorial plan
S4 S3
Figure 5.3.2.3-2: Geostationary satellite trajectory
In the figure above, the trajectory of the satellite is represented as seen from a point at the equator having same
longitude as the nominal Geostationary satellite orbital position.
A geostationary satellite will cover the whole trajectory in 24 Hours.
The average tangential velocity of a geostationary satellite with respect to an earth point is around 2.74 m/s.
Satellite maintained in a typical trajectory box (see figure 2) characterized by
- +/- 37.5 km in both latitude and longitude directions corresponding to an aperture angle of +/- 0.05 °.
- +/- 17.5 km in the equatorial plane.
To illustrate the Doppler shift computation using the Doppler Shift formula, we shall consider some concrete cases:
- Geostationary Satellite at 10 ° E (over Europe)
- First a High Speed Train (500 km/h) from Paris to Lille (France) and from Paris to Strasbourg (France)
- Secondly an air plane (1000 km/h and 10 km altitude) moving in the same directions.
We first compute the Doppler shift while supposing the satellite has no motion, but on a moving UE, and secondly we
evaluate the impact of the satellite motion on a fixed UE.
First case: Satellite is considered fixed with respect to earth point
In the high speed train going north from Paris the obtained Doppler shift is provided here below
Table 5.3.2.3-1: Examples of Doppler shift with GEO and a terminal on board a High Speed Train in
opposition direction
Frequency 2GHz 20 GHz 30 GHz

Doppler shift (Hz) -707 -7074 -10612
3GPP
Release 15 31 3GPP TR 38.811 V15.1.0 (2019-06)
Table 5.3.2.3-2: Example of Doppler shift with GEO and a terminal on board an aircraft in opposition
direction

Doppler shift (Hz) -1414 -14149 -21224
And going from Paris to east.
Table 5.3.2.3-3: Example of Doppler shift with GEO and in High Speed Train

Doppler shift (Hz) in
Paris 147 1478 2217
Doppler shift (Hz) in
Lille 138 1383 2075
Table 5.3.2.3-4: Example of Doppler shift with GEO and in a plane

Doppler shift (Hz) 295 2956 4434
The Doppler shift is highest for a platform moving along longitude compared to a moving platform along latitude.
In the figure below, we consider a train travelling North from the Equator at a constant speed of 500 km/h. The two
extreme points in the figure are 3600 seconds apart. The Doppler shift plotted as a function of the latitude. The
maximum rate of change of Doppler shift is approximately -23 mHz/s
Figure 5.3.2.3-3: Doppler Shift at 2 GHz for a High Speed Train travelling along a longitude (North
direction)
3GPP
Release 15 32 3GPP TR 38.811 V15.1.0 (2019-06)
In the figure below, we consider an aircraft travelling North from the Equator at a constant speed of 1000 km/h.
Similarly to the previous plot, the two extreme points are 1800 seconds apart. The Doppler shift plotted as a function of
the latitude. The maximum rate of change of Doppler shift is approximately -90 mHz/s.
Figure 5.3.2.3-4: Doppler Shift at 2 GHz for an aircraft travelling north
However these values are reached only when cos (θ) is equal to 1 or -1. Where θ is the angle between the vector speed
and the direction of wave propagation (the axis UE-Satellite).
Second case: Satellite is moving on its "8"
When satellite is moving from S2 to S1, the Doppler shift in Paris is the following
Table 5.3.2.3-4:Example of Doppler shift when satellite is moving

Doppler shift (Hz) -0.25 -2.4 -4.0
When satellite is moving from S1 to S4, the Doppler shift in Paris is the following
Table 5.3.2.3-5:Example of Doppler shift when satellite is moving

Doppler shift (Hz) 2.25 22.5 34
Doppler shift is higher than in the previous case, though still very low compared to the case of Non GEO satellites.
When satellite is in near GEO orbit with inclination up to 6°, the Doppler shift can reach around 300 Hz at 2 GHz, then
3000 Hz at 20 GHz and 4500 Hz at 30 GHz, which are still compatible with OFDM.
5.3.3 Aerial vehicle

The altitude of aerial vehicle also called UAS (Unmanned Aircraft Systems) can be between 8 and 50 km, including
HAPS. HAPS is a station located on an aerial object at an altitude of 20 to 50 km and at a specified, nominal, fixed
point relative to the Earth.
3GPP
Release 15 33 3GPP TR 38.811 V15.1.0 (2019-06)
The general UAS based system architectures are presented in the following picture.
Figure 5.3.3-1: UAS based system architectures
The coverage can be divided into small cells, and usually the minimum elevation angle for a Mobile Terminal is 5°.
The platform can move around its nominal position within a few kilometres at a maximum tangential velocity of 15 m/s.
This can result in maximum Doppler shift (in absolute value) of 100 Hz @ 2GHz, 1000 Hz @ 20 GHz and 1500 Hz @
30 GHz due to Haps motion.
In S band, at 100 km/h a car will suffer a Doppler shift of +/- 185 Hz maximum
The Doppler variation can be evaluated with a car moving from one edge to another at 100 km/h , covering 450 km .
The Doppler variation will be around -0.0025 Hz/s. So no impact on demodulation
At 5° elevation angle the distance to the aerial vehicle is 229 km, and if we suppose the Gateway is at same distance
- One way delay is around 1.526 ms
- Round Trip Time is 3.053 ms
- Differential delay between nadir and Edge of Coverage: 0.697 ms
5.3.4 Non geostationary satellites

The Non-Geostationary case comprises three examples
- LEO at 600 km
- LEO at 1500 km
- MEO at 10000 km

In the case of bent pipe satellites, one way propagation delay is the sum of feeder link propagation delay and user link
propagation delay, thus the propagation delay between Gateway and UE.
In the case of regenerative satellite, one way propagation delay is the satellite to UE propagation delay.
In both cases the transit time and/or processing time are not taken into account.
In the case of bent pipe satellite, the Round Trip Time is the physical path duration of the path : Gateway-Satellite-UE-
Satellite-Gateway, that is in fact twice the one way propagation delay.
In the case of regenerative satellite, the round trip delay is the delay corresponding to the following path :satellite-UE-
satellite.
In the computation, Gateway is set at 5° (TBC) elevation angle, and terminal can be set at various elevation angles, but
we consider that the reference case is 10° elevation angle for the propagation delay computation.
3GPP
Release 15 34 3GPP TR 38.811 V15.1.0 (2019-06)
The following table summarises the different situations and the different distances in km and the different propagation
delays in ms.
The results for the three cases of NGSO satellites are summarized in the next table.
Table 5.3.4.1-1: Propagation delays for different NGSO satellites (altitude and payload types)
LEO at 600 km LEO at 1500 km MEO at 10000 km

Elevation Distance D Distance D Distance D
angle Path (km) Delay (ms) (km) Delay (ms) (km) Delay (ms)
UE: 10° satellite - UE 1932.24 6,440 3647.5 12,158 14018.16 46.727
GW: 5° satellite - gateway 2329.01 7.763 4101.6 13.672 14539.4 48.464
90° satellite - UE 600 2 1500 5 10000 33.333
Bent pipe satellite
One way
4261.2 14.204 7749.2 25.83 28557.6 95.192
delay Gateway-satellite_UE
Round
8522.5 28.408 15498.4 51.661 57115.2 190.38
Trip Delay Twice
Regenerative satellite
One way
1932.24 6.44 3647.5 12.16 14018.16 46.73
delay Satellite -UE
Round
3864.48 12.88 7295 24.32 28036.32 93.45
Trip Delay Satellite-UE-Satellite

In this clause, we compute the differential delays between some specific positions : for instance at nadir and Edge of
Coverage.
The path to gateway is supposed to be the same for all terminals.
Table 5.3.4.2-1: Differential Delay for LEO satellite
LEO 600 km LEO 1500 km MEO 10000 km

Delta Distance Delta Delay Delta Distance Delta Delay Delta Distance Delta Delay
Differential One way
delay between nadir 1332.2 km 4.44 ms 2147.5 km 7.158 ms 4018.16 km 13.4 ms
and EOC paths
Percentage of the
maximum delay (bent 31.26 % 27.8 % 14.1 %
pipe)
Percentage of the
maximum delay
67 % 58.9 % 28.7 %
(regenerative
satellite)
5.3.4.3 Doppler Shift and variation rate

The following picture summarizes the methodology used for Non Geostationary systems. We evaluate the Doppler shift
which is maximum when the UE is located in the orbital plane.
3GPP
Release 15 35 3GPP TR 38.811 V15.1.0 (2019-06)
Figure 5.3.4.3-1: System Geometry for Doppler computation
The picture describing the system geometry shows the satellite "S" on its circular orbit. The vector V corresponds to the
orbital speed vector.
The Doppler shift is computed for a Mobile Terminal "M" located in the orbital plan, and corresponds to the maximum
value.
One of the impacting factors on the Doppler shift value is the angle between and the speed vector . Angle called θ.
- The satellite is at an altitude h, and R is the earth radius.
- The satellite has the velocity , and the transmitted frequency is Fc.
- The Doppler shift value Fd due to satellite motion is expressed by the formula
- Where
- θ is the angle between satellite velocity and the radius to the earth SO.
- u the angle between and
- is the vector between earth centre and point on earth
- is the vector between earth centre and the satellite
- angle u is varying with the satellite motion: u(t) = V*t/(R+h) with t the time
- α is the elevation angle to the satellite of the UT in M.
- γ can be computed as follow:
- The Doppler formula is obtained after some computation.
At this altitude the speed of the satellite in circular orbit is 7.5622 km.s -1. So we can use non-relativistic approximation
to compute Doppler shift.
3GPP
Release 15 36 3GPP TR 38.811 V15.1.0 (2019-06)
Also at first order we neglect the speed of earth which is 327 m/s at 45° latitude and 464 m/s at the equator.
For all Non GSO cases, the satellite speeds are the following
- At 600 km : V =7.5622 km.s-1
- At 1500 km : V =7.1172 km.s-1
- At 600 km : V =4.9301 km.s-1
5.3.4.3.1 Case at 2 GHz

Both (Downlink) D/L and (Uplink) U/L the signal is around 2GHz and we limit the curves at 2 GHz.
If we consider now a moving UE at 1000 km/h, and moving in the same direction than the satellite, we have determined
the worst case impact in the following graph. We can define the bounds by adding the Doppler shift due to the satellite
motion and the Doppler shift due to the UE motion.
All the curves are gathered in the next graphs, showing clearly the boundaries of the Doppler shift depending on the
sense of motion between the satellites and the UE.
Three NON GSO satellites cases are provided.
Figure 5.3.4.3.1-1: Case with 2 GHz signal at 600 km on D/L and U/L: fixed UE and UE in motion
3GPP
Release 15 37 3GPP TR 38.811 V15.1.0 (2019-06)
Figure 5.3.4.3.1-2: Case with 2 GHz signal at 1500 km: fixed UE and UE in motion
Figure 5.3.4.3.1-3: Case with 2 GHz signal at 10000 km: fixed UE and UE in motion
5.3.4.3.2 Case in Ka band

There are two cases
- Downlink at 20 GHz
- Uplink at 30 GHz
If we consider now a moving UE at 1000 km/h, and moving in the same direction than the satellite, we have determined
the worst case impact in the following graph.
This impact is at maximum 18 kHz in one sense or the other at 20 GHz, and 27 kHz at 30 GHz, all the curves are
gathered in the next graph, showing clearly the boundaries of the Doppler shift depending on the sense of motion
between the satellite and the UE.
3GPP
Release 15 38 3GPP TR 38.811 V15.1.0 (2019-06)
Figure 5.3.4.3.2-1: Case with 20 GHz signal at 600 km on D/L: fixed UE and UE in motion
Figure 5.3.4.3.2-2 Case with 30 GHz signal at 600 km on D/L: fixed UE and UE in motion
3GPP
Release 15 39 3GPP TR 38.811 V15.1.0 (2019-06)
3GPP
Release 15 40 3GPP TR 38.811 V15.1.0 (2019-06)
The different cases are summarized here below with the ratio of maximum Doppler shift in absolute value to the central
frequency.
3GPP
Release 15 41 3GPP TR 38.811 V15.1.0 (2019-06)
Table 5.3.4.3.2-7: Summary of Doppler shift and shift variation for different altitudes
Max Doppler shift

Frequency (GHz) Max Doppler Relative Doppler variation
2 +/- 48 kHz 0.0024 % - 544 Hz/s LEO at 600 km altitude
20 +/- 480 kHz 0.0024 % -5.44 kHz/s
30 +/- 720 kHz 0.0024 % -8.16 kHz/s
2 +/- 40 kHz 0.002 % -180 Hz/s LEO at 1500 km
20 +/- 400 kHz 0.002 % -1.8 kHZ/s altitude
30 +/- 600 kHz 0.002 % -2.7 kHz/s
2 +/- 15 kHz 0.00075 % -6 Hz/s MEO at 10000 km
20 +/- 150 kHz 0.00075 % -60 Hz/s altitude
30 +/- 225 kHz 0.00075 % -90 Hz/s
Note that the Maximum Doppler shift variation in absolute value is always negative and observed when the Doppler
shift is zero.
5.3.4.4 Doppler Shift and variation rate

The following picture summarizes the methodology used to compute Doppler shift for Non Geostationary satellite
systems. It assumes a Cartesian coordinate system such that the moving satellite and the receiver are on the y-z plane.
The Doppler shift experienced by a stationary receiver can be computed as follows as a function of time:
where f0 is the carrier frequency, d(t) is the distance vector between the satellite and the receiver, and xSAT(t) is the
vector of the satellite position. These vectors can be expressed as:
where RE is the Earth radius, h is the satellite altitude, and is the satellite angular velocity.
Figure 5.3.4.4-1: System geometry for Doppler computation (Satellite moves in the Y-Z plane)
After some mathematical manipulation [7], the Doppler shift as a function of the elevation angle is computed in a
closed-form expression as follows:
3GPP
Release 15 42 3GPP TR 38.811 V15.1.0 (2019-06)
where the angular velocity is , with G the gravitational constant and ME the Earth mass.
If the receiver is placed on board an aircraft or a high speed train, there will be an additional term of Doppler shift
resulting from its own velocity. In case of Non Geostationary satellites, the Doppler shift due to satellite movement is
much higher than the one caused by UE movement. For GEO and HAPS, the Doppler shift component is mainly caused
by the UE movement.
5.3.5 Synthesis for each scenarios

Following table summarises the different Doppler shift, Doppler Shift variation and propagation delays. The main
impairments are the Doppler shifts and Doppler variation in deployment scenarios D3 and D4. (Non-geostationary
platforms at 600 km).
Table 5.3.5-1: Summary of Doppler shift, Doppler Shift variation and propagation delay for LEO at 600
km, GEO and HAPS
Deployment- Deployment- Deployment- Deployment- Deployment- Cellular ( 10 km

D1 D2 D3 D4 D5 Radius)
Platform GEO at 35 786 GEO at 35 Non-GEO Non-GEO Airborne
orbit and km 786 km down to 600 down to 600 vehicle up to
altitude when km km 20 km
relevant
Frequency Ka band S band S band Ka band S band S band
band (Below 6
GHz)
Max One way Bentpipe: 272.37 ms 14.204 ms 14.204 ms 1.526 ms 0.03333 ms
Propagation 272.37 ms
delay (ms) gNB on board:
135.28 ms
Max 16 (between 16 (between 4.44 4.44 0.697 0.00333(between
Differential Edge of Edge of (between (between (between cell centre and cell
delay (ms) satellite satellite Edge of Edge of Edge of edge) equal to
coverage and coverage and satellite satellite satellite maximum delay
Nadir) Nadir) coverage and coverage and coverage and
Nadir) Nadir) Nadir)
Max Doppler For plane For plane +/- 48 kHz @20 GHz : @ 2 GHz: +/- In case of UE on
shift in kHz @ 20 GHz: +/- 1.851 kHz @ +/- 480 kHz 100 Hz board a high speed
18.51 kHz 20 GHz @30 GHz : mainly due to train:
@30 GHz: +/- +/- 720 kHz platform +/- 925 Hz
27.7 kHz motion
% of the 10-4 % 10-4 % 0.0024% 0.0024%
carrier
frequency
(Ratio of
Doppler Shift
over the
central signal
frequency
Max Doppler Negligible Negligible -544 Hz/s @ -5.44 kHz/s Negligible Negligible
variation in 2 GHz @ 20Ghz
Hz/s. (Downlink)
-8.16 kHz/s
@30 GHz
(uplink)
NOTE : In some cases like UE on board an aircraft during taking off, the acceleration can add a supplementary
Doppler variation in absolute value of respectively 262 Hz/s @ 20 Ghz, and 393 Hz/s @ 30 GHz.
3GPP
Release 15 43 3GPP TR 38.811 V15.1.0 (2019-06)
6 Non-Terrestrial Networks channel models
6.1. Status/expectation of existing information for satellite/HAPS

channels
6.1.1 Channel modeling works outside of 3GPP
ITU recommendations are encompassing most recent works and measurements on satellite channel models.
- ITU-R P.681 [10] defines the Land Mobile Satellite channel with measurements up to 20 GHz
- ITU-R P.618 [11] describes atmospheric effects such as gas attenuation, scintillation, rain and cloud attenuation.
6.1.2 Targeted user environment

Only outdoor conditions are considered for satellite operations, since performance requirements are not expected to be
met with the available link budget for indoor communications.
Since HAPS are closer to the Earth, resulting in less path loss than in satellite access networks, additional indoor
conditions are also considered for HAPS.
Several user environments will be considered, depending on the frequency band: open, rural, suburban, urban and dense
urban. In open environments (such as fixed terminals or terminals mounted on boats/aircrafts), an AWGN channel is
assumed.
6.1.3 Modeling objectives

The requirements for channel modelling are as follows:
- Support a frequency range from 0.5 GHz up to 100 GHz. Two frequency bands are targeted in particular: below
6 GHz and Ka bands. For Ka band communications, the uplink frequency is around 30 GHz while the downlink
frequency is around 20 GHz.
- Accommodate UE mobility. For satellite channel models, mobility speed up to 1000 km/h is supported; this
corresponds to aircrafts that can be served by satellite access. For HAPS channel models, mobility speed up to
around 500 km/h is supported, corresponding to high speed trains.
6.2 Differences between satellite/HAPS and cellular channel

modelling
Figures 6.2-1 and 6.2-2 compare the macro-cellular and satellite access links in NLOS and LOS cases, respectively. The
terrestrial propagation is quite similar: multipath propagation is caused by objects near the user. However, the angular
spread from satellite is almost zero (due to its distant location) while it can still be several degrees from the base station.
3GPP
Release 15 44 3GPP TR 38.811 V15.1.0 (2019-06)
Figure 6.2-1: Macro-cellular vs. satellite channel, NLOS
Figure 6.2-2: Macro-cellular vs. satellite channel, LOS
New satellite link model Terrestrial model based on TR38.901
Terrestrial
Dynamic Dynamic Convolution
Path loss &
delay attenuation
shadowing
Satellite Satellite Scintillation, Scenario- LSPs, SSPs

orbit Doppler Rain & cloud Specific
shift macro-cell
parameters
Figure 6.2-3: Combined satellite and terrestrial channels (conceptual drawing).
3GPP
Release 15 45 3GPP TR 38.811 V15.1.0 (2019-06)
6.3 Coordinate system

A three-dimensional global coordinate system is considered, described as "Earth Centred Earth fixed". The Earth is
approximated as a true sphere with radius of 6371 km. The coordinates' origin O lies in the center of the earth, x-y plane
locates in the equator plane with x-axis pointing to 0 degree longitude, y axis pointing to 90 degree longitude, and z-
axis pointing to geographical north pole from the origin O.
A UE or a satellite position is described by a set of three parameters ( x, y, z ) , with x ² y ² z ²  6371km for all
UEs and x ² y ² z ²  6371km for all satellites.
The proposed coordinate system is illustrated in Figure 6.3-1 for a non-GEO satellite constellation.
Figure 6.3-1: Illustration of the coordinate system
6.4 Antenna modelling

6.4.1 HAPS/Satellite antenna
Satellite antenna pattern
The following normalized antenna gain pattern, corresponding to a typical reflector antenna with a circular aperture, is
considered
1 for  0
2
𝐽1 𝑘𝑎 sin 𝜃
4
𝑘𝑎 sin 𝜃 for 0 < 𝜃 ≤ 90°
where J1(x) is the Bessel function of the first kind and first order with argument x, a is the radius of the antenna's
circular aperture, k = 2f/c is the wave number, f is the frequency of operation, c is the speed of light in a vacuum and 
is the angle measured from the bore sight of the antenna's main beam. Note that ka equals to the number of wavelengths
on the circumference of the aperture and is independent of the operating frequency.
The normalized gain pattern for a = 10 c/f (aperture radius of 10 wavelengths) is shown in Figure 6.4.1-1.
3GPP
Release 15 46 3GPP TR 38.811 V15.1.0 (2019-06)
Figure 6.4.1-1: Satellite antenna gain pattern for aperture radius 10 wavelengths, a=10 c/f
HAPS antenna pattern

Two different antenna patterns are considered:
- The above antenna pattern defined for satellite scenarios, based on the Bessel function.
- The 3GPP antenna pattern defined for the base station in Section 7.3 of [12], corresponding to a uniform
rectangular panel array with dual linear polarization.
6.4.2 UE antenna pattern

The following reference UE antenna patterns are adopted for fast fading:
- Quasi Isotropic - Linear polarisation (Quasi isotropic refers to dipole antenna which is omni-directional in one
plane)
- Co-phased array - Dual Linear polarisation (one for below 6 GHz band and one for above 6 GHz band as
described in [48])
- "VSAT type - circular polarization: fixed or tracking" UE antenna pattern (only in deployment scenarios
featuring flat fading conditions)
6.5 Methodology to define channel models

6.5.1 System-level methodology
Only drop-based simulations are considered, similarly to [12].
The baseline model to generate channel coefficients is the one described in [12] for terrestrial links, and depicted in
Figure 6.5.1-1.
3GPP
Release 15 47 3GPP TR 38.811 V15.1.0 (2019-06)
General parameters:
Set scenario, Assign propagation Generate correlated

network layout and condition (NLOS/ large scale
Calculate pathloss
antenna parameters LOS) parameters (DS,
AS, SF, K)
Small scale parameters:
Perform random Generate arrival & Generate cluster

Generate XPRs Generate delays
coupling of rays departure angles powers
Coefficient generation:
Draw random initial Generate channel Apply pathloss and

phases coefficient shadowing
Figure 6.5.1-1: Channel coefficient generation procedure issued from [12]
An alternative and simplified model can be applied if all UE meet the flat fading criteria. In this case, the channel
coefficients reduce to a single tap, since the channel is not frequency selective. This simplified model, derived from [10]
is depicted in Figure 6.5.1-2.
General parameters:
Set scenario, Assign propagation Generate large

network layout and condition (NLOS/ Calculate pathloss scale parameters
antenna parameters LOS) (SF only)
Small scale parameters:
Generate LMS
parameters
Coefficient generation:
Generate channel Apply pathloss and

coefficient shadowing
Figure 6.5.1-2: Simplified channel coefficient generation issued from [10]
6.5.2 Link-level methodology

Similarly to [12], reference CDL and TDL are considered for link-level simulations. For given environment and
elevation angle, they are obtained from a single instance of the baseline model used for system level simulations (i.e.
the one derived from [12]).
3GPP
Release 15 48 3GPP TR 38.811 V15.1.0 (2019-06)
For flat fading conditions (including AWGN) no channel model is needed for link-level simulations.
6.6 Large scale model

6.6.1 LOS probability
Line-Of-Sight (LOS) probability depends on UE environment and elevation angle, and is obtained from Table 6.6.1-1.
Reference elevation angles are considered from 10° to 90° with a 10° step. For an UE-to-satellite or UE-to-HAPS link,
the LOS probability is taken from the nearest reference elevation angle.
Table 6.6.1-1 LOS probability
Dense urban Urban scenario Suburban and Rural

Elevation
scenario scenarios
10° 28.2% 24.6% 78.2%
20° 33.1% 38.6% 86.9%
30° 39.8% 49.3% 91.9%
40° 46.8% 61.3% 92.9%
50° 53.7% 72.6% 93.5%
60° 61.2% 80.5% 94.0%
70° 73.8% 91.9% 94.9%
80° 82.0% 96.8% 95.2%
90° 98.1% 99.2% 99.8%
6.6.2 Path loss and Shadow fading

The signal path between a satellite or HAPS transmitter and an NTN terminal undergoes several stages of propagation
and attenuation. The path loss (PL) is composed of components as follows:
PL  PLb  PLg  PLs  PLe , (6.6-1)
where PL is the total path loss in dB,
PLb is the basic path loss in dB,

PL g is the attenuation due to atmospheric gasses in dB,
PLs is the attenuation due to either ionospheric or tropospheric scintillation in dB,
PLe is building entry loss in dB.
This section specifies the basic path loss model ( PLb ) which accounts for the signal's free space propagation, clutter
loss, and shadow fading. Attenuations due to building entry loss, atmospheric gasses and scintillation are described in
Sections 6.6.3, 6.6.4 and 6.6.6, respectively.
The free space path loss (FSPL) in dB for a separation distance d in meter and frequency in GHz is given by
FSPL(d , f c )  32.45 20 log10 ( f c )  20 log10 (d ) (6.6-2)
For a ground terminal, the distance d (a.k.a. slant range), as shown in Figure 6.6.2-1, can be determined by the
satellite/HAPS altitude h0 and elevation angle α by
d  RE2 sin 2   h02  2h0 RE  RE sin 

, (6.6-3)
where RE denotes Earth radius.
3GPP
Release 15 49 3GPP TR 38.811 V15.1.0 (2019-06)
d h0
horizon
nadir
α
terminal
Earth surface RE
Figure 6.6.2-1: Slant range d between a satellite and a ground terminal
Clutter loss (CL) models the attenuation of signal power caused by surrounding buildings and objects on the ground. It
depends on the elevation angle α, the carrier frequency fc, and the environment. Shadow fading (SF) is modeled by a
log-normal distribution, which when expressed in decibel unit, is a zero-mean normal distribution with a standard
deviation  SF
2
, i.e., N (0, SF ) .
2
The basic path loss in dB unit is modeled as
PLb  FSPL(d , f c )  SF  CL( , f c ) , (6.6-4)
where FSPL (d , f c ) is the free space path loss, CL( , f c ) is clutter loss, and SF is shadow fading loss represented
by a random number generated by the normal distribution, i.e., SF ~ N (0, SF
2
) . When the UE is in LOS condition,
clutter loss is negligible and should be set to 0 dB in the basic path loss model.
The values of  SF
2
and CL are given in tables 6.6.2-1 to 6.6.2-3 at reference elevation angles for different scenarios.
The UE in a particular scenario should take the values corresponding to the reference angle nearest to its elevation angle
α.
Table 6.6.2-1: Shadow fading and clutter loss for dense urban scenario
S-band Ka-band
LOS NLOS LOS NLOS
Elevation  SF (dB)  SF (dB) CL (dB)  SF (dB)  SF (dB) CL (dB)

10° 3.5 15.5 34.3 2.9 17.1 44.3
20° 3.4 13.9 30.9 2.4 17.1 39.9
30° 2.9 12.4 29.0 2.7 15.6 37.5
40° 3.0 11.7 27.7 2.4 14.6 35.8
50° 3.1 10.6 26.8 2.4 14.2 34.6
60° 2.7 10.5 26.2 2.7 12.6 33.8
70° 2.5 10.1 25.8 2.6 12.1 33.3
80° 2.3 9.2 25.5 2.8 12.3 33.0
90° 1.2 9.2 25.5 0.6 12.3 32.9
3GPP
Release 15 50 3GPP TR 38.811 V15.1.0 (2019-06)
Table 6.6.2-2: Shadow fading and clutter loss for urban scenario
S-band Ka-band
LOS NLOS LOS NLOS

10° 4 6 34.3 4 6 44.3
20° 4 6 30.9 4 6 39.9
30° 4 6 29.0 4 6 37.5
40° 4 6 27.7 4 6 35.8
50° 4 6 26.8 4 6 34.6
60° 4 6 26.2 4 6 33.8
70° 4 6 25.8 4 6 33.3
80° 4 6 25.5 4 6 33.0
90° 4 6 25.5 4 6 32.9
Table 6.6.2-3: Shadow fading and clutter loss for suburban and rural scenarios
S-band Ka-band
LOS NLOS LOS NLOS

10° 1.79 8.93 19.52 1.9 10.7 29.5
20° 1.14 9.08 18.17 1.6 10.0 24.6
30° 1.14 8.78 18.42 1.9 11.2 21.9
40° 0.92 10.25 18.28 2.3 11.6 20.0
50° 1.42 10.56 18.63 2.7 11.8 18.7
60° 1.56 10.74 17.68 3.1 10.8 17.8
70° 0.85 10.17 16.50 3.0 10.8 17.2
80° 0.72 11.52 16.30 3.6 10.8 16.9
90° 0.72 11.52 16.30 0.4 10.8 16.8
6.6.3 O2I penetration loss

For an indoor Earth-based station, account must be taken of the additional loss between the station and the adjacent
outdoor path. The additional loss varies greatly with the location and construction details of buildings, and a statistical
evaluation is required. Recommendation ITU-R P.2109 gives a suitable building entry/exit-loss model for this purpose.
Experimental results, such as those collated in Report ITU-R P.2346, shows that, when characterised in terms of entry
loss, buildings fall into two distinct populations: where modern, thermally-efficient building methods are used
(metallised glass, foil-backed panels) building entry loss is generally significantly higher than for 'traditional' buildings
without such materials. The model therefore gives predictions for these two cases.
This classification, of 'thermally efficient' and 'traditional', refers purely to the thermal efficiency of construction
materials. No assumption should be made on the year of construction, type (single or multi-floors), heritage or building
method.
For building entry loss, it is important to consider the thermal efficiency of the complete building (or the overall thermal
efficiency). A highly thermally efficient main structure with poorly insulated windows (e.g. single glazed with thin
glass) can make the building thermally inefficient and vice versa.
Thermal transmittance, commonly referred as U-value, provides a quantifiable description of thermal efficiency. Low
U-values represent high thermal efficiency. Typically, the presence of metallised glass windows, insulated cavity walls,
thick reinforced concrete and metal foil back cladding is a good indication of a thermally efficient building.
NOTE: For example, U-values of < 0.3 and < 0.9 are representative of thermally efficient main structure and
metallised glass, respectively.
3GPP
Release 15 51 3GPP TR 38.811 V15.1.0 (2019-06)
Building entry loss will vary depending on building type, location within the building and movement in the building.
The building entry loss distribution is given by a combination of two lognormal distributions. The building entry loss
not exceeded for the probability, P, is given by:
LBEL ( P)  10 log(100.1 A( P )  100.1B ( P )  100.1C )dB (6.6-5)
with:
A( P )  F 1 ( P ) 1  1
B ( P )  F 1 ( P ) 2   2
C  3.0
1  Lh  Le
 2  w  x log( f )
 1  u  v log( f )
 2  y  z log( f )
where:
Lh is the median loss for horizontal paths, given by:

Lh  r  s log( f )  t (log( f )) 2
(6.6-6)
Le is the correction for elevation angle of the path at the building façade:
Le 0.212
(6.6-7)
and:
f= frequency (GHz)
θ= elevation angle of the path at the building façade (degrees)
P= probability that loss is not exceeded (0.0 < P < 1.0)
-1
F (P) = inverse cumulative normal distribution as a function of probability.
and the coefficients are as given in Table 6.6.3-1:
Table 6.6.3-1: Model coefficients
Building r s t u v w x y z
type
Related to: Median BEL (μ1) σ1 μ2 σ2
Traditional 12.64 3.72 0.96 9.6 2.0 9.1 −3.0 4.5 −2.0
Thermally- 28.19 −3.00 8.48 13.5 3.8 27.8 −2.9 9.4 −2.1
efficient
6.6.4 Atmospheric absorption

Attenuation by atmospheric gases which is entirely caused by absorption depends mainly on frequency, elevation angle,
altitude above sea level and water vapour density (absolute humidity). At frequencies below 10 GHz, it may normally
be neglected. However, for elevation angles below 10 degrees it is recommended that the calculation is performed for
any frequency above 1 GHz. Annex 1 of Recommendation ITU-R P.676 gives a complete method for calculating
gaseous attenuation, while Annex 2 of the same Recommendation gives an approximate method for frequencies up to
350 GHz.
For system level simulations, the baseline method is as follows:
3GPP
Release 15 52 3GPP TR 38.811 V15.1.0 (2019-06)
- The method of Annex 2 in ITU-R P.676 is considered (except for UE altitude higher than 10km and for
frequencies within 0.5 GHz of the centres of resonance lines at any altitude).
- For all UEs, a geometric height of 0km is considered, corresponding to the sea level.
- For all UEs, dry air pressure, water-vapour density, water vapour partial pressure and temperature correspond to
the mean annual global reference atmosphere given in Recommendation ITU-R P835.
For all UEs, this corresponds to the following values:
T = 288.15 K
p = 1013.25 hPa
  7.5 g / m3
T
e=  9.98hPa
216.5
where T denotes the temperature, p the dry air pressure, ρ the water-vapour density and e the water vapour partial
pressure.
Figure 6 of ITU-R P.676 shows the corresponding zenith attenuation Azenith ( f ) for frequencies between 1 and
350 GHz. For an elevation angle α, the corresponding attenuation PL A ( , f ) is given by:
Azenith ( f )
PL A ( , f ) 
sin( ) (6.6-8)
6.6.5 Rain and cloud attenuation

Rain and cloud attenuation is considered as negligible for frequencies below 6 GHz. Section 2.2 of [11] describes a
method to estimate the long-term statistics of attenuation due to rain, which are location specific.
For system-level simulations, the baseline is to consider clear sky conditions only.
Alternatively, the following procedure shall be followed to define rain and cloud attenuation (adaptation of [11] for
drop-based simulations):
- For each UE, determine its CDF of rain and cloud attenuation (location specific) using Section 2.2 of [11].
- For each drop, draw the attenuation due to rain and cloud attenuation for each UE from its corresponding CDF
NOTE: The spatial correlation of rain and cloud attenuation is not taken into account in this procedure.
6.6.6 Scintillation
6.6.6.1 Ionospheric scintillation

Scintillation corresponds to rapid fluctuations of the received signal amplitude and phase. Ionosphere propagation shall
only be considered for frequencies below 6 GHz.
These phenomena are among the most severe disruptions along a trans-ionospheric propagation path for signals below 3
GHz, and may be observed occasionally up to 10 GHz [13]. Scintillations depend on location, time-of-day (as observed
in Figure 6.6.6.1-1), season, solar and geomagnetic activity. During nominal conditions, strong levels of scintillation are
rarely observed in mid-latitudes, but they may be encountered daily during post-sunset hours in low latitude regions. At
high (auroral and polar) latitudes, moderate to strong levels of scintillations have been observed.
3GPP
Release 15 53 3GPP TR 38.811 V15.1.0 (2019-06)
Figure 6.6.6.1-1: Occurrence of different scintillation events around the solar maximum of 2014 at low
(top) and high (bottom) latitudes
6.6.6.1.1 Ionospheric scintillation indices

The most commonly used parameter to characterize intensity fluctuations (amplitude scintillations) is the amplitude
scintillation index S4 [4], defined by equation:
I2  I
2
S4  ( 2
)1/ 2
I
(6.6-9)
where I is intensity (proportional to the square of the signal amplitude) and denotes averaging, usually over a
period of 60 seconds. Likewise, phase scintillations are characterized by the standard deviation of the phase variations,
the phase scintillation index σφ:
  2  
2
(6.6-10)
where φ is carrier phase in radians and denotes averaging, usually over a period of 60 seconds.
For convenience, scintillation strength can be classified into three regimes:
Table 6.6.6.1.1-1: Definition of scintillation regime based on S4 values
Scintillation Regime Amplitude Scintillation Phase Scintillation (rad)

Weak S4 < 0.3 σφ < 0.25-0.3
Moderate 0.3 ≤ S4 ≤ 0.6 0.25-0.3 ≤ σφ ≤ 0.5-0.7
Strong S4 > 0.6 σφ > 0.5-0.7
6.6.6.1.2 Ionospheric scintillation location dependence

As previously mentioned, scintillation effects differ at low and high-latitudes, and they are not observed at mid-latitudes
except for strong geomagnetic storms. At high-latitudes (e.g., above 60°), the effect mainly occurs from the high-
latitude edge of the Van Allen outer belt into polar region. On the other hand, Equatorial scintillations occur around
±20° of latitude of the magnetic equator and they are due to large (~100 km) depleted ionization volumes driven
through the F region, leaving a plume of small-scale (tens of cm to m) irregularities surrounding the depletion, which
can extend well through F-layer peak. They are produced by convective plasma processes. Irregularities with this range
of scales are not independent from larger-scale plasma structures to those of smaller-scale irregularities.
3GPP
Release 15 54 3GPP TR 38.811 V15.1.0 (2019-06)
The cross-correlation between S4 and σφ is markedly different between high geomagnetic latitudes and low latitudes.
Amplitude scintillations dominate at low-latitudes, and phase scintillations dominate at high latitudes, however, they are
not exclusively and both effects can be expected in the two regions.
6.6.6.1.3 Frequency scaling

For frequency scaling, typically the following relation on amplitude scintillation S4 index is used:
f 2 n
s4, f 2  s4, f1 ( )
f1 (6.6-11)
with n=1.5 recommended for L-band frequencies. In [14], values of n derived from satellite measurement data between
several pairs of frequencies from 30 MHz up to 6 GHz are presented, ranging from 1 to 2. This relationship is valid
particularly for weak scattering assumptions (higher elevations and low to moderate S4 values below 0.6). For high S4
values (S4=1), the relation saturates with n equal to 0.
For phase scintillations, an equivalent relation is used:
f 2 n
s , f 2  s , f1 ( )
f1 (6.6-12)
with n=1 recommended for L-band frequencies and also reaching saturation for high σφ values.
As an illustrative example, the frequency scaling between GPS L1, L2 and L5 bands (1.57542, 1.22760 and 1.17645
GHz, respectively) is presented in Figure 6.6.6.1.3-1, where scintillation events in two bands are compared against each
other and against the theoretical values described by 6.6-11 and 6.6-12.
Figure 6.6.6.1.3-1: Frequency correlation of scintillation events observed in GPS L1, L2 and L5 bands
6.6.6.1.4 Model for Ionospheric scintillation loss

The proposed method for the ionospheric scintillation loss is based on the so-called Gigahertz scintillation model ([13],
Section 4.8), and it is valid only for the regions located approximately 20° north and south of the magnetic equator. At
high-latitudes (e.g., above 60°), this model is not applicable, whereas for other latitude locations the ionospheric
scintillation can be neglected.
To evaluate the scintillation effects that can be expected in a given situation the following steps may be used:
3GPP
Release 15 55 3GPP TR 38.811 V15.1.0 (2019-06)
Step 1: Figure 6.6.6.1.4-1 provides scintillation occurrence statistics on equatorial ionospheric paths: peak-to-
peak amplitude fluctuations, Pfluc, (dB), for 4 GHz reception from satellites in the East at elevation angles of
about 20° (P solid curves) and in the West at about 30° elevation (I dotted curves). The data are given for
different times of year and sunspot number.
Step 2: Since Figure 6.6.6.1.4-1 relates to 4 GHz, values for other frequencies are found by multiplying these
values by ( f /4)–1.5 where f is the frequency of interest (GHz).
Step 3: Since one element of link budget calculations is related to signal loss due to ionospheric scintillation, AIS,
the following relationship is recommended:
AIS  Pfluc / 2
(6.6-13)
NOTE: The most widely used parameter in describing amplitude scintillations phenomena is the amplitude index
S4. It is adopted to define three main regime conditions (see Table 6.6.6.1.1-1), and it is related to Pfluc
using the empirical approximation:
Pfluc = 27.5 S41.26 (6.6-14)
where the empirical conversion table is presented in Table 1 of [13].
3GPP
Release 15 56 3GPP TR 38.811 V15.1.0 (2019-06)
10
5
P6
P5
2 P4
Percentage of time the abscissa is exceeded
1 P3
0.5
P2
0.2
I6
0.1
I5
P1
I4
0.05 I3
I2
I1
0.02
0.01
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14
Peak-to-peak fluctuation (dB)
Curve Period SSN range

I1, P1 March 75-76 10-15
I2, P2 June 76-77 12-26
I3, P3 March 77-78 20-70
I4, P4 October 77-78 44-110
I5, P5 November 78-79 110-160
I6, P6 June 79-80 153-165
Figure 6.6.6.1.4-1: Annual statistics of peak-to-peak fluctuations observed at Hong Kong earth station
(Curves I1, P1, I3-I6, P3-P6) and Taipei earth station (Curves P2 and I2). Extracted from [13].
For system-level simulations below 6 GHz, PLS is equal to AIS from Equation (6.6-13) for latitudes of maximum ±20°.
For latitudes between ±20° and ±60° of latitude, PLS  0 . Finally, for latitude above ±60°, S4 equal to 0.7 can be
assumed (i.e., strong scintillation regime), therefore Pfluc equal to 17 from Equation (6.6-14) and PLS derived from
Equation (6.6-13).
6.6.6.2 Tropospheric scintillation

Scintillation corresponds to rapid fluctuations of the received signal amplitude and phase. Tropospheric propagation
shall only be considered for frequencies above 6 GHz.
Tropospheric scintillation is a phenomenon that causes rapid amplitude and phase fluctuations of signals from satellite
communication systems. Unlike, ionospheric scintillation, the effect of tropospheric scintillation increases with the
carrier frequency of the signal, being especially significant above 10 GHz. In this case, the signal fluctuations are
caused by sudden changes in the refractive index due to the variation of temperature, water vapor content, and
barometric pressure.
Besides increasing with the carrier frequency, the effects of scintillation also increase with low elevation angles, due to
the longer path of the signal, and wide beam width receiving antennas.
3GPP
Release 15 57 3GPP TR 38.811 V15.1.0 (2019-06)
6.6.6.2.1 Model for Tropospheric scintillation loss

The ITU-R recommendation algorithm for fading prediction [11] permits an accurate prediction of the amplitude. This
method consists in three parts:
- Prediction of the amplitude scintillation fading at free-space elevation angles ≥ 5° (Section 2.4.1 in [13]).
- Prediction of the amplitude scintillation fading for fades ≥ 25 dB (section 2.4.2 in [13]).
- Prediction of the amplitude scintillation in the transition region between the above two distributions (section
2.4.3 in [13]).
An illustrative example of typical power attenuation levels as a function of the elevation angle is depicted in Figure
6.6.6.2.1-1. The user location is Toulouse (France), the carrier frequency is set to 20 GHz, and circular polarization is
assumed. Even though tropospheric scintillation is latitude dependent, it is suggested to take this plot as a reference for
satellite link margin computation.
Figure 6.6.6.2.1-1: Complementary cumulative probability function of the tropospheric scintillation

attenuation at 20 GHz in Toulouse (France).
As baseline for system-level simulations above 6 GHz, the fading due to scintillation PLS is equal to the tropospheric
attenuation level at 99% of the time derived from Figure 6.6.6.2.1-1 and summarized in Table 6.6.6.2.1-1. Alternatively,
it can be drawn for each UE based on its corresponding CDF from Figure 6.6.6.2.1-1, assuming no correlation between
different UE.
3GPP
Release 15 58 3GPP TR 38.811 V15.1.0 (2019-06)
Table 6.6.6.2.1-1: Tropospheric attenuation in dB with 99% probability at 20 GHz in Toulouse
Elevation angle [deg] Tropospheric attenuation,

P{AIS > x} < 0.01
10 1.08 dB
20 0.48 dB
30 0.30 dB
40 0.22 dB
50 0.17 dB
60 0.13 dB
70 0.12 dB
80 0.12 dB
90 0.12 dB
6.7 Fast fading model

The generic fast fading model in 3GPP is based on TR 38.901 [12]. For narrowband SISO simulations, an optional
simplified channel model can be used as long as the flat fading assumption is valid.
Most literature on satellite channel models rely on flat fading models [15], [16], [17], i.e. non frequency selective
channel models, the ITU two-state model described in [1] being the most up-to-date model.
1
The formula BC  is defined to calculate the coherence bandwidth from the 95th percentile rms delay spread.
10 rms
Coherence bandwidth depends on environment, antenna pattern and elevation. The channel is assumed to be flat if the
UE bandwidth is lower than the coherence bandwidth of the channel. In satellite channels, where small fade margins are
usually considered,  rms can be assessed only for cases where the fade event is below a given threshold (otherwise,
system outage will occur regardless of channel dispersion).
In [47], values of  rms were estimated considering an omnidirectional UE antenna for suburban and urban
environments, several elevation angles and fade margins. Based on these results, the ITU two-state model can at least
(NOTE 1) be used as a simplified alternative to the TR 38.901 [12] methodology for satellite links if all of the following
conditions are met:
- S-band scenario
- Minimum elevation angle is 20° or above
- Quasi-LOS conditions (i.e. fading margin is approx. 5dB maximum)
- Channel bandwidth is 5 MHz or below
- Environment is rural, suburban or urban
NOTE 1: Flat fading conditions are more easily achievable when using highly directive UE antennas located in less
scattering environment, like on a rooftop or on an open field. In every case, the flat fading criterion
described above shall be fulfilled.
6.7.1 Flat fading

We consider here the ITU two-state model. This model includes already the clutter loss and the shadow fading, so that
the basis path loss is calculated as follows:
PLb  FSPL(d , f c ) (6.7-1)
In the two-state model, the signal level is statistically described with a good state (corresponding to LOS and slightly
shadowed conditions) and a bad state (corresponding to severe shadowed conditions). The state duration is described by
a semi-Markov model. Within each state fading is described by a Loo distribution where the received signal is the sum
of the direct path signal and the diffuse multipath.
3GPP
Release 15 59 3GPP TR 38.811 V15.1.0 (2019-06)
The Loo distribution is therefore defined with the following parameters:
- Mean of the direct signal
- Standard deviation of the direct signal
- Mean of the multipath
The following procedure shall be followed for system-level evaluations:
Step 1: Set general parameters related to environment and satellite link as follows:
- Set the center frequency from 1.5 GHz to 20 GHz;
- Choose one of the following LMS scenarios available (in S band: urban, suburban, rural wooded, residential – in
Ka band: suburban, rural wooded);
- Set the link elevation assuming a rounded value towards the closest available elevation for the
frequency/environment chosen (20°, 30°, 34°, 45°, 60°, 70°);
- Give UE position, array orientation, speed and direction of motion in the global coordinate system.
Step 2: Determine the (µ,)G,B , (  M A ,  M A ) G , B , (g1,g2)G,B, (h1,h2)G,B, (durmin)G,B ,(f1,f2), pB,min and pB,max from the input
parameters table provided in Annex 2 of [1] and summarized in Table 6.7.1-1.
Table 6.7.1-1: Model parameters of the 2-state model
Parameter Description
(µ,)G,B Mean and standard deviation of the log-normal law assumed for events duration (m)
durminG,B Minimum possible events duration (m)
( M M Parameters of the MA G,B distribution (MA being the average value of the direct path
GB, GB)
A A amplitude A over one event) (dB)
MP = h1G,BMA+h2G,B Multipath power, MPG,B (one 1st order polynomial for each state), (dB)
 AG,B = g1G,BMA+g2G,B Standard deviation of A,  AG,B (one 1st order polynomial for each state)
LcorrG,B NOTE 2 Direct path amplitude correlation distance (m)
f1ΔMA+f2 Transition length, Ltrans (one single 1st order polynomial), (m)
[pB,min , pB,max] Probability range to consider for the MA B distribution
NOTE 1: G stands for the GOOD state and B stands for the BAD state.
NOTE 2: Only for generative modelling.
Assign propagation condition (GOOD/BAD) states in the original procedure from [15].
NOTE: Unlike the frequency selective case, the LOS probability defined in clause 6.6.1 is not used when
considering the flat fading model
The propagation conditions for different Earth-space links are uncorrelated.
The GOOD and BAD state probability is calculated as follows :
log durmin, B ,G  (  G ,B   G2 ,B   G2 ,B )
1  erf ( )
 G2 ,B  2
dur  exp( G ,B  )
G ,B
2 log durmin B ,G  G ,B
1  erf ( )
 2 (6.7-2)
p N ( M A,min ;  M A ,B ,  M A ,B )  pN ( M A,max ;  M A,B ,  M A, B )

dur  f1  (  M A ,G   M A ,B   M2 A ,B  )  f2
T
FN ( M A,max ;  M A ,B ,  M A ,B )  FN ( M A,min ;  M A ,B ,  M A ,B )
(6.7-3)
3GPP
Release 15 60 3GPP TR 38.811 V15.1.0 (2019-06)
M A,min/ max, B   M A, B  2 M A, B erf 1 (2 pB ,min/ max  1)

(6.7-4)
dur  dur
PG  G T
dur G
 dur B
 2 dur T (6.7-5)
- Where subscripts G, B and T stand respectively for good, bad and transition states,  dur  the mean duration
of the considered state in meters, durmin the minimal state duration in meters, µ and σ respectively the mean and
standard deviation of the assumed log-normal law in m.
- pN(x; ,) and FN(x; ,) are respectively the probability density function and the cumulative distribution
function of a normal distribution with mean  and standard deviation  as defined in Recommendation ITU-R
P.1057
- Where M A
GB, M A
GB are the parameters of the average value of the direct path amplitude A over one event,
[pB,min , pB,max] the probability range to consider for the M A,B distribution.
Step 3: Draw MAi, the mean power of the direct signal, as a normally distributed parameter function of (  M A ,  M A ) G , B
expressed in dB.
M Ai ~ N( MAi ,  M Ai )
Compute ΣAi and MPi, respectively the standard deviation of the direct signal and the mean multipath power both
expressed in dB where suscript i designate the good or bad state, as follow:
ΣAi = g1iMAi + g2i (6.7-6)
MPi = h1iMAi + h2i (6.7-7)
Step 4: Draw the power of the direct signal PD following the normally distribution( M Ai ,  Ai ) and derive the K factor
as PD - MPi (all expressed in dB).
For each UE, the channel is therefore characterized by a single Rice distributed tap.
Note that the above procedure is only valid for simulation durations up to a few TTIs. It is further assumed that no state
change occurs during the simulation duration.
For longer simulations, the procedure described in clause 6.2 from [10] must be applied, as the K factor must not be
considered as constant.
Note that individual fading values at a given time may directly be obtained after step 3 of the above procedure, based on
the Loo distribution:

8.686 x 1 (20 log10 (a)  M Ai )? x ² a ² xa
p Loo ( x) 
 Ai  i2 2  a exp(
0
2 Ai
2

2 i2
) I 0 ( 2 )da
i
(6.7-8)
With 2  i being the multipath mean received power expressed in dB, i.e. MPi = 10log (2  i )
2 2
Similarly to [10], a Jake’s Doppler spectrum is considered for UE mobility.
3GPP
Release 15 61 3GPP TR 38.811 V15.1.0 (2019-06)
Additional Doppler shift due to satellite motion should be taken into account according to the following formula:
 R 
f d , shift  (vsat / c)   cos  mod el   f c ,
 Rh 
Where vsat denotes the satellite speed, c denotes the speed of light, R denotes the earth radius, h denotes the satellite
altitude,  mod el denotes the satellite elevation angle, and f c denotes the carrier frequency.
The satellite speed, satellite elevation angle and UE speed should be considered to be constant during the simulation
duration, if limited to few TTIs.
6.7.2 Frequency selective fading

In the fast fading model, the process in 7.5 of TR 38.901 [12] is used. This section is not a stand-alone description of the
fast fading model, but it describes the differences between the channel models used for terrestrial and satellite/HAPS
communications. As can be seen from Figure 6.7.2-1, there is not much difference in local scattering between the HAPS
and satellite cases. Therefore, the same fast fading parameters can be used for the both cases, including different
satellite orbits as well. The critical parameter is the elevation angle of the LOS path of the satellite/HAPS vs. ground
horizon.
8 – 200 km 600 – 36 000 km
Figure 6.7.2-1: HAPS to UE vs. satellite to UE propagation
Instead of the parameterization tables in TR 38.901 [12] (Table 7.5-6 Part-1 and Part-2) the following tables shall be
used.
NOTE 1: Some channel models may lead to pessimistic results of the performance of satellite/HAPS to UE link
especially in the higher elevations due to the high number of clusters and low K factor.
NOTE 2: In some cases, the correlation distances are shorter in real world conditions.
Angular scaling factors in cluster generation need to be added to the NTN scenarios that have lower number of clusters
than the scenarios described in TR 38.901 [12] (Table 6.7.2-1aa below corresponds to Table 7.5-2 in TR 38.901 [12]
and Table 6.7.2-1ab below corresponds to Table 7.5-3 in TR 38.901 [12]).
3GPP
Release 15 62 3GPP TR 38.811 V15.1.0 (2019-06)
Table 6.7.2-1aa: Scaling factors for AOA, AOD generation

# clusters 2 3 4 5 8 10 11 12 14 15 16 19 20
NLOS
C 0.501 0.680 0.779 0.860 1.018 1.090 1.123 1.146 1.190 1.211 1.226 1.273 1.289
Table 6.7.2-1ab: Scaling factors for ZOA, ZOD generation

# clusters 2 3 4 8 10 11 12 15 19 20
CNLOS 0.430 0.594 0.697 0.889 0.957 1.031 1.104 1.1088 1.184 1.178
Table 6.7.2-1a: Channel model parameters for Dense Urban Scenario (LOS) in S band
Dense Urban LOS

Scenarios
10° 20° 30° 40° 50° 60° 70° 80° 90°
Delay spread (DS) lgDS -7.12 -7.28 -7.45 -7.73 -7.91 -8.14 -8.23 -8.28 -8.36
lgDS=log10(DS/1s) lgDS 0.80 0.67 0.68 0.66 0.62 0.51 0.45 0.31 0.08
AOD spread (ASD) lgASD -3.06 -2.68 -2.51 -2.40 -2.31 -2.20 -2.00 -1.64 -0.63
lgASD=log10(ASD/1) lgASD 0.48 0.36 0.38 0.32 0.33 0.39 0.40 0.32 0.53
AOA spread (ASA) lgASA 0.94 0.87 0.92 0.79 0.72 0.60 0.55 0.71 0.81
lgASA=log10(ASA/1) lgASA 0.70 0.66 0.68 0.64 0.63 0.54 0.52 0.53 0.62
ZOA spread (ZSA) lgZSA 0.82 0.50 0.82 1.23 1.43 1.56 1.66 1.73 1.79
lgZSA=log10(ZSA/1) lgZSA 0.03 0.09 0.05 0.03 0.06 0.05 0.05 0.02 0.01
ZOD spread (ZSD) lgZSD -2.52 -2.29 -2.19 -2.24 -2.30 -2.48 -2.64 -2.68 -2.61
lgZSA=log10(ZSD/1) lgZSD 0.50 0.53 0.58 0.51 0.46 0.35 0.31 0.39 0.28
Shadow fading (SF) [dB] SF 3.5 3.4 2.9 3.0 3.1 2.7 2.5 2.3 1.2
K 4.4 9.0 9.3 7.9 7.4 7.0 6.9 6.5 6.8
K-factor (K) [dB]
K 3.3 6.6 6.1 4.0 3.0 2.6 2.2 2.1 1.9
ASD vs DS 0.4 0.4 0.4 0.4 0.4 0.4 0.4 0.4 0.4
ASA vs DS 0.8 0.8 0.8 0.8 0.8 0.8 0.8 0.8 0.8
ASA vs SF -0.5 -0.5 -0.5 -0.5 -0.5 -0.5 -0.5 -0.5 -0.5
ASD vs SF -0.5 -0.5 -0.5 -0.5 -0.5 -0.5 -0.5 -0.5 -0.5
DS vs SF -0.4 -0.4 -0.4 -0.4 -0.4 -0.4 -0.4 -0.4 -0.4
Cross-Correlations
ASD vs ASA 0 0 0 0 0 0 0 0 0
ASD vs  0 0 0 0 0 0 0 0 0
ASA vs  -0.2 -0.2 -0.2 -0.2 -0.2 -0.2 -0.2 -0.2 -0.2
DS vs  -0.4 -0.4 -0.4 -0.4 -0.4 -0.4 -0.4 -0.4 -0.4
SF vs  0 0 0 0 0 0 0 0 0
ZSD vs SF 0 0 0 0 0 0 0 0 0
ZSA vs SF -0.8 -0.8 -0.8 -0.8 -0.8 -0.8 -0.8 -0.8 -0.8
ZSD vs K 0 0 0 0 0 0 0 0 0
ZSA vs K 0 0 0 0 0 0 0 0 0
ZSD vs DS -0.2 -0.2 -0.2 -0.2 -0.2 -0.2 -0.2 -0.2 -0.2
Cross-Correlations
ZSA vs DS 0 0 0 0 0 0 0 0 0
ZSD vs ASD 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5
ZSA vs ASD 0 0 0 0 0 0 0 0 0
ZSD vs ASA -0.3 -0.3 -0.3 -0.3 -0.3 -0.3 -0.3 -0.3 -0.3
ZSA vs ASA 0.4 0.4 0.4 0.4 0.4 0.4 0.4 0.4 0.4
ZSD vs ZSA 0 0 0 0 0 0 0 0 0
Delay scaling parameter r 2.5 2.5 2.5 2.5 2.5 2.5 2.5 2.5 2.5
XPR 24.4 23.6 23.2 22.6 21.8 20.5 19.3 17.4 12.3
XPR [dB]
XPR 3.8 4.7 4.6 4.9 5.7 6.9 8.1 10.3 15.2
3GPP
Release 15 63 3GPP TR 38.811 V15.1.0 (2019-06)
Number of clusters N 3 3 3 3 3 3 3 3 3
Number of rays per cluster M 20 20 20 20 20 20 20 20 20
Cluster DS ( cDS ) in [ns] 3.9 3.9 3.9 3.9 3.9 3.9 3.9 3.9 3.9
Cluster ASD ( c ASD ) in [deg] 0 0 0 0 0 0 0 0 0
Cluster ASA ( c ASA ) in [deg] 11 11 11 11 11 11 11 11 11
Cluster ZSA ( cZSA ) in [deg] 7 7 7 7 7 7 7 7 7
Per cluster shadowing std  [dB] 3 3 3 3 3 3 3 3 3

DS 30 30 30 30 30 30 30 30 30
ASD 18 18 18 18 18 18 18 18 18
ASA 15 15 15 15 15 15 15 15 15
Correlation distance in the
horizontal plane [m]
SF 37 37 37 37 37 37 37 37 37
 12 12 12 12 12 12 12 12 12
ZSA 15 15 15 15 15 15 15 15 15
ZSD 15 15 15 15 15 15 15 15 15
fc is carrier frequency in GHz; d2D is BS-UT distance in km.
NOTE 1: DS = rms delay spread, ASD = rms azimuth spread of departure angles, ASA = rms azimuth spread of arrival angles, ZSD =
rms zenith spread of departure angles, ZSA = rms zenith spread of arrival angles, SF = shadow fading, and K = Ricean K-
factor.
NOTE 2: The sign of the shadow fading is defined so that positive SF means more received power at UT than predicted by the path
loss model.
NOTE 3: All large scale parameters are assumed to have no correlation between different floors.
NOTE 4: The following notation for mean (μlgX=mean{log10(X) }) and standard deviation (σlgX=std{log10(X) }) is used for logarithmized
parameters X.
NOTE 5: For all considered scenarios the AOD/AOA distributions are modelled by a wrapped Gaussian distribution, the ZOD/ZOA
distributions are modelled by a Laplacian distribution and the delay distribution is modelled by an exponential distribution.
NOTE 6: For UMa and frequencies below 6 GHz, use fc = 6 when determining the values of the frequency-dependent LSP values
NOTE 7: For UMi and frequencies below 2 GHz, use fc = 2 when determining the values of the frequency-dependent LSP values
NOTE 8: For satellite (e.g.GEO/LEO), the departure angle spreads are zeros, i.e. µ lgASD and µ lgZSD are –∞, and corresponding
standard deviations are zeros.
NOTE 9: The number of clusters is based on a limited data. The number may be different in the real field conditions.
Table 6.7.2-1b: Channel model parameters for Dense Urban Scenario (LOS) in Ka band
Dense Urban LOS

Scenarios
10° 20° 30° 40° 50° 60° 70° 80° 90°
lgDS=log10(DS/1s) lgDS 0.90 0.78 0.80 0.72 0.61 0.47 0.39 0.26 0.01
K-factor (K) [dB] K  6.1 13.7 12.9 10.3 9.2 8.4 8.0 7.4 7.6
K  2.6 6.8 6.0 3.3 2.2 1.9 1.5 1.6 1.3
ASD vs DS 0.4 0.4 0.4 0.4 0.4 0.4 0.4 0.4 0.4
ASA vs DS 0.8 0.8 0.8 0.8 0.8 0.8 0.8 0.8 0.8
ASA vs SF -0.5 -0.5 -0.5 -0.5 -0.5 -0.5 -0.5 -0.5 -0.5
ASD vs SF -0.5 -0.5 -0.5 -0.5 -0.5 -0.5 -0.5 -0.5 -0.5
Cross-Correlations DS vs SF -0.4 -0.4 -0.4 -0.4 -0.4 -0.4 -0.4 -0.4 -0.4
ASD vs ASA 0 0 0 0 0 0 0 0 0
ASD vs  0 0 0 0 0 0 0 0 0
ASA vs  -0.2 -0.2 -0.2 -0.2 -0.2 -0.2 -0.2 -0.2 -0.2
DS vs  -0.4 -0.4 -0.4 -0.4 -0.4 -0.4 -0.4 -0.4 -0.4
SF vs  0 0 0 0 0 0 0 0 0
Cross-Correlations ZSD vs SF 0 0 0 0 0 0 0 0 0
3GPP
Release 15 64 3GPP TR 38.811 V15.1.0 (2019-06)
ZSA vs SF -0.8 -0.8 -0.8 -0.8 -0.8 -0.8 -0.8 -0.8 -0.8
ZSD vs K 0 0 0 0 0 0 0 0 0
ZSA vs K 0 0 0 0 0 0 0 0 0
ZSD vs DS -0.2 -0.2 -0.2 -0.2 -0.2 -0.2 -0.2 -0.2 -0.2
ZSA vs DS 0 0 0 0 0 0 0 0 0
ZSD vs ASD 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5
ZSA vs ASD 0 0 0 0 0 0 0 0 0
ZSD vs ASA -0.3 -0.3 -0.3 -0.3 -0.3 -0.3 -0.3 -0.3 -0.3
ZSA vs ASA 0.4 0.4 0.4 0.4 0.4 0.4 0.4 0.4 0.4
ZSD vs ZSA 0 0 0 0 0 0 0 0 0
XPR 24.7 24.4 24.4 24.2 23.9 23.3 22.6 21.2 17.6
XPR [dB]
XPR 2.1 2.8 2.7 2.7 3.1 3.9 4.8 6.8 12.7
1.6 1.6 1.6 1.6 1.6 1.6 1.6 1.6 1.6
Cluster DS ( cDS ) in [ns]
0 0 0 0 0 0 0 0 0
Cluster ASD ( c ASD ) in [deg]
11 11 11 11 11 11 11 11 11
Cluster ASA ( c ASA ) in [deg]
7 7 7 7 7 7 7 7 7
Cluster ZSA ( cZSA ) in [deg]
DS 30 30 30 30 30 30 30 30 30
ASD 18 18 18 18 18 18 18 18 18
ASA 15 15 15 15 15 15 15 15 15
Correlation distance in 37 37 37 37 37 37 37 37 37
SF
the horizontal plane [m]
 12 12 12 12 12 12 12 12 12
ZSA 15 15 15 15 15 15 15 15 15
ZSD 15 15 15 15 15 15 15 15 15
factor.
loss model.
parameters X.
NOTE 7: For UMi and frequencies below 2 GHz, use fc = 2 when determining the values of the frequency-dependent
LSP values
Table 6.7.2-2a: Channel model parameters for Dense Urban Scenario (NLOS) in S band
Dense Urban NLOS

Scenarios
10° 20° 30° 40° 50° 60° 70° 80° 90°
lgDS=log10(DS/1s) lgDS 0.82 0.61 0.49 0.49 0.51 0.47 0.44 0.42 0.55
Cross-Correlations ASD vs DS 0.4 0.4 0.4 0.4 0.4 0.4 0.4 0.4 0.4
ASA vs DS 0.6 0.6 0.6 0.6 0.6 0.6 0.6 0.6 0.6
3GPP
Release 15 65 3GPP TR 38.811 V15.1.0 (2019-06)
ASA vs SF 0 0 0 0 0 0 0 0 0
ASD vs SF -0.6 -0.6 -0.6 -0.6 -0.6 -0.6 -0.6 -0.6 -0.6
DS vs SF -0.4 -0.4 -0.4 -0.4 -0.4 -0.4 -0.4 -0.4 -0.4
ASD vs ASA 0.4 0.4 0.4 0.4 0.4 0.4 0.4 0.4 0.4
ASD vs  N/A N/A N/A N/A N/A N/A N/A N/A N/A
ASA vs  N/A N/A N/A N/A N/A N/A N/A N/A N/A
DS vs  N/A N/A N/A N/A N/A N/A N/A N/A N/A
SF vs  N/A N/A N/A N/A N/A N/A N/A N/A N/A
ZSD vs SF 0 0 0 0 0 0 0 0 0
ZSA vs SF -0.4 -0.4 -0.4 -0.4 -0.4 -0.4 -0.4 -0.4 -0.4
ZSD vs K N/A N/A N/A N/A N/A N/A N/A N/A N/A
ZSA vs K N/A N/A N/A N/A N/A N/A N/A N/A N/A
ZSD vs DS -0.5 -0.5 -0.5 -0.5 -0.5 -0.5 -0.5 -0.5 -0.5
Cross-Correlations ZSA vs DS 0 0 0 0 0 0 0 0 0
ZSD vs ASD 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5
ZSA vs ASD -0.1 -0.1 -0.1 -0.1 -0.1 -0.1 -0.1 -0.1 -0.1
ZSD vs ASA 0 0 0 0 0 0 0 0 0
ZSA vs ASA 0 0 0 0 0 0 0 0 0
ZSD vs ZSA 0 0 0 0 0 0 0 0 0
XPR 23.8 21.9 19.7 18.1 16.3 14.0 12.1 8.7 6.4
XPR [dB]
XPR 4.4 6.3 8.1 9.3 11.5 13.3 14.9 17.0 12.3
3.9 3.9 3.9 3.9 3.9 3.9 3.9 3.9 3.9
0 0 0 0 0 0 0 0 0
15 15 15 15 15 15 15 15 15
7 7 7 7 7 7 7 7 7
DS 40 40 40 40 40 40 40 40 40
ASD 50 50 50 50 50 50 50 50 50
ASA 50 50 50 50 50 50 50 50 50
Correlation distance in
SF 50 50 50 50 50 50 50 50 50
 N/A N/A N/A N/A N/A N/A N/A N/A N/A
ZSA 50 50 50 50 50 50 50 50 50
ZSD 50 50 50 50 50 50 50 50 50
factor.
loss model.
parameters X.
Table 6.7.2-2b: Channel model parameters for Dense Urban Scenario (NLOS) in Ka band
Dense Urban NLOS

Scenarios
10° 20° 30° 40° 50° 60° 70° 80° 90°
lgDS=log10(DS/1s) lgDS 0.81 0.61 0.56 0.56 0.57 0.55 0.47 0.42 0.59
3GPP
Release 15 66 3GPP TR 38.811 V15.1.0 (2019-06)
ASD vs DS 0.4 0.4 0.4 0.4 0.4 0.4 0.4 0.4 0.4
ASA vs DS 0.6 0.6 0.6 0.6 0.6 0.6 0.6 0.6 0.6
ASA vs SF 0 0 0 0 0 0 0 0 0
ASD vs SF -0.6 -0.6 -0.6 -0.6 -0.6 -0.6 -0.6 -0.6 -0.6
ASD vs ASA 0.4 0.4 0.4 0.4 0.4 0.4 0.4 0.4 0.4
ZSD vs SF 0 0 0 0 0 0 0 0 0
ZSA vs SF -0.4 -0.4 -0.4 -0.4 -0.4 -0.4 -0.4 -0.4 -0.4
ZSD vs DS -0.5 -0.5 -0.5 -0.5 -0.5 -0.5 -0.5 -0.5 -0.5
ZSD vs ASD 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5
ZSA vs ASD -0.1 -0.1 -0.1 -0.1 -0.1 -0.1 -0.1 -0.1 -0.1
ZSD vs ASA 0 0 0 0 0 0 0 0 0
ZSA vs ASA 0 0 0 0 0 0 0 0 0
ZSD vs ZSA 0 0 0 0 0 0 0 0 0
XPR 23.7 21.8 19.6 18.0 16.3 15.9 12.3 10.5 10.5
XPR [dB]
XPR 4.5 6.3 8.2 9.4 11.5 12.4 15.0 15.7 15.7
3.9 3.9 3.9 3.9 3.9 3.9 3.9 3.9 3.9
0 0 0 0 0 0 0 0 0
15 15 15 15 15 15 15 15 15
7 7 7 7 7 7 7 7 7
DS 40 40 40 40 40 40 40 40 40
ASD 50 50 50 50 50 50 50 50 50
ASA 50 50 50 50 50 50 50 50 50
SF 50 50 50 50 50 50 50 50 50
ZSA 50 50 50 50 50 50 50 50 50
ZSD 50 50 50 50 50 50 50 50 50
factor.
loss model.
parameters X.
3GPP
Release 15 67 3GPP TR 38.811 V15.1.0 (2019-06)
Table 6.7.2-3a: Channel model parameters for Urban Scenario (LOS) at S band
Urban LOS
Scenarios
10° 20° 30° 40° 50° 60° 70° 80° 90°
lgDS=log10(DS/1s) lgDS 1 0.83 0.68 0.48 0.38 0.24 0.18 0.13 0.09
AOA spread (ASA) lgASA 0.18 0.42 0.41 0.18 -0.07 -0.43 -0.64 -0.91 -0.54
ZOA spread (ZSA) lgZSA -0.63 -0.15 0.54 0.35 0.27 0.26 -0.12 -0.21 -0.07
ZOD spread (ZSD) lgZSD -2.54 -2.67 -2.03 -2.28 -2.48 -2.56 -2.96 -3.08 -3
Shadow fading (SF) SF 4 4 4 4 4 4 4 4 4
[dB]
K-factor (K) [dB] K 31.83 18.78 10.49 7.46 6.52 5.47 4.54 4.03 3.68
K 13.84 13.78 10.42 8.01 8.27 7.26 5.53 4.49 3.14
ASD 0.4 0.4 0.4 0.4 0.4 0.4 0.4 0.4 0.4
vs
DS
ASA 0.8 0.8 0.8 0.8 0.8 0.8 0.8 0.8 0.8
vs
DS
ASA -0.5 -0.5 -0.5 -0.5 -0.5 -0.5 -0.5 -0.5 -0.5
vs
SF
ASD -0.5 -0.5 -0.5 -0.5 -0.5 -0.5 -0.5 -0.5 -0.5
vs
SF
Cross-Correlations DS -0.4 -0.4 -0.4 -0.4 -0.4 -0.4 -0.4 -0.4 -0.4
vs
SF
ASD 0 0 0 0 0 0 0 0 0
vs
ASA
ASD 0 0 0 0 0 0 0 0 0
vs 
ASA -0.2 -0.2 -0.2 -0.2 -0.2 -0.2 -0.2 -0.2 -0.2
vs 
DS -0.4 -0.4 -0.4 -0.4 -0.4 -0.4 -0.4 -0.4 -0.4
vs 
SF 0 0 0 0 0 0 0 0 0
vs 
ZSD 0 0 0 0 0 0 0 0 0
vs
SF
ZSA -0.8 -0.8 -0.8 -0.8 -0.8 -0.8 -0.8 -0.8 -0.8
vs
SF
ZSD 0 0 0 0 0 0 0 0 0
vs K
ZSA 0 0 0 0 0 0 0 0 0
vs K
ZSD -0.2 -0.2 -0.2 -0.2 -0.2 -0.2 -0.2 -0.2 -0.2
vs
DS
Cross-Correlations ZSA 0 0 0 0 0 0 0 0 0
vs
DS
ZSD 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5
vs
ASD
ZSA 0 0 0 0 0 0 0 0 0
vs
ASD
ZSD -0.3 -0.3 -0.3 -0.3 -0.3 -0.3 -0.3 -0.3 -0.3
vs
ASA
ZSA 0.4 0.4 0.4 0.4 0.4 0.4 0.4 0.4 0.4
vs
ASA
3GPP
Release 15 68 3GPP TR 38.811 V15.1.0 (2019-06)
ZSD 0 0 0 0 0 0 0 0 0
vs
ZSA
XPR 8 8 8 8 8 8 8 8 8
XPR [dB]
XPR 4 4 4 4 4 4 4 4 4
Number of rays per cluster 20 20 20 20 20 20 20 20 20
M
3.9 3.9 3.9 3.9 3.9 3.9 3.9 3.9 3.9
0.09 0.09 0.12 0.16 0.2 0.28 0.44 0.9 2.87
12.55 12.76 14.36 16.42 17.13 19.01 19.31 22.39 27.8
1.25 3.23 4.39 5.72 6.17 7.36 7.3 7.7 9.25
Per cluster shadowing std  3 3 3 3 3 3 3 3 3
[dB]
DS 30 30 30 30 30 30 30 30 30
ASD 18 18 18 18 18 18 18 18 18
Correlation distance ASA 15 15 15 15 15 15 15 15 15
in the horizontal SF 37 37 37 37 37 37 37 37 37
plane [m]  12 12 12 12 12 12 12 12 12
ZSA 15 15 15 15 15 15 15 15 15
ZSD 15 15 15 15 15 15 15 15 15
NOTE 1: DS = rms delay spread, ASD = rms azimuth spread of departure angles, ASA = rms azimuth spread of arrival angles, ZSD = rms
zenith spread of departure angles, ZSA = rms zenith spread of arrival angles, SF = shadow fading, and K = Ricean K-factor.
NOTE 2: The sign of the shadow fading is defined so that positive SF means more received power at UT than predicted by the path loss
model.
parameters X.
NOTE 8: For satellite (e.g.GEO/LEO), the departure angle spreads are zeros, i.e. µ lgASD and µ lgZSD are –∞, and corresponding standard
deviations are zeros.
Table 6.7.2-3b: Channel model parameters for Urban Scenario (LOS) at Ka band
Urban LOS
Scenarios
10° 20° 30° 40° 50° 60° 70° 80° 90°
lgDS=log10(DS/1s) lgDS 0.92 0.79 0.65 0.48 0.46 0.34 0.27 0.19 0.14
AOA spread (ASA) lgASA -0.4 -0.15 -0.18 -0.31 -0.58 -0.9 -1.16 -1.48 -1.14
ZOA spread (ZSA) lgZSA -0.67 -0.34 0.07 -0.08 -0.21 -0.25 -0.61 -0.79 -0.58
[dB]
K-factor (K) [dB] K 40.18 23.62 12.48 8.56 7.42 5.97 4.88 4.22 3.81
K 16.99 18.96 14.23 11.06 11.21 9.47 7.24 5.79 4.25
ASD 0.4 0.4 0.4 0.4 0.4 0.4 0.4 0.4 0.4
vs
DS
ASA 0.8 0.8 0.8 0.8 0.8 0.8 0.8 0.8 0.8
Cross-Correlations
vs
DS
ASA -0.5 -0.5 -0.5 -0.5 -0.5 -0.5 -0.5 -0.5 -0.5
vs
SF
3GPP
Release 15 69 3GPP TR 38.811 V15.1.0 (2019-06)
ASD -0.5 -0.5 -0.5 -0.5 -0.5 -0.5 -0.5 -0.5 -0.5
vs
SF
DS -0.4 -0.4 -0.4 -0.4 -0.4 -0.4 -0.4 -0.4 -0.4
vs
SF
ASD 0 0 0 0 0 0 0 0 0
vs
ASA
ASD 0 0 0 0 0 0 0 0 0
vs 
ASA -0.2 -0.2 -0.2 -0.2 -0.2 -0.2 -0.2 -0.2 -0.2
vs 
DS -0.4 -0.4 -0.4 -0.4 -0.4 -0.4 -0.4 -0.4 -0.4
vs 
SF 0 0 0 0 0 0 0 0 0
vs 
ZSD 0 0 0 0 0 0 0 0 0
vs
SF
ZSA -0.8 -0.8 -0.8 -0.8 -0.8 -0.8 -0.8 -0.8 -0.8
vs
SF
ZSD 0 0 0 0 0 0 0 0 0
vs K
ZSA 0 0 0 0 0 0 0 0 0
vs K
ZSD -0.2 -0.2 -0.2 -0.2 -0.2 -0.2 -0.2 -0.2 -0.2
vs
DS
ZSA 0 0 0 0 0 0 0 0 0
vs
Cross-Correlations
DS
ZSD 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5
vs
ASD
ZSA 0 0 0 0 0 0 0 0 0
vs
ASD
ZSD -0.3 -0.3 -0.3 -0.3 -0.3 -0.3 -0.3 -0.3 -0.3
vs
ASA
ZSA 0.4 0.4 0.4 0.4 0.4 0.4 0.4 0.4 0.4
vs
ASA
ZSD 0 0 0 0 0 0 0 0 0
vs
ZSA
XPR 8 8 8 8 8 8 8 8 8
XPR [dB]
XPR 4 4 4 4 4 4 4 4 4
4 3 3 3 3 3 3 3 3
Number of clusters N
M
1.6 1.6 1.6 1.6 1.6 1.6 1.6 1.6 1.6
0.09 0.09 0.11 0.15 0.18 0.27 0.42 0.86 2.55
11.78 11.6 13.05 14.56 15.35 16.97 17.96 20.68 25.08
1.14 2.78 3.87 4.94 5.41 6.31 6.66 7.31 9.23
[dB]
DS 30 30 30 30 30 30 30 30 30
ASD 18 18 18 18 18 18 18 18 18
plane [m]  12 12 12 12 12 12 12 12 12
ZSA 15 15 15 15 15 15 15 15 15
ZSD 15 15 15 15 15 15 15 15 15
3GPP
Release 15 70 3GPP TR 38.811 V15.1.0 (2019-06)

model.
parameters X.
Table 6.7.2-4a: Channel model parameters for Urban Scenario (NLOS) at S band
Urban NLOS
Scenarios
10° 20° 30° 40° 50° 60° 70° 80° 90°
lgDS=log10(DS/1s) lgDS 1.19 0.98 0.84 0.73 0.73 0.72 0.71 0.78 0.67
ZOA spread (ZSA) lgZSA -0.97 0.49 1.03 1.12 1.3 1.32 1.35 1.31 1.5
Shadow fading (SF) 6 6 6 6 6 6 6 6 6
SF
[dB]
K-factor (K) [dB] K N/A N/A N/A N/A N/A N/A N/A N/A N/A
K N/A N/A N/A N/A N/A N/A N/A N/A N/A
ASD 0.72 0.59 0.76 0.66 0.82 0.81 0.85 0.77 0.84
vs
DS
ASA 0.47 0.41 0.34 0.35 0.24 0.28 0.26 0.27 0.33
vs
DS
ASA 0 0 0 0 0 0 0 0 0
vs
SF
ASD -0.6 -0.6 -0.6 -0.6 -0.6 -0.6 -0.6 -0.6 -0.6
vs
SF
vs
SF
ASD 0.53 0.48 0.4 0.44 0.27 0.29 0.3 0.37 0.34
vs
ASA
ASD N/A N/A N/A N/A N/A N/A N/A N/A N/A
vs 
ASA N/A N/A N/A N/A N/A N/A N/A N/A N/A
vs 
DS N/A N/A N/A N/A N/A N/A N/A N/A N/A
vs 
SF N/A N/A N/A N/A N/A N/A N/A N/A N/A
vs 
ZSD 0 0 0 0 0 0 0 0 0
vs
SF
ZSA -0.4 -0.4 -0.4 -0.4 -0.4 -0.4 -0.4 -0.4 -0.4
Cross-Correlations vs
SF
ZSD N/A N/A N/A N/A N/A N/A N/A N/A N/A
vs K
ZSA N/A N/A N/A N/A N/A N/A N/A N/A N/A
vs K
3GPP
Release 15 71 3GPP TR 38.811 V15.1.0 (2019-06)
ZSD 0.98 0.97 0.96 0.9 0.89 0.92 0.93 0.89 0.8
vs
DS
ZSA -0.01 0.06 0.06 0.06 -0.05 -0.19 -0.18 -0.39 -0.34
vs
DS
ZSD 0.72 0.6 0.75 0.69 0.7 0.71 0.79 0.77 0.92
vs
ASD
ZSA 0 0.08 0.11 0.12 -0.04 -0.13 -0.17 -0.35 -0.3
vs
ASD
ZSD 0.47 0.43 0.38 0.41 0.34 0.33 0.28 0.3 0.42
vs
ASA
ZSA 0.04 0.15 0.18 0.15 0.15 0.16 -0.03 -0.26 -0.55
vs
ASA
ZSD -0.01 0.06 0.09 0.09 -0.02 -0.14 -0.16 -0.35 -0.36
vs
ZSA
XPR 7 7 7 7 7 7 7 7 7
XPR [dB]
XPR 3 3 3 3 3 3 3 3 3
3 3 3 3 3 3 2 2 2
M
3.9 3.9 3.9 3.9 3.9 3.9 3.9 3.9 3.9
0.08 0.1 0.14 0.23 0.33 0.53 1 1.4 6.63
15.07 16.2 18.14 19.96 21.53 22.44 23.59 26.57 32.7
1.66 4.71 7.33 9.82 11.52 11.75 10.93 12.19 16.68
[dB]
DS 40 40 40 40 40 40 40 40 40
ASD 50 50 50 50 50 50 50 50 50
plane [m]  N/A N/A N/A N/A N/A N/A N/A N/A N/A
ZSA 50 50 50 50 50 50 50 50 50
ZSD 50 50 50 50 50 50 50 50 50
model.
parameters X.
Table 6.7.2-4b: Channel model parameters for Urban Scenario (NLOS) at Ka band
Urban NLOS
Scenarios
10° 20° 30° 40° 50° 60° 70° 80° 90°
lgDS=log10(DS/1s) lgDS 1.26 0.99 0.86 0.75 0.77 0.76 0.73 0.79 0.7
AOD spread (ASD) lgASD -1.58 -1.67 -1.84 -2.02 -2.06 -1.99 -2.19 -1.88 -2
3GPP
Release 15 72 3GPP TR 38.811 V15.1.0 (2019-06)
ZOA spread (ZSA) lgZSA -1.13 0.49 0.95 1.15 1.14 1.13 1.16 1.28 1.42
Shadow fading (SF) [dB] SF 6 6 6 6 6 6 6 6 6
ASD vs DS 0.74 0.6 0.75 0.65 0.72 0.78 0.85 0.8 0.86
ASA vs DS 0.47 0.43 0.35 0.35 0.23 0.24 0.27 0.25 0.51
ASA vs SF 0 0 0 0 0 0 0 0 0
ASD vs SF -0.6 -0.6 -0.6 -0.6 -0.6 -0.6 -0.6 -0.6 -0.6
ASD vs ASA 0.53 0.51 0.41 0.44 0.29 0.24 0.36 0.35 0.47
ZSD vs SF 0 0 0 0 0 0 0 0 0
ZSA vs SF -0.4 -0.4 -0.4 -0.4 -0.4 -0.4 -0.4 -0.4 -0.4
Cross-Correlations ZSD vs DS 0.98 0.97 0.96 0.88 0.84 0.92 0.91 0.88 0.83
ZSA vs DS 0 0.06 0.08 0.1 -0.04 -0.17 -0.16 -0.32 -0.37
ZSD vs ASD 0.72 0.61 0.73 0.67 0.56 0.68 0.79 0.75 0.96
ZSA vs ASD 0.01 0.08 0.12 0.16 0 -0.11 -0.16 -0.31 -0.28
ZSD vs ASA 0.47 0.45 0.38 0.41 0.32 0.29 0.29 0.29 0.46
ZSA vs ASA 0.05 0.15 0.18 0.18 0.21 0.24 0.02 -0.16 -0.43
ZSD vs ZSA 0 0.07 0.11 0.13 -0.02 -0.11 -0.16 -0.3 -0.32
XPR 7 7 7 7 7 7 7 7 7
XPR [dB]
XPR 3 3 3 3 3 3 3 3 3
3 3 3 3 3 3 2 2 2
1.6 1.6 1.6 1.6 1.6 1.6 1.6 1.6 1.6
0.08 0.1 0.14 0.22 0.31 0.49 0.97 1.52 5.36
14.72 14.62 16.4 17.86 19.74 19.73 20.5 26.16 25.83
1.57 4.3 6.64 9.21 10.32 10.3 10.2 12.27 12.75
DS 40 40 40 40 40 40 40 40 40
ASD 50 50 50 50 50 50 50 50 50
ASA 50 50 50 50 50 50 50 50 50
Correlation distance in the
horizontal plane [m]
SF 50 50 50 50 50 50 50 50 50
ZSA 50 50 50 50 50 50 50 50 50
ZSD 50 50 50 50 50 50 50 50 50
factor.
loss model.
parameters X.
NOTE 7: For UMi and frequencies below 2 GHz, use fc = 2 when determining the values of the frequency-dependent
LSP values
3GPP
Release 15 73 3GPP TR 38.811 V15.1.0 (2019-06)
Table 6.7.2-5a: Channel model parameters for Suburban Scenario (LOS) in S band
Suburban LOS
Scenarios
10° 20° 30° 40° 50° 60° 70° 80° 90°
lgDS=log10(DS/1s) lgDS 0.99 0.96 0.79 0.81 1.12 1.23 0.55 0.63 0.63
AOA spread (ASA) lgASA 0.05 -0.38 -0.56 -0.59 -0.58 -0.55 -0.28 -0.17 -0.17
ZOA spread (ZSA) lgZSA -1.78 -1.84 -1.67 -1.59 -1.55 -1.51 -1.27 -1.28 -1.28
K-factor (K) [dB] K  11.40 19.45 20.80 21.20 21.60 19.75 12.00 12.85 12.85
K  6.26 10.32 16.34 15.63 14.22 14.19 5.70 9.91 9.91
ASD vs DS 0.4 0.4 0.4 0.4 0.4 0.4 0.4 0.4 0.4
ASA vs DS 0.8 0.8 0.8 0.8 0.8 0.8 0.8 0.8 0.8
ASA vs SF -0.5 -0.5 -0.5 -0.5 -0.5 -0.5 -0.5 -0.5 -0.5
ASD vs SF -0.5 -0.5 -0.5 -0.5 -0.5 -0.5 -0.5 -0.5 -0.5
ASD vs ASA 0 0 0 0 0 0 0 0 0
ASD vs  0 0 0 0 0 0 0 0 0
ASA vs  -0.2 -0.2 -0.2 -0.2 -0.2 -0.2 -0.2 -0.2 -0.2
DS vs  -0.4 -0.4 -0.4 -0.4 -0.4 -0.4 -0.4 -0.4 -0.4
SF vs  0 0 0 0 0 0 0 0 0
ZSD vs SF 0 0 0 0 0 0 0 0 0
ZSA vs SF -0.8 -0.8 -0.8 -0.8 -0.8 -0.8 -0.8 -0.8 -0.8
ZSD vs K 0 0 0 0 0 0 0 0 0
ZSA vs K 0 0 0 0 0 0 0 0 0
ZSD vs DS -0.2 -0.2 -0.2 -0.2 -0.2 -0.2 -0.2 -0.2 -0.2
ZSD vs ASD 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5
ZSA vs ASD 0 0 0 0 0 0 0 0 0
ZSD vs ASA -0.3 -0.3 -0.3 -0.3 -0.3 -0.3 -0.3 -0.3 -0.3
ZSA vs ASA 0.4 0.4 0.4 0.4 0.4 0.4 0.4 0.4 0.4
ZSD vs ZSA 0 0 0 0 0 0 0 0 0
XPR 21.3 21.0 21.2 21.1 20.7 20.6 20.3 19.8 19.1
XPR [dB]
XPR 7.6 8.9 8.5 8.4 9.2 9.8 10.8 12.2 13.0
3 3 3 3 3 3 2 2 2
1.6 1.6 1.6 1.6 1.6 1.6 1.6 1.6 1.6
0 0 0 0 0 0 0 0 0
11 11 11 11 11 11 11 11 11
7 7 7 7 7 7 7 7 7
DS 30 30 30 30 30 30 30 30 30
ASD 18 18 18 18 18 18 18 18 18
ASA 15 15 15 15 15 15 15 15 15
SF 37 37 37 37 37 37 37 37 37
 12 12 12 12 12 12 12 12 12
ZSA 15 15 15 15 15 15 15 15 15
ZSD 15 15 15 15 15 15 15 15 15
3GPP
Release 15 74 3GPP TR 38.811 V15.1.0 (2019-06)

factor.
loss model.
parameters X.
3GPP
Release 15 75 3GPP TR 38.811 V15.1.0 (2019-06)
Table 6.7.2-5b: Channel model parameters for Suburban Scenario (LOS) in Ka band
Suburban LOS
Scenarios
10° 20° 30° 40° 50° 60° 70° 80° 90°
lgDS=log10(DS/1s) lgDS 0.46 0.45 0.28 0.17 0.14 0.15 0.09 0.05 0.02
AOA spread (ASA) lgASA 0.89 0.31 0.02 -0.10 -0.19 -0.54 -0.24 -0.52 -0.15
K-factor (K) [dB] K  8.9 14.0 11.3 9.0 7.5 6.6 5.9 5.5 5.4
K  4.4 4.6 3.7 3.5 3.0 2.6 1.7 0.7 0.3
ASD vs DS 0.4 0.4 0.4 0.4 0.4 0.4 0.4 0.4 0.4
ASA vs DS 0.8 0.8 0.8 0.8 0.8 0.8 0.8 0.8 0.8
ASA vs SF -0.5 -0.5 -0.5 -0.5 -0.5 -0.5 -0.5 -0.5 -0.5
ASD vs SF -0.5 -0.5 -0.5 -0.5 -0.5 -0.5 -0.5 -0.5 -0.5
ASD vs ASA 0 0 0 0 0 0 0 0 0
ASD vs  0 0 0 0 0 0 0 0 0
ASA vs  -0.2 -0.2 -0.2 -0.2 -0.2 -0.2 -0.2 -0.2 -0.2
DS vs  -0.4 -0.4 -0.4 -0.4 -0.4 -0.4 -0.4 -0.4 -0.4
SF vs  0 0 0 0 0 0 0 0 0
ZSD vs SF 0 0 0 0 0 0 0 0 0
ZSA vs SF -0.8 -0.8 -0.8 -0.8 -0.8 -0.8 -0.8 -0.8 -0.8
ZSD vs K 0 0 0 0 0 0 0 0 0
ZSA vs K 0 0 0 0 0 0 0 0 0
ZSD vs DS -0.2 -0.2 -0.2 -0.2 -0.2 -0.2 -0.2 -0.2 -0.2
ZSD vs ASD 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5
ZSA vs ASD 0 0 0 0 0 0 0 0 0
ZSD vs ASA -0.3 -0.3 -0.3 -0.3 -0.3 -0.3 -0.3 -0.3 -0.3
ZSA vs ASA 0.4 0.4 0.4 0.4 0.4 0.4 0.4 0.4 0.4
ZSD vs ZSA 0 0 0 0 0 0 0 0 0
XPR 23.2 23.6 23.5 23.4 23.2 23.3 23.4 23.2 23.1
XPR [dB]
XPR 5.0 4.5 4.7 5.2 5.7 5.9 6.2 7.0 7.6
3 3 3 3 3 3 2 2 2
1.6 1.6 1.6 1.6 1.6 1.6 1.6 1.6 1.6
0 0 0 0 0 0 0 0 0
11 11 11 11 11 11 11 11 11
7 7 7 7 7 7 7 7 7
DS 30 30 30 30 30 30 30 30 30
ASD 18 18 18 18 18 18 18 18 18
ASA 15 15 15 15 15 15 15 15 15
SF 37 37 37 37 37 37 37 37 37
 12 12 12 12 12 12 12 12 12
ZSA 15 15 15 15 15 15 15 15 15
ZSD 15 15 15 15 15 15 15 15 15
3GPP
Release 15 76 3GPP TR 38.811 V15.1.0 (2019-06)

factor.
loss model.
parameters X.
Table 6.7.2-6a: Channel model parameters for Suburban Scenario (NLOS) in S band
Suburban NLOS
Scenarios
10° 20° 30° 40° 50° 60° 70° 80° 90°
lgDS=log10(DS/1s) lgDS 1.42 1.46 1.46 1.21 1.18 1.13 1.37 1.50 1.50
ASD vs DS 0.4 0.4 0.4 0.4 0.4 0.4 0.4 0.4 0.4
ASA vs DS 0.6 0.6 0.6 0.6 0.6 0.6 0.6 0.6 0.6
ASA vs SF 0 0 0 0 0 0 0 0 0
ASD vs SF -0.6 -0.6 -0.6 -0.6 -0.6 -0.6 -0.6 -0.6 -0.6
ASD vs ASA 0.4 0.4 0.4 0.4 0.4 0.4 0.4 0.4 0.4
ZSD vs SF 0 0 0 0 0 0 0 0 0
ZSA vs SF -0.4 -0.4 -0.4 -0.4 -0.4 -0.4 -0.4 -0.4 -0.4
ZSD vs DS -0.5 -0.5 -0.5 -0.5 -0.5 -0.5 -0.5 -0.5 -0.5
ZSD vs ASD 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5
ZSA vs ASD -0.1 -0.1 -0.1 -0.1 -0.1 -0.1 -0.1 -0.1 -0.1
ZSD vs ASA 0 0 0 0 0 0 0 0 0
ZSA vs ASA 0 0 0 0 0 0 0 0 0
ZSD vs ZSA 0 0 0 0 0 0 0 0 0
XPR 20.6 16.7 13.2 11.3 9.6 7.5 9.1 11.7 11.7
XPR [dB]
XPR 8.5 12.0 12.8 13.8 12.5 11.2 10.1 13.1 13.1
3GPP
Release 15 77 3GPP TR 38.811 V15.1.0 (2019-06)
4 4 4 4 4 3 3 3 3
1.6 1.6 1.6 1.6 1.6 1.6 1.6 1.6 1.6
0 0 0 0 0 0 0 0 0
15 15 15 15 15 15 15 15 15
7 7 7 7 7 7 7 7 7
DS 40 40 40 40 40 40 40 40 40
ASD 50 50 50 50 50 50 50 50 50
ASA 50 50 50 50 50 50 50 50 50
SF 50 50 50 50 50 50 50 50 50
ZSA 50 50 50 50 50 50 50 50 50
ZSD 50 50 50 50 50 50 50 50 50
factor.
loss model.
parameters X.
Table 6.7.2-6b: Channel model parameters for Suburban Scenario (NLOS) in Ka band
Suburban NLOS
Scenarios
10° 20° 30° 40° 50° 60° 70° 80° 90°
lgDS=log10(DS/1s) lgDS 0.50 0.61 0.56 0.58 0.59 0.47 0.36 0.29 0.29
AOA spread (ASA) lgASA 1.49 1.24 1.06 0.91 0.98 0.49 0.73 -0.04 -0.04
ASD vs DS 0.4 0.4 0.4 0.4 0.4 0.4 0.4 0.4 0.4
ASA vs DS 0.6 0.6 0.6 0.6 0.6 0.6 0.6 0.6 0.6
ASA vs SF 0 0 0 0 0 0 0 0 0
ASD vs SF -0.6 -0.6 -0.6 -0.6 -0.6 -0.6 -0.6 -0.6 -0.6
ASD vs ASA 0.4 0.4 0.4 0.4 0.4 0.4 0.4 0.4 0.4
ZSD vs SF 0 0 0 0 0 0 0 0 0
ZSA vs SF -0.4 -0.4 -0.4 -0.4 -0.4 -0.4 -0.4 -0.4 -0.4
Cross-Correlations ZSA vs K N/A N/A N/A N/A N/A N/A N/A N/A N/A
ZSD vs DS -0.5 -0.5 -0.5 -0.5 -0.5 -0.5 -0.5 -0.5 -0.5
ZSA vs DS 0 0 0 0 0 0 0 0 0
ZSD vs ASD 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5
3GPP
Release 15 78 3GPP TR 38.811 V15.1.0 (2019-06)
ZSA vs ASD -0.1 -0.1 -0.1 -0.1 -0.1 -0.1 -0.1 -0.1 -0.1
ZSD vs ASA 0 0 0 0 0 0 0 0 0
ZSA vs ASA 0 0 0 0 0 0 0 0 0
ZSD vs ZSA 0 0 0 0 0 0 0 0 0
XPR 22.5 19.4 15.5 13.9 11.7 9.8 10.3 15.6 15.6
XPR [dB]
XPR 5.0 8.5 10.0 10.6 10.0 9.1 9.1 9.1 9.1
4 4 4 4 4 3 3 3 3
1.6 1.6 1.6 1.6 1.6 1.6 1.6 1.6 1.6
0 0 0 0 0 0 0 0 0
15 15 15 15 15 15 15 15 15
7 7 7 7 7 7 7 7 7
DS 40 40 40 40 40 40 40 40 40
ASD 50 50 50 50 50 50 50 50 50
ASA 50 50 50 50 50 50 50 50 50
SF 50 50 50 50 50 50 50 50 50
ZSA 50 50 50 50 50 50 50 50 50
ZSD 50 50 50 50 50 50 50 50 50
factor.
loss model.
parameters X.
Table 6.7.2-7a: Channel model parameters for Rural Scenario (LOS) at S band
Rural LOS
Scenarios
10° 20° 30° 40° 50° 60° 70° 80° 90°
lgDS=log10(DS/1s) lgDS 0.66 0.44 0.28 0.19 0.14 0.1 0.1 0.05 0.06
AOD spread (ASD) lgASD -3.42 -3 -2.86 -2.78 -2.7 -2.66 -2.53 -2.21 -1.78
lgZSA=log10(ZSA/1) lgZSA 6.3 5.04 0.81 1.16 0.82 0.67 1 1.18 0.91
[dB]
K-factor (K) [dB] K 24.72 12.31 8.05 6.21 5.04 4.42 3.92 3.65 3.59
K 5.07 5.75 5.46 5.23 3.95 3.75 2.56 1.77 1.77
ASD 0 0 0 0 0 0 0 0 0
vs
DS
ASA 0 0 0 0 0 0 0 0 0
Cross-Correlations
vs
DS
ASA 0 0 0 0 0 0 0 0 0
vs
SF
3GPP
Release 15 79 3GPP TR 38.811 V15.1.0 (2019-06)
ASD 0 0 0 0 0 0 0 0 0
vs
SF
DS -0.5 -0.5 -0.5 -0.5 -0.5 -0.5 -0.5 -0.5 -0.5
vs
SF
ASD 0 0 0 0 0 0 0 0 0
vs
ASA
ASD 0 0 0 0 0 0 0 0 0
vs 
ASA 0 0 0 0 0 0 0 0 0
vs 
DS 0 0 0 0 0 0 0 0 0
vs 
SF 0 0 0 0 0 0 0 0 0
vs 
ZSD
vs 0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.01
SF
ZSA
vs -0.17 -0.17 -0.17 -0.17 -0.17 -0.17 -0.17 -0.17 -0.17
SF
ZSD
vs K
0 0 0 0 0 0 0 0 0
ZSA
vs K
-0.02 -0.02 -0.02 -0.02 -0.02 -0.02 -0.02 -0.02 -0.02
ZSD
vs -0.05 -0.05 -0.05 -0.05 -0.05 -0.05 -0.05 -0.05 -0.05
DS
ZSA
Cross-Correlations
vs 0.27 0.27 0.27 0.27 0.27 0.27 0.27 0.27 0.27
DS
ZSD
vs 0.73 0.73 0.73 0.73 0.73 0.73 0.73 0.73 0.73
ASD
ZSA
vs -0.14 -0.14 -0.14 -0.14 -0.14 -0.14 -0.14 -0.14 -0.14
ASD
ZSD
vs -0.20 -0.20 -0.20 -0.20 -0.20 -0.20 -0.20 -0.20 -0.20
ASA
ZSA
vs 0.24 0.24 0.24 0.24 0.24 0.24 0.24 0.24 0.24
ASA
ZSD
vs -0.07 -0.07 -0.07 -0.07 -0.07 -0.07 -0.07 -0.07 -0.07
ZSA
XPR 12 12 12 12 12 12 12 12 12
XPR [dB]
XPR 4 4 4 4 4 4 4 4 4
2 2 2 2 2 2 2 2 2
M
N/A N/A N/A N/A N/A N/A N/A N/A N/A
0.39 0.31 0.29 0.37 0.61 0.9 1.43 2.87 5.48
10.81 8.09 13.7 20.05 24.51 26.35 31.84 36.62 36.77
1.94 1.83 2.28 2.93 2.84 3.17 3.88 4.17 4.29
[dB]
DS 50 50 50 50 50 50 50 50 50
ASD 25 25 25 25 25 25 25 25 25
plane [m]  40 40 40 40 40 40 40 40 40
ZSA 15 15 15 15 15 15 15 15 15
ZSD 15 15 15 15 15 15 15 15 15
3GPP
Release 15 80 3GPP TR 38.811 V15.1.0 (2019-06)

model.
parameters X.
NOTE 7: For UMi and frequencies below 2 GHz, use fc = 2 when determining the values of the frequency-dependent LSP
values
Table 6.7.2-7b: Channel model parameters for Rural Scenario (LOS) at Ka band.
Rural LOS
Scenarios
10° 20° 30° 40° 50° 60° 70° 80° 90°
lgDS=log10(DS/1s) lgDS 0.46 0.29 0.18 0.19 0.14 0.15 0.13 0.03 0.07
[dB]
K-factor (K) [dB] K 25.43 12.72 8.40 6.52 5.24 4.57 4.02 3.70 3.62
K 7.04 7.47 7.18 6.88 5.28 4.92 3.40 2.22 2.28
ASD 0 0 0 0 0 0 0 0 0
vs
DS
ASA 0 0 0 0 0 0 0 0 0
vs
DS
ASA 0 0 0 0 0 0 0 0 0
vs
SF
ASD 0 0 0 0 0 0 0 0 0
vs
SF
vs
SF
ASD 0 0 0 0 0 0 0 0 0
vs
ASA
ASD 0 0 0 0 0 0 0 0 0
vs 
ASA 0 0 0 0 0 0 0 0 0
vs 
DS 0 0 0 0 0 0 0 0 0
vs 
SF 0 0 0 0 0 0 0 0 0
vs 
ZSD
vs 0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.01
SF
ZSA
Cross-Correlations vs -0.17 -0.17 -0.17 -0.17 -0.17 -0.17 -0.17 -0.17 -0.17
SF
ZSD
vs K
0 0 0 0 0 0 0 0 0
ZSA
vs K
-0.02 -0.02 -0.02 -0.02 -0.02 -0.02 -0.02 -0.02 -0.02
3GPP
Release 15 81 3GPP TR 38.811 V15.1.0 (2019-06)
ZSD
vs -0.05 -0.05 -0.05 -0.05 -0.05 -0.05 -0.05 -0.05 -0.05
DS
ZSA
vs 0.27 0.27 0.27 0.27 0.27 0.27 0.27 0.27 0.27
DS
ZSD
vs 0.73 0.73 0.73 0.73 0.73 0.73 0.73 0.73 0.73
ASD
ZSA
vs -0.14 -0.14 -0.14 -0.14 -0.14 -0.14 -0.14 -0.14 -0.14
ASD
ZSD
vs -0.20 -0.20 -0.20 -0.20 -0.20 -0.20 -0.20 -0.20 -0.20
ASA
ZSA
vs 0.24 0.24 0.24 0.24 0.24 0.24 0.24 0.24 0.24
ASA
ZSD
vs -0.07 -0.07 -0.07 -0.07 -0.07 -0.07 -0.07 -0.07 -0.07
ZSA
XPR 12 12 12 12 12 12 12 12 12
XPR [dB]
XPR 4 4 4 4 4 4 4 4 4
2 2 2 2 2 2 2 2 2
M
0.36 0.3 0.25 0.35 0.53 0.88 1.39 2.7 4.97
4.63 6.83 12.91 18.9 22.44 25.69 27.95 31.45 28.01
0.75 1.25 1.93 2.37 2.66 3.23 3.71 4.17 4.14
[dB]
DS 50 50 50 50 50 50 50 50 50
ASD 25 25 25 25 25 25 25 25 25
plane [m]  40 40 40 40 40 40 40 40 40
ZSA 15 15 15 15 15 15 15 15 15
ZSD 15 15 15 15 15 15 15 15 15
model.
parameters X.
Table 6.7.2-8a: Channel model parameters for Rural Scenario (NLOS) at S band
Rural NLOS
Scenarios
10° 20° 30° 40° 50° 60° 70° 80° 90°
lgDS=log10(DS/1s) lgDS 1.59 0.95 0.92 0.92 0.87 0.87 0.82 0.86 0.81
AOA spread (ASA) lgASA -3.33 -0.74 0.08 0.32 0.53 0.33 0.55 0.45 0.4
3GPP
Release 15 82 3GPP TR 38.811 V15.1.0 (2019-06)
ZOA spread (ZSA) lgZSA -0.88 -0.07 0.75 0.72 0.95 0.97 1.1 0.97 1.35
lgZSA=log10(ZSD/1) lgZSD 3.96 4.07 1.7 2.01 2 1.82 1.66 2.58 2.52
[dB]
ASD 0.5 0.3 0.32 0.39 0.25 0.16 0.23 0.18 0.91
vs
DS
ASA 0.2 0.3 0.31 0.38 0.44 0.38 0.49 0.27 -0.01
vs
DS
ASA 0 0 0 0 0 0 0 0 0
vs
SF
ASD 0.6 0.6 0.6 0.6 0.6 0.6 0.6 0.6 0.6
vs
SF
vs
SF
ASD 0.57 0.14 0.08 0.25 0.17 0.15 0.23 -0.04 -0.05
vs
ASA
vs 
vs 
vs 
vs 
ZSD
vs -0.04 -0.04 -0.04 -0.04 -0.04 -0.04 -0.04 -0.04 -0.04
SF
ZSA
vs -0.25 -0.25 -0.25 -0.25 -0.25 -0.25 -0.25 -0.25 -0.25
SF
ZSD
vs K
ZSA
vs K
ZSD 0.75 0.79 0.74 0.9 0.93 0.93 0.94 0.74 0.61
vs
DS
ZSA 0.05 0.03 0 -0.09 -0.2 -0.22 -0.33 -0.41 -0.36
vs
Cross-Correlations
DS
ZSD 0.7 0.42 0.38 0.33 0.19 0.17 0.21 0.15 0.31
vs
ASD
ZSA 0.12 -0.04 -0.1 -0.13 -0.13 -0.12 -0.13 -0.14 -0.29
vs
ASD
ZSD 0.35 0.35 0.31 0.38 0.47 0.45 0.52 0.28 0.1
vs
ASA
ZSA 0.09 0.21 0.22 0.06 -0.06 -0.15 -0.36 -0.28 -0.4
vs
ASA
ZSD 0.15 0.07 -0.06 -0.13 -0.21 -0.22 -0.31 -0.33 -0.27
vs
ZSA
XPR 7 7 7 7 7 7 7 7 7
XPR [dB]
XPR 3 3 3 3 3 3 3 3 3
3GPP
Release 15 83 3GPP TR 38.811 V15.1.0 (2019-06)
3 3 2 2 2 2 2 2 2
M
0.03 0.05 0.07 0.1 0.15 0.22 0.5 1.04 2.11
18.16 26.82 21.99 22.86 25.93 27.79 28.5 37.53 29.23
2.32 7.34 8.28 8.76 9.68 9.94 8.9 13.74 12.16
[dB]
DS 36 36 36 36 36 36 36 36 36
ASD 30 30 30 30 30 30 30 30 30
ZSA 50 50 50 50 50 50 50 50 50
ZSD 50 50 50 50 50 50 50 50 50
model.
parameters X.
NOTE 6: For UMa and frequencies below 6 GHz, use fc = 6 when determining the values of the frequency-dependent LSP values.
Table 6.7.2-8b: Channel model parameters for Rural Scenario (NLOS) at Ka band
Rural NLOS
Scenarios
10° 20° 30° 40° 50° 60° 70° 80° 90°
lgDS=log10(DS/1s) lgDS 1.91 0.94 0.92 0.94 0.89 0.87 0.82 1.65 0.76
AOA spread (ASA) lgASA -3.4 -0.51 0.06 0.2 0.4 0.32 0.46 0.33 0.24
ZOA spread (ZSA) lgZSA -1.19 -0.11 0.72 0.69 0.84 0.99 0.95 0.92 1.29
[dB]
ASD 0.51 0.36 0.29 0.41 0.27 0.19 0.44 0.2 0.9
vs
DS
ASA 0.22 0.33 0.32 0.41 0.47 0.37 0.34 0.28 -0.02
vs
DS
ASA 0 0 0 0 0 0 0 0 0
Cross-Correlations
vs
SF
ASD 0.6 0.6 0.6 0.6 0.6 0.6 0.6 0.6 0.6
vs
SF
DS -0.5 -0.5 -0.5 -0.5 -0.5 -0.5 -0.5 -0.5 -0.5
vs
SF
3GPP
Release 15 84 3GPP TR 38.811 V15.1.0 (2019-06)
ASD 0.57 0.15 0.09 0.24 0.11 0.16 0.26 -0.03 -0.03
vs
ASA
vs 
vs 
vs 
vs 
ZSD
vs -0.04 -0.04 -0.04 -0.04 -0.04 -0.04 -0.04 -0.04 -0.04
SF
ZSA
vs -0.25 -0.25 -0.25 -0.25 -0.25 -0.25 -0.25 -0.25 -0.25
SF
ZSD
vs K
ZSA
vs K
ZSD 0.72 0.78 0.72 0.91 0.91 0.86 0.93 0.82 0.75
vs
DS
ZSA 0.05 0.02 0.04 -0.06 -0.16 -0.17 -0.31 -0.35 -0.29
vs
Cross-Correlations
DS
ZSD 0.71 0.42 0.4 0.45 0.15 0.2 0.32 0.11 0.44
vs
ASD
ZSA 0.1 -0.04 -0.07 -0.11 -0.12 -0.1 -0.13 -0.14 -0.31
vs
ASD
ZSD 0.4 0.35 0.33 0.4 0.47 0.46 0.35 0.3 0.08
vs
ASA
ZSA 0.1 0.21 0.22 0.1 0 -0.11 -0.25 -0.18 -0.21
vs
ASA
ZSD 0.19 0.06 -0.02 -0.11 -0.19 -0.19 -0.3 -0.28 -0.22
vs
ZSA
XPR 7 7 7 7 7 7 7 7 7
XPR [dB]
XPR 3 3 3 3 3 3 3 3 3
3 3 2 2 2 2 2 2 2
M
0.03 0.05 0.07 0.09 0.16 0.22 0.51 0.89 1.68
18.21 24.08 22.06 21.4 24.26 24.15 25.99 36.07 24.51
2.13 6.52 7.72 8.45 8.92 8.76 9 13.6 10.56
[dB]
DS 36 36 36 36 36 36 36 36 36
ASD 30 30 30 30 30 30 30 30 30
ZSA 50 50 50 50 50 50 50 50 50
ZSD 50 50 50 50 50 50 50 50 50
3GPP
Release 15 85 3GPP TR 38.811 V15.1.0 (2019-06)

model.
parameters X.
NOTE 6: For UMa and frequencies below 6 GHz, use fc = 6 when determining the values of the frequency-dependent LSP values.
NOTE 7: For UMi and frequencies below 2 GHz, use fc = 2 when determining the values of the frequency-dependent LSP values.
6.8. Additional modelling components

6.8.1 Time-varying Doppler shift
The Doppler shift generally depends on the time evolution of the channel as the joint results due to the movement of Tx
and Rx, or scatterer movement. As mentioned above, the movement of BS and UE are time-varying, especially in the
spaceborne case. Then, the more general form to describe the phase rotation due to the Doppler shift can be calculated
as:
 rˆ
t T
(t )  v  t    rˆ (t )  vsat  t  
t T
exp  j 2  rx ,n ,m dt   exp  j 2  tx ,n ,m dt  .
     
 t0 0   t0 0 
Here, rˆrx ,n ,m (t ) is the normalized vector that points into the direction of the incoming wave as seen from the Rx at time

t . v t  denotes the velocity vector of the Rx at time t . rˆtx,n,m (t ) is the normalized vector that points into the direction
of the outgoing wave as seen from the Tx at time t . vsat  t  denotes the velocity vector of the Tx at time t . While t0
denotes a reference point in time that defines the initial phase, e.g. t0  0 .
6.8.2 Faraday rotation

The Faraday rotation is introduced to describe the rotation of the polarization due to the interaction of the
electromagnetic wave with the ionized medium in the earth's magnetic field along the path. In case of propagation for
spaceborne BS above the ionosphere, the Faraday rotation should be calculated as:
cos( n ,m ) sin( n ,m ) 
Fr   
 sin( n ,m ) cos( n ,m ) (6.8-1)
The equations 7.5-28 and 7.5-29 in [12] should be updated by post-multiplied with Fr for the channel coefficient
generation of the mth path in the nth cluster, as equations 6.8-1a and 6.8-1b, respectively:
NLOS
(t ) 
Pn  Frx ,u ,  n ,m ,ZOA , n ,m , AOA 


exp j
T
n ,m  
 n ,m 1 exp j 
n ,m



 rx ,u ,  n ,m ,ZOA , n ,m, AOA 
H F 
u , s ,n ,m
M   n ,m 1 exp j
 n ,m  exp  j 
n ,m 


cos( n ,m ) sin( n ,m )   Ftx,s ,  n ,m,ZOD , n ,m , AOD   rˆT .d   rˆT .d   rˆT .v 
 sin( ) cos( )  F  exp j 2 rx ,n ,m rx ,u  exp j 2 tx,n ,m tx,s  exp j 2 rx ,n ,m t 
n ,m , ZOD , n ,m , AOD 

 n ,m n ,m   tx, s ,  0   0   0 
(6.8-1a)
3GPP
Release 15 86 3GPP TR 38.811 V15.1.0 (2019-06)
 Frx ,u ,  LOS , ZOA ,  LOS , AOA  1 0  cos( n , m ) sin( n , m )   Ftx, s ,  LOS , ZOD ,  LOS , AOD 
T
LOS
(t )  
 Frx ,u ,  LOS , ZOA ,  LOS , AOA  0  1  sin( n , m ) cos( n , m )  Ftx, s ,  LOS , ZOD ,  LOS , AOD  (6.8-1b)
H u , s ,1     
 d   rˆrxT , LOS .d rx ,u   rˆ T .d   rˆ T .v 
 exp  j 2 3D 
 exp j 2  exp j 2 tx, LOS tx, s  exp j 2 rx , LOS t
0  0   0   0 
       
where  n,m is calculated as [11]:
108

f² (6.8-2)
where ψ is the faraday rotation in degree and f is the central carrier frequency in GHz.
3GPP
Release 15 87 3GPP TR 38.811 V15.1.0 (2019-06)
6.9 Channel models for link level simulations

The link level models follow the same principles as the link level models specified in TR38.901 section 7.7.
6.9.1 CDL models

The CDL models are defined for the S and Ka bands and are applicable to different environments and elevation angles.
NTN-CDL-A and NTN-CDL-B are constructed to represent two different channel profiles for NLOS, while NTN-CDL-
C and NTN-CDL-D are constructed for LOS. The parameters of these models can be found in Tables 6.9.1-1 to 6.9.1-4.
Table 6.9.1-1a: Void
Table 6.9.1-1b: Void
Due to the long propagation distance between the NTN gNB and ground UE, the azimuth and elevation angular spreads
of departure, ASD and ZSD, can be considered zero, and all cluster AOD and ZOD angles the same. The coordinate
system can be chosen in such a way that all AODs are zero. With the NTN gNB at an elevation angle α with respect to
the UE, ZOD of all clusters is 90⁰+α, as shown in Figure 6.9.1-1. The CDL models in Tables 6.9.1-1 to 6.9.1-4 use 50⁰
elevation angle between the NTN gNB and UE, resulting in ZODs equal to 190⁰. For a desired elevation angle αdesired,
the cluster ZOD needs to be set to
𝜃𝑛,ZOD ,desired = 90 + 𝛼desired

(6.9-1)
Figure 6.9.1-1 Elevation angle α at UE with NTN gNB
Depending on the frequency band, environment scenario, and elevation angle of the intended link level simulations,
suitable DS, ASA, ZSA, and Rician K-factor in case of LOS, can be determined from the channel model parameters in
Section 6.7.2.
Each CDL model can be scaled in delay and AOA to achieve desired RMS delay spread and ASA according to the
procedures specified respectively in subclauses 7.7.3 and 7.7.5.1 of TR 38.901 [12]. With the same angle scaling
principle, each CDL model can also be scaled in ZOA to achieve desired ZSA, taking into account the difference
between the desired elevation angle and reference elevation angle, by
ZSAdesired
𝜃𝑛,ZOA ,scaled = 𝜃𝑛,ZOA ,model − 𝜇ZOA ,model + 𝜇ZOA ,desired − Δ𝛼
ZSAmodel
(6.9-2)
3GPP
Release 15 88 3GPP TR 38.811 V15.1.0 (2019-06)
where
𝜃𝑛,ZOA ,scaled
is the scaled ZOA,
ZSAdesired
is the desired ZSA,
ZSAmodel
is the RMS zenith angular spread of the reference model,
𝜃𝑛,ZOA ,model
is the cluster ZOA of the reference model,
𝜇ZOA ,model
is the mean angle of ZOA of the reference model,
𝜇ZOA ,desired
is the desired mean ZOA,
Δ𝛼 = 𝛼desired − 𝛼model
is the difference between the desired and reference elevation angles between the
NTN gNB and UE.
The resultant ZOA angles after scaling may need to be wrapped to the domain [0,180] by the same rule in TR 38.901: if
𝜃𝑛,ZOA ∈ [1800 , 3600 ] 3600 − 𝜃𝑛,ZOA
, it should be set to .
For LOS channel models, the K-factor of NTN-CDL-C and NTN-CDL-D can be set to a desired value following the
procedure described in subclause 7.7.6 of TR 38.901.
Doppler shift due to UE and satellite motion should be calculated based on section 6.8.1. For simulations over a few
TTIs, constant speed for the UE and the satellite and constant satellite elevation angle may be considered
𝜶𝐦𝐨𝐝𝐞𝐥 = 𝟓𝟎°
Table 6.9.1-1 NTN-CDL-A at elevation
Cluster # Normalized delay Power in [dB] AOD in [°] AOA in [°] ZOD in [°] ZOA in [°]
1 0 0 0 178.8 140 35.6
2 1.0811 -4.675 0 -115.7 140 22.9
3 2.8416 -6.482 0 111.5 140 127.4
Per-Cluster Parameters
Parameter cASD in [°] cASA in [°] cZSD in [°] cZSA in [°] XPR in [dB]
Value 0 15 0 7 10
Table 6.9.1-2 NTN-CDL-B at elevation
Cluster # Normalized delay Power in [dB] AOD in [°] AOA in [°] ZOD in [°] ZOA in [°]
1 0 0 0 -174.6 140 42.2
2 0.7249 -1.973 0 144.9 140 63.4
3 0.7410 -4.332 0 -119.8 140 89.7
4 5.7392 -11.914 0 -88.8 140 174.1
Value 0 15 0 7 10
3GPP
Release 15 89 3GPP TR 38.811 V15.1.0 (2019-06)
Table 6.9.1-3 NTN-CDL-C at elevation
Cluster # Cluster PAS Normalized Delay Power in [dB] AOD in [°] AOA in [°] ZOD in [°] ZOA in [°]
Specular(LOS path) 0 -0.394 0 -180 140 40
1
Laplacian 0 -10.618 0 -180 140 40
2 Laplacian 14.8124 -23.373 0 -75.9 140 87.1
Value 0 11 0 7 16
Table 6.9.1-4 NTN-CDL-D at elevation
Cluster # Cluster PAS Normalized Delay Power in [dB] AOD in [°] AOA in [°] ZOD in [°] ZOA in [°]
Specular(LOS path) 0 -0.284 0 -180 140 40
1
Laplacian 0 -11.991 0 -180 140 40
2 Laplacian 0.5596 -9.887 0 -135.4 140 146.2
3 Laplacian 7.3340 -16.771 0 -121.5 140 136.0
Value 0 11 0 7 16
3GPP
Release 15 90 3GPP TR 38.811 V15.1.0 (2019-06)
6.9.2 TDL models

The Tapped Delay Line (TDL) models are filtered from the CDL models according to the section 7.7.4 of TR 38.901
[12] by assuming isotropic UE antenna. Two TDL models, namely NTN-TDL-A and NTN-TDL-B are constructed to
represent two different channel profiles for NLOS, while NTN-TDL-C and NTN-TDL-D are constructed for LOS. The
parameter of these models can be found in Tables 6.9.2-1 to 6.9.2-4.
Table 6.9.2-1a: Void
Table 6.9.2-1b: Void
The Doppler spectrum for each tap is defined as described in subclause 7.7.2 of TR 38.901. Each TDL model can be
scaled in delay to achieve desired RMS delay spread according to the procedure specified in subclause 7.7.3 of TR
38.901. For LOS channel models, the K-factor of NTN-TDL-C and NTN-TDL-D can be set to a desired value
following the procedure described in subclause 7.7.6 of TR 38.901.
Additional Doppler shift due to satellite motion should be taken into account according to the following formula:
 R 
f d , shift  (vsat / c)   cos  mod el   f c ,
 Rh 
Where vsat denotes the satellite speed, c denotes the speed of light, R denotes the earth radius, h denotes the satellite
altitude,  mod el denotes the satellite elevation angle, and f c denotes the carrier frequency.
This additional Doppler shift should be applied to all taps of the TDL model.
The satellite speed, satellite elevation angle and UE speed should be considered to be constant during the simulation
duration, if limited to few TTIs.
An illustration of the effect of additional Doppler shift due to satellite motion on the Doppler power spectrum is
displayed on the next figure.
3GPP
Release 15 91 3GPP TR 38.811 V15.1.0 (2019-06)
Pure Doppler peak

due to mobile movement
P fc + fd,shift + (v/c)*(fc + fd,shift)*(cos(Ɵ))
f
fc fc+ fd,shift
Doppler spread due to local scattering

around the terrestrial device
Carrier frequency shift

due to the satellite movement
Figure 6.9.2-1: Illustration of Doppler power spectrum in NTN in LOS conditions
Table 6.9.2-1. NTN-TDL-A at elevation
Tap # Normalized delay Power in [dB] Fading distribution

1 0 0 Rayleigh
2 1.0811 -4.675 Rayleigh
3 2.8416 -6.482 Rayleigh
Table 6.9.2-2. NTN-TDL-B at elevation

1 0 0 Rayleigh
2 0.7249 -1.973 Rayleigh
3 0.7410 -4.332 Rayleigh
4 5.7392 -11.914 Rayleigh
3GPP
Release 15 92 3GPP TR 38.811 V15.1.0 (2019-06)
Table 6.9.2-3. NTN-TDL-C at elevation

0 -0.394 LOS path
1
0 -10.618 Rayleigh
2 14.8124 -23.373 Rayleigh
NOTE: The first tap follows a Ricean distribution with a K-factor of K1 = 10.224 dB and a mean power of 0 dB.
Table 6.9.2-4. NTN-TDL-D at elevation

0 -0.284 LOS path
1
0 -11.991 Rayleigh
2 0.5596 -9.887 Rayleigh
3 7.3340 -16.771 Rayleigh
NOTE: The first tap follows a Ricean distribution with a K-factor of K1 = 11.707 dB and a mean power of 0 dB.
3GPP
Release 15 93 3GPP TR 38.811 V15.1.0 (2019-06)
6.10 Channel model calibration

6.10.1 NTN channel model features per deployment scenarios
Table 6.10.1-1: NTN channel model features per deployment scenarios
Deployment-D1 Deployment-D2 Deployment-D3 Deployment-D4 Deployment-D5

Platform orbit and GEO at 35 786 km GEO at 35 786 km Non-GEO down to Non-GEO down to HAPS between 8 km
altitude 600 km 600 km and 50 km
Carrier Frequency Around 20 GHz for Around 2 GHz for Around 2 GHz for Around 20 GHz for Below 6 GHz
on the link DL both DL and UL (S both DL and UL (S DL
between Air / Around 30 GHz for band) band) Around 30 GHz for
space-borne UL (Ka band) UL (Ka band)
platform and UE
Maximum Channel Up to 2 * 800 MHz Up to 2 * 20 MHz Up to 2 * 20 MHz Up to 2 * 800 MHz Up to 2 * 80 MHz
Bandwidth
(DL + UL)
UE antenna VSAT type - circular Quasi Isotropic - Quasi Isotropic - VSAT type - circularQuasi Isotropic -
pattern + polarisation Linear polarisation Linear polarisation polarisation Linear polarisation
polarisation Co-phased array - (Note 4) (Note 4) Co-phased array - (Note 4)
Dual Linear Co-phased array - Co-phased array - Dual Linear Co-phased array -
polarisation (Note 1) Dual Linear Dual Linear polarisation (Note 1)
Dual Linear
polarisation (Note 2) polarisation (Note 2) polarisation (Note 2)
UE type Handheld, nomadic, Handheld, moving Handheld, moving Handheld, nomadic, Handheld, moving
fixed, moving platform mounted platform mounted fixed, moving platform mounted
platform mounted platform mounted
Airborne & space Bessel function and Bessel function and Bessel function and Bessel function and - Bessel function
borne antenna circular polarisation circular polarisation circular polarisation circular polarisation and circular
pattern modelling polarisation
+ polarisation - 3GPP antenna
pattern of Base
Station (Dual
Linear
polarisation)
Doppler cause Mainly UE mobility Mainly UE mobility UE + satellite UE + satellite UE + HAPS mobility
mobility mobility
O2I penetration No No No No Possible
loss
Atmospheric Mandatory Negligible Negligible Mandatory Negligible
absorption
Rain attenuation (Note 3) Negligible Negligible (Note 3) Negligible
Cloud attenuation (Note 3) Negligible Negligible (Note 3) Negligible
Scintillation Tropospheric Ionospheric Ionospheric Tropospheric Negligible
Fast fading Flat fading (Note 6) Flat fading (Note 6) Flat fading (Note 6) Flat fading (Note 6) Frequency selective
models (system or frequency or frequency fading (note 5)
level) selective fading selective fading according to
(note 5) according to (note 5) according to elevation and
elevation and elevation and environments
environments environments
Link level model Flat fading (Note 6) CDL or TDL CDL or TDL Flat fading (Note 6) CDL or TDL
Shadowing model LMS (Land Mobile LMS LMS LMS 3GPP TR38.901
Satellite) based
Note 1: As described in [48] as [M,N,P] = [2,4,2].
Note 2: As described in [48] as [M,N,P] = [1,2,2].
Note 3: Rain and cloud attenuation are not needed for system or link level simulations related to channel model, if
they are already considered in the system dimensioning (e.g. link budget). If they need to be taken into
account, ITU-R P618 models (Rain) and ITU-R P840 models (Cloud) shall be used.
Note 4: Quasi isotropic refers to dipole antenna which is omni-directional in one plane.
3GPP
Release 15 94 3GPP TR 38.811 V15.1.0 (2019-06)
Note 5: The frequency selective fading refers to Geometry based Stochastic Channel Model or GSCM which is
defined in (SCM/FP7 WINNER, 3GPP TR 38.901 etc.)
Note 6: Flat fading model refers to the 2 state model from ITU-R P681 (section 6). Since this model is based on
time series, R1-1802975 proposes a method to adapt it to drop based simulations for system level
evaluation.
3GPP
Release 15 95 3GPP TR 38.811 V15.1.0 (2019-06)
7 Potential key impact areas on NR to support NTN
7.1 Specific constraints associated to NTN

This clause describes selected specific Non Terrestrial Network design constraints that need to be addressed when
considering the Non-Terrestrial Network deployment scenarios.
Table 7.1-1: Specific design constraints of Non Terrestrial Network
Non Terrestrial Differences between NTN and cellular systems

Network Design
Constraints
Propagation channel As detailed in clause 6, the main differences lies in different multi path delay and Doppler
spectrum model .
For narrowband signals and frequency bands below 6 GHz, the time dispersion may be
ignored.
We shall assume outdoor conditions and line-of-sight operations for the UE for communication
via satellite. In HAPS system, indoor conditions are also addressed, which implies the need to
consider non-line-of-sight conditions as well.
Frequency Plan and The allocated spectrum to a satellite system is respectively 2 x 15 MHz (UL & DL) at S band
channel Bandwidth and about 2 x 2500 MHz (UL & DL) at Ka band. In addition, satellite systems at S and Ka
bands use mostly circular polarizations.
Frequency re-use and flexibility of spectrum allocation in the different cells may be supported
and therefore the maximum channel bandwidth per cell is assumed to be respectively 2 x 15
MHz (UL & DL) at S band and 2 x 800 MHz (UL & DL) at Ka band.
For efficient spectrum usage, the system should always minimise the risk of inter cell
interference.
Power limited link The main design drivers of satellite and HAPS based communication systems are;
budget - To maximise the throughput for a given transmit power from the UE on the UL and from
the satellite/HAPS on the DL.
- To maximise the availability of the service under deep fading situations (typically between
20 and 30 dB in Ka band for 99.95% availability)
Cell pattern generation Space and HAPS systems typically feature larger cells compared to cellular networks. In
addition, the cells may be moving (without a fixed earth reference point) in case of NGSO
(Non Geostationary Satellite Orbit) satellite or HAPS system.
These large cells especially at low operational elevation angles will create a significant
differential propagation delay between a UE at cell centre and UE at cell edge and the ratio of
the differential increases as the altitude of the satellite and HAPS decreases. In other words,
the ratio between propagation delays at cell centre and cell edge is likely to be higher in the
context of HAPS as compared to geostationary satellite systems.
This will impact contention based access channels when the position of UEs is not known by
the network.
Propagation Delay Satellite systems feature much larger propagation delays than terrestrial systems. The one-
characteristics way delay between the UE and the RAN (whether on-board the satellite/HAPS or on the
ground) may reach up to 272.4 ms for GSO (Geostationary Synchronous Orbit) systems, and
is greater than 14.2 ms for NGSO (Non-Geosynchronous Orbit) systems. In the case of
HAPS, the one way delay is less than 1.6 ms and hence comparable with cellular networks.
This larger delay will likely impact all signalling loops especially at access and transport (data
transfer) levels.
The analysis of the propagation delay is detailed in clause 5.3.
Mobility of the In cellular networks, transmission equipment refers to base stations (gNB) or Remote Radio
infrastructure's Heads (RRH). They are usually fixed, except when on board a moving platform such as a
transmission equipment train.
In Non-Terestrial Networks, the transmission equipment is on board the space/HAPS. For
GSO systems, the transmission equipment is quasi static with respect to the UE with only
small Doppler effects. For HAPS, the transmission equipment is moving around or across a
theoretical central point and hence creates Doppler. For NGSO systems, the satellites move
relative to the earth and creates higher Doppler effects than for GSO systems.
The Doppler depends on the relative satellite/HAPS velocity with respect to the UE, and on
the frequency band. Doppler can be characterized by a maximum Doppler Shift and a Doppler
variation rate.
This effect will continuously modify the carrier frequency, phase and spacing and may create
Inter Carrier Interference (ICI).
As detailed in clause 5.3, assuming a worst case UE velocity of 1000 km/h:
3GPP
Release 15 96 3GPP TR 38.811 V15.1.0 (2019-06)
Non Terrestrial Differences between NTN and cellular systems

Network Design
Constraints
- For LEO in S band (2 GHz): +/- 48kHz Doppler Shift and - 544 Hz/s Doppler variation rate
- For LEO in Ka band (20 GHz): +/- 480kHz Doppler Shift and – 5.44 kHz/s Doppler variation
rate
- For LEO in Ka band (30 GHz): +/- 720kHz Doppler Shift and – 8.16 kHz/s Doppler variation
rate
Note, although these values are high, most of the Doppler shift and Doppler variation rate can
be pre/post compensated, exploiting the knowledge of the satellite/HAPS motion that can be
anticipated through modelling (e.g. ephemeris of satellites) as well as UE location if known.
Service continuity Whenever a UE leaves or enters cellular coverage, a hand-over to or from the satellite/HAPS
between land-based 5G system can take place to ensure service continuity. The handover triggering mechanisms
access and non- might be different for each direction e.g. leave satellite as soon as there is enough cellular
terrestrial based access signal, but only leave cellular when there is a very low cellular signal.
networks The hand-over procedure should take into account the service enablers, the characteristics
and the measurement reports of both access technologies. The following aspects should be
considered:
- Support both non-transparent air/spaceborne (on-board processing) and bent-pipe
architectures
- Handover preparation and HO failure/RLF handling
- Time synchronization
- Measurement object coordination – incl. gap allocation & alignment
- Lossless handover support
- Specifics related to intra-Non Terrestrial network mobility, as well as between Non-
Terrestrial and Cellular networks
Radio resource In order to support varying traffic demand while also taking UE mobility requirements into
management adapted account, requires minimal response time of the access control functionality.
to network topology In cellular systems, access control is typically located in the gNB close to the UE. Moreover
coordination between gNBs is possible through the Xn interface (between gNBs interface) or
via a central entity.
In satellite systems, access control is mostly located at satellite base station, gateway or hub
level which may prevent optimal response time for access control. Hence, pre-grants, Semi
Persistent Scheduling (SPS) and/or grant free access scheme would be beneficial.
Terminal mobility The challenge is to support very high speed UEs such as aircraft systems featuring maximum
speeds of 1000 km/h (see 3GPP TS 22.261 [6]).
3GPP
Release 15 97 3GPP TR 38.811 V15.1.0 (2019-06)
7.2 NR features/protocols potentially affected

This clause identifies the NR Features/protocols in bold that may require some adaptations to support operation via
Satellite or HAPS. Other solutions may be identified but don't require adaptations of the standard.
Table 7.2-1: NR Features/protocols that may require some adaptations to support operations via
Satellite or HAPS
Non-Terrestrial Impacted NR NR Features that may require adaptations to support NR operation via
Network design areas Satellite or HAPS
constraints
Propagation channel Physical Layer For satellite-based systems the signal is mostly direct LOS and follows a
Ricean distibution with strong direct signal component; slow fading is
possible, due to temporary signal masking e.g. under trees and bridges.
For HAPS based systems, the signal contains significant multipath
components and follows a Ricean model. Similarly to cellular systems, a
frequent and fast fading of max 100 ms coherence time is expected, mainly
due to signal components recombination.
To improve performance, the receivers' synchronisation configuration at both
UE and gNB level, could be adapted especially:
- The Reference signals in the physical signals (e.g. DL: PSS, SSS,
Reference Signals; UL: DMRS, SRS), the Preamble sequence and
aggregation (related to random access channel), to take into account the
Doppler and possibly some specific multi path channel model.
- The Cyclic Prefix to compensate for the delay spread and the
jitter/phase. The sub-carrier spacing (SCS) of the OFDM signal may be
extended with greater SCS values to accommodate larger Doppler (To Be
Confirmed).
Frequency plan and Physical Layer The carrier numbering could be reviewed to support the targeted spectrum
channel Bandwidth (S band, Ka band) and the pairing between UL/DL bands with specific
band separation.
Specifically, for Ka band NTN deployment scenarios, the 5G radio interface
mode foreseen for above 6 GHz bands shall be configured to support FDD
access scheme (e.g. with two mono directional carriers operated in opposite
direction on both UL and DL bands). This will require the system to configure
and possibly adapt the MAC and network layer signalling in a specific
manner.
The carrier bandwidth could be extended up to a maximum of 800 MHz.
Alternatively carrier aggregation method can be used to provide equivalent
throughput while enabling a greater flexibility of carrier allocation between the
cells while respecting frequency reuse constraints.
Power limited link Physical Layer To maximise the throughput / power ratio, the operation point in the power
budget amplifier at satellite or at the UE shall be set as close as possible to the
saturation point, when needed. To support this, several techniques can be
considered and possibly combined together:
- Extended multicarrier modulation and coding schemes especially for
the UL that features low Peak to Average Power Ratio (PAPR) that are more
robust against distortions
- PAPR reduction and nonlinear distortion mitigation through signal
processing techniques (e.g. pre-distortion mechanisms).
- Operating the high power amplifier (UE and satellite/HAPS level) with the
minimum output back-off if necessary.
To maximise the signal availability with slow and deep fading, especially for
UE at cell edge, it is recommended to provide modulation and coding
schemes featuring very low SNR operating points or other alternatives
especially for mMTC service enablers. This may lead to extend the
Modulation Coding Scheme of NR towards very low Es/No to meet the
reliability requirements of critical communications or low energy consumption
scenarios.
Power limited link MAC layer To maximise the spectral efficiency and accommodate limited power
budget (Resource terminals, the MAC layer should be able to allocate Physical Resource
Allocation) Blocks in the most flexible way. Reduced size of Physical Resource
Blocks should be considered (e.g. single tone transmission or transmission
over one OFDM sub carrier of same or larger bandwidth than a single OFDM
carrier).
3GPP
Release 15 98 3GPP TR 38.811 V15.1.0 (2019-06)
constraints
Cell pattern Physical layer The differential delay due to large cell size may create near far effect
generation between UEs during the initial access procedure when the UEs positions
are not known. This may require an extended acquisition window to improve
performance.
Note however that if the UE position becomes known during a session, the
differential delay can be compensated by the network.
For broadcast service, specific signalling may be needed to accommodate
the larger and moving cells.
Propagation Delay Physical layer User traffic, such as voice or video conference services, is latency and jitter
characteristics sensitive.
In cellular networks, HARQ retransmission may lead to jitter (e.g. typically up
to 8ms in the case of LTE FDD mode). A mitigation scheme called TTI
Bundling on UL is defined. It allows to re-transmit the same symbols over up
to 4 consecutive sub frames, without waiting for HI acknowledgment. It is
optimised for short jitter.
In Satellite/HAPS systems, as the propagation delays are significantly longer,
HARQ scheme would create unacceptable jitter. The UL slot aggregation
may need to be adapted to compensate for higher jitter. Possible solutions
could be to increase the periodicity, the number of re-transmissions of
symbols and/or decrease the slot duration.
Further enhancements of HARQ process should not be precluded to
compensate for the propagation delay characteristics.
Propagation Delay Access layer In satellite/HAPS systems, the longer propagation delay impacts various
characteristics (MAC, RLC) protocol layers, retransmission mechanisms and response times in resource
scheduling.
The following mechanisms should be adapted to accommodate longer delays
and provide latency compensation/reduction especially on delay sensitive
applications. We recommend to minimise the number of exchanges between
the UE and the network via the following features
- an Initial Access procedure (based on a random access scheme) by
implementing novel methods, where data and access signalling are sent
together, would help to meet this requirement (e.g. grant free access, 2 step
RACH)
- a Data Transfer procedure, by implementing flexible/extended receive
window size, flexible acknowledgement policy in terms of frequency, event,
flexible and ARQ/HARQ cross coordination, radio resource allocation, ACK
free scheme or a latency adaptive HARQ-ACK feedback, to accommodate
long delay channel
Propagation Delay Physical Layer In terrestrial cellular systems, the gNB selects the most appropriate MCS
characteristics (Modulation and Coding Scheme) based on the CQI (Channel Quality
Indicator) reported by the UE as part of the AMC (Adaptive Modulation and
Coding) procedure.
In satellite systems, the propagation delay creates a larger response time for
the AMC loop and hence requires a margin to compensate for the possible
outdated CQI. This leads to a suboptimal use of the useful transmission
capacity (lower spectral efficiency).
In order to improve its efficiency, the AMC procedure could be modified with
potential signalling extension.
Node B or RRH Physical Layer In cellular networks based on an OFDM radio interface, the SCS (Sub Carrier
mobility Spacing) may be scaled in order to mitigate both the Doppler shift and
Doppler variation rate.
In satellite / HAPS systems, the SCS range of values may need to be
extended especially for Ka band and large channel bandwidth (e.g. 800
MHz).
Service Continuity Physical Layer, To support service continuity between the cellular and non-terrestrial
between land based MAC layer networks or within non-terrestrial network, it is recommended to adopt a hard
5G access and (Resource hand-over scheme or dual/multi-connectivity between terrestrial and non-
satellite based Allocation) and terrestrial networks, when possible.
access networks above (Mobility The differences in propagation delay between cellular and non-terrestrial
Management) access network will create a significant jitter or possible data starvation.
- For delay sensitive applications, a temporary QoS degradation may be
accommodated provided that it doesn't occur too often.
- For data services with high reliability requirements, buffering or
retransmission techniques (e.g. PDCP inherent capability) may be needed to
compensate for such impairment.
3GPP
Release 15 99 3GPP TR 38.811 V15.1.0 (2019-06)
constraints
- Delays may be compensated before starting the hand-over
The support of hand-over between cellular and non-terrestrial network will
require:
- Extension of PDCP retransmission scheme in terms of repetition rate
of data and duplication handling in PDCP layer
- Extension of hand-over signalling at RRC, RLC, MAC layer
Radio resource NAS, RRC, The mobility management should accommodate the specific cell patterns
management RLC, MAC, (size and position) of NTN networks. Moreover, cells in the non-terrestrial
adapted to network Physical Layer network may cross borders between countries.
topology In particular, this will affect the identification method of cells, the design
of tracking and location areas, the roaming and billing procedures as
well as location based services.
Furthermore, NGSO and HAPS systems generate cells which move around.
This causes hand-over to occur not only for mobile units but also for fixed
UEs. However, the hand-over caused by such NTN may be anticipated, by
exploiting trajectory models such as satellite ephemeris.
Radio resource RAN In cellular networks, the access controller, which control radio resources
management Architecture allocation, is implemented in the gNB. The access controller may control the
adapted to network interface between the nominal gNB and the UE or the interface between a
topology neighbouring gNB and the attached UE (in the latter case, the neighbouring
gNB acts as a relay node).
In Non-Terrestrial Networks, the access controller function is typically
implemented
- on board the HAPS
- at the satellite gateway/gateway level or on board the satellite for GSO
and NGSO systems
In case of Ka band deployment scenarios, the NTN terminal may
accommodate gNB functions as part of a Relay Node.
The NTN specific topology may require some adaptations to the inter gNB
protocols to cope with the propagation delay, the cell pattern and the cell
mobility (NGSO and HAPS systems)
Terminal mobility Physical layer For NTN terminals moving at 1000 Km/h the response time of the power
control loop should be decreased.
- Physical Frame & Sub-Frame structures:
- The reduction of transmission slot duration could be considered
(e.g. mini slot approach) to decrease the Power Control Loop response time
- The extension of SCS values could be considered to support very
high speed mobility UEs
- Physical signals:
- The mapping and scheduling of the power control command on
physical radio resources may be revisited to enable a faster response time.
3GPP
Release 15 100 3GPP TR 38.811 V15.1.0 (2019-06)
7.3 NR modifications to support the Non-Terrestrial Network

deployment scenarios
For each deployment scenarios, the actual impacts on NR are identified.
7.3.1 Methodology
For each of the NR Features that may require adaptations to support NR operation via Satellite or HAPS, identified in
chapter 7.2, the problem statement in terms of issue are assessed, potential areas of impact on NR protocol are identified.
Table 7.3.1-1: Areas of impacts on NR to support Non-Terrestrial networks
Non-Terrestrial network specifics Effects Impacted NR features

Moving cell pattern Hand-over/paging
Motion of the space/aerial vehicles (especially for Non Delay variation TA adjustment
GEO based access network) Init synchro downlink
Doppler
DMRS time density
HARQ
MAC/RLC Procedures
Altitude Long latency
Physical layer Procedures (ACM,
power control)
TA in Random access response
Cell size Differential delay message
RACH
DMRS frequency density
Propagation channel Impairments
Cyclic prefix
Regulatory Access scheme (TDD/FDD)
Duplex mode
constraints
Phase noise PT-RS
Satellite or aerial Payload performance impairment
Back-off PAPR
Network architecture RAN Mapping Protocols
7.3.2 Motion of the space/aerial vehicles
7.3.2.1 Hand-Over and paging
7.3.2.1.1 Problem statement

NGSO satellites move rapidly with respect to any given UE location. As an example, on a 2-hour orbit, a LEO satellite
is in view of a stationary UE from horizon to horizon for about 20 minutes. Since each LEO satellite may have many
beams, the time such a UE stays within a beam is typically for only a few minutes. The fast pace of change creates
problems for paging as well as handoffs for a stationary UE as well as a moving UE.
Since handover happens in general when the UE or relay is in CM-ACTIVE and RRC-CONNECTED state, the
procedure is time critical to avoid loss data. In NTN systems based on NGSO satellites, the cells or spot beams are
moving at high speeds and so the handover procedure from one spot beam to the next or from one satellite to the next
has to be executed quickly otherwise the UE may not make use the target beam and/or satellite resources efficiently and
in the worst case may suffer loss of data.
NR beam management for mobility between spot-beams on the same base station cannot be ported to satellite to
minimize the handoff overhead. The NR beam management might assume same frequency on the adjacent beams, but
for, the adjacent beams on the same satellite may use different frequencies or different polarization. Thus, the beam
management procedures may have to be modified.
The problem for paging needs more detailed explanation. In NR based cellular networks, a UE camps on a cell. The cell
is uniquely identified by the RAN from which the UE is receiving the radio signals from. A collection of cells is called
a Tracking Area. A collection of Tracking Areas is called as a Registration Area. A cell belongs to a Tracking area and
3GPP
Release 15 101 3GPP TR 38.811 V15.1.0 (2019-06)
a Registration Area. As long as the UE stays within a Registration Area, no location update is needed. The UE in the
CM-IDLE state will perform a Registration Area update when it moves out of a Tracking Area.
The AMF only needs be aware of the UE location to the granularity of Registration Area when a UE is in the CM-IDLE
state. If a packet arrives from internet for this UE in CM-IDLE state, the AMF attempts to page the UE on all cells
belonging to the Registration Area in order to notify the arrival of packets to it. All RANs that receive the page
transmits a page in the corresponding cells to reach UE that may be anywhere in the Registration Area.
In Non-GEO satellite access network, a UE camps on a beam of a satellite, but as beams move, it ends up camping on
different beams and different satellites over time even though UE may not have moved. Unlike terrestrial framework
where a cell on the ground is tied to radio communication with a RAN, in Non-GEO satellite access network, the
satellite beams are moving. There is no correspondence between cells on the ground and satellite beams. The same cell
on the ground is covered by different satellites and different beams over time. Therefore, for the initial Registration, the
Satellite based radio access network will not be able to provide the Tracking Area information to AMF based on which
beam and which satellite the Registration Request was received. Given that tracking areas are defined on the ground
and Non-GEO beams are moving, there is no one-to-one correspondence between moving beams and fixed tracking
areas or registration areas on the ground. However, this information is necessary for UE to determine if it needs to
perform a registration area update with AMF in NR.
7.3.2.1.2 Assessment of conditions for NR operation in Non-Terrestrial networks

For handover and paging to operate successfully and efficiently in the NGSO satellite networks, the NR UEs may need
to be capable of geo-location. One possibility is that the NR UE may be aware of their locations and may report this to
the satellite RAN if needed. (For fixed installations, their location can be reported once at time of installation.)
The ephemeris information of the NGSO satellites can be used to determine their footprints of each of the beams, and
its velocity all the time. Therefore, for a given UE location at any given time, the network has information as to which
beam of which satellite covers that location best. It also knows the duration that UE location would remain to be
covered by the beam and which beam on the same satellite, or a different satellite will be the best candidate to switch
over next, and at what time. If positioning information of the UEs is available at the network, the possibility to simplify
handover procedure and reduce measurement reporting overhead require further study.
7.3.2.1.3 NR impact considerations

No impact for HAPS or GEO satellite based NTN. For Non-GEO satellite based NTN, adaptation of the NR handover
and paging protocols needs further study possibly taking advantage of the knowledge of the UE location and satellite
ephemeris information.
7.3.2.2 TA adjustment

MEO, LEO and HAPS systems feature a strong varying delay because satellite/HAPS and UE are fast-moving and are
not relatively static. In this case, the individual timing advances of the UEs may need to be dynamically updated and
appropriate TA index values may be needed to solve the long strong delay in the overall distance of the propagation on
NTN link.
The issues or technical problems to solve, related to TA alignment in Satellite communications, are as follows:
1. A strong delay variation is caused by moving satellites generating a fast change in the overall distance of the
propagation from UE over Satellite to BS.
2. The delay is much longer over a satellite link than one TTI.
A strong delay variation is caused by moving satellites generating a fast change in the overall distance of the radio link
between UE and BS via satellite. The delay is much higher and variable over a satellite radio link than over a terrestrial
radio link. This delay largely exceeds the TTI (Equivalent to one frame) of NR which is equal to or less than 1 ms.
However, the delay variation is quite predictable knowing the satellite orbits and UE position.
Hence, TA alignment is an important feature of NR that will be impacted by introduction of NTN in 5G to ensure that
all uplink transmissions are synchronized at gNB reception point.
3GPP
Release 15 102 3GPP TR 38.811 V15.1.0 (2019-06)

A timing advance command [20], TA , for a TAG indicates adjustment of the current N TA value, N TA_old , to the new
N TA value, N TA_new , by index values of TA = 0, 1, 2,..., 63, where for a subcarrier spacing of 2  15 kHz,
N TA_new  N TA_old  TA  31  16  64 2  .
As shown in Figure 7.3.2.2.2-1, transmission of uplink frame number from the UE shall start
before the start of the corresponding downlink frame at the UE, where can be
derived by the UE based on the index value from gNB, depends on the duplex mode and frequency range
in uplink transmission, and is the basic timing unit [21][22][23].
Figure 7.3.2.2.2-1: Time alignment at gNB with TA
In case of satellite, the high delay drift of individual non-GEO satellites is quite predictable because the motion of the
satellites follows known paths. The very fast update of the TA is neither required in terrestrial links, nor in GEO
satellite links. In both scenarios, the terminal mobility is dominating the TA requirements.
Another technical issue that arises is that the delay variation over the satellite link is much more than a TTI. E.g. if the
SCS is increasing from 15 kHz to 60 kHz, the TTI goes down from 1 ms to 250 μs. The required TA adjustment range
for satellite links will become larger that the TTI with any SCS selection and the transmission timing of the UE has to
be adjusted over the borders of individual TTIs.
Table 7.3.2.2.2-1: TA granularity, and step size with SCS
Subcarrier spacing (SCS) SCS RB band- TA gran- Tstep Max Ta step

configuration parameter, µ [kHz] width [kHz] ularity [Ts] [ns] size [µs]
0 15 180 1024 520.83 16,6
1 30 360 512 260.42 8,3
2 60 720 256 130.21 4,15
3 120 1440 128 65.10 2,1
4 240 2880 64 32.55 1
This max Ta step should be valid only for extended cyclic prefix. With normal cyclic prefix, we expect this maximum
step not to exceed the normal CP length (e.g. CP length is 4.7 µs in 15kHz SCS case). It means that the number of Ta
commands to be sent per second to track the maximum drift of 35 µs/s is about 10 per second for 15 kHz SCS case, 40
per second for 60 kHz SCS, and 80 per second for 120 kHz SCS case. The number of Ta commands to be sent per
second is important but can be implemented.
3GPP
Release 15 103 3GPP TR 38.811 V15.1.0 (2019-06)

Solutions need to be studied to ensure alignment of uplink signals over the NTN links to overcome the predictable delay
of NTN.
7.3.2.3 Initial synchronization in downlink

In order to access the 5G network, the UE has to detect the PSS and the SSS. Those synchronization signals allow time
and frequency correction, and Cell Id detection. UE in cellular network has to get good one-shot detection probability at
-6 dB received baseband SNR condition with less than 1% false alarm rate as defined in chapter 7.1.5 of [24], with
robustness against initial frequency offset up to 5 ppm.
We expect that requirements (SNR level of -6 dB, frequency error robustness of 5 ppm and 1% false alarm) defined for
terrestrial UE will be kept the same for NTN UE. Even if the SNR level of NTN systems is typically in the range of -3
to 13 dB SNR.
In cellular networks, transmission equipment (gNB or RRH) are usually fixed, except when on board a moving platform
such as a train.
In Non-Terrestrial Networks, the transmission equipment is on board the satellite/HAPS.
- For geostationary systems, the transmission equipment is quasi static with respect to the UE with only small
Doppler shift.
- For HAPS, the transmission equipment is moving around or across a theoretical central point but creates small
Doppler shift.
- For non-geostationary systems, the satellites move relative to the earth and creates higher Doppler shift than for
geostationary systems.
The Doppler shift depends on the relative satellite/HAPS velocity with respect to the UE, and on the frequency band.
In term of Doppler shift, the worst case for NTN systems corresponds to non-geostationary systems, at lowest altitude
(i.e. 600 km), where the speed of the satellite embedding transmission equipment is 7.5 km/s.
As detailed in clause 5.3, assuming a worst case NTN terminal velocity of 1000 km/h:
- For LEO in S band (2 GHz): Up to +/- 48kHz Doppler Shift in downlink for the whole satellite coverage (spot)
- For LEO in Ka band (20 GHz): Up to +/- 480kHz Doppler Shift in downlink for the whole satellite coverage
(spot)
If frequency error robustness requirement is 5 ppm (i.e. 10kHz for S band, 100 kHz for Ka band), it means that this
worst case described above is not covered by current 5G specifications. To be inferior or equal to 5ppm error, the
satellite altitude has to be above 13000 km.
Actually, the Doppler shift amplitude to be compensated is less than 48 kHz for S band and 480 kHz for Ka band,
because in satellite communication systems, all satellites whether GSO or NGSO generate multi beams and each beam
foot print corresponds to a cell. The Max Doppler shift amplitude observed at the satellite coverage foot print edges will
be reduced within each beam footprint. The lower the beam width, the less the Doppler shift amplitude.

In case the altitude of the satellite is above a certain altitude (about 13 000 km) or if center beam pre-compensation is
sufficient to reach the 5 ppm requirements, no impact is foreseen.

In case the above conditions are not met, further studies are required to accommodate the high Doppler shift during the
cell synchronization procedure in non-Terrestrial networks.
3GPP
Release 15 104 3GPP TR 38.811 V15.1.0 (2019-06)
Note that this high Doppler shift can be known based on UE location and satellite ephemeris and hence could be for
example pre compensated. This could prevent impact on the NR but may need to be confirmed with further study.
7.3.2.4 DMRS time density

As analyzed in Section 5.3, the maximum Doppler variation rates during the field tests appeared in the scenarios of
LEO at the altitude of 600 km, and are -544 Hz/s with carrier frequency of 2 GHz, -5.44 kHz/s with carrier frequency of
20Ghz and -8.16 kHz/s with carrier frequency of 30 GHz, where the maximum Doppler variation rate is the maximum
slope (maximum derivative) of the Doppler shift curves. In NR network, there are up to 4 DMRS symbols per slot and
the duration of one slot can be configured to 1 ms, 0.5 ms, 0.25 ms, 0.125 ms or 0.0625 ms, which correspond to the
subcarrier spacing of 15 kHz, 30 kHz, 60 kHz, 120 kHz and 240 kHz respectively, as shown in Figure 7.3.2.4.1-1. Table
7.3.2.4.1-1 summarizes the potential maximum Doppler shift in one slot in NTN based on the observed maximum
Doppler variation rate during the field tests. As shown in Table 7.3.2.4.1-1, it is observed that the potential maximum
Doppler shifts are up to 0.544 Hz with carrier frequency of 2 GHz in DL/UL, up to 5.44 Hz with carrier frequency of 20
GHz in DL and up to 8.16 Hz with carrier frequency of 30 GHz in UL.
Figure 7.3.2.4.1-1: Slot duration depending on µ configuration
3GPP
Release 15 105 3GPP TR 38.811 V15.1.0 (2019-06)
Table 7.3.2.4.1-1 Potential maximum Doppler Shift in one slot in NTN
Carrier frequency Potential maximum

Maximum Doppler Duration of one slot
when the altitude of Doppler shift in one
variation (Hz/s) (ms)
LEO satellite is 600km slot (Hz)
1 0.544
0.5 0.272
2 GHz
- 544 0.25 0.136
(DL/UL)
0.125 0.068
0.0625 0.034
1 5.44
0.5 2.72
20 GHz
-5440 0.25 1.36
(DL)
0.125 0.68
0.0625 0.34
1 8.16
0.5 4.08
30 GHz
-8160 0.25 2.04
(UL)
0.125 1.02
0.0625 0.51
In the existing LTE network, the frequency error at the UE side shall be within ±0.1 PPM observed over a period of 0.5
ms [25]. In NR network, the frequency error at the UE side shall be within ±0.1 PPM observed over a period of 1
ms[26].

The frequency error corresponding to ±0.1 PPM is ±200 Hz with carrier frequency of 2 GHz, ±2 kHz with carrier
frequency of 20 GHz and ±3 kHz with carrier frequency of 30 GHz. Compared to the requirement of frequency error at
the UE side, the maximum Doppler shift in one slot in LEO scenarios during field test is negligible, see Table 7.3.2.4.1-
1. This result is also valid for other configurations with higher DMRS time density per time slot as configurable in 5G
NR. Similar conclusion could be observed at the eNB side too [27][40].

As a conclusion, no impact on NR specifications is needed for DMRS positioning in time in NTN deployment scenarios
because the maximum Doppler Variation in NTN is negligible compared to the minimum requirement of frequency
error at both gNB and UE sides in NR.
7.3.3 Altitude of the space/aerial vehicles
7.3.3.1 HARQ
The HARQ process is a very time-critical mechanism. HARQ operation becomes even more critical at extremely long
RTT, i.e., as in case of NTN and extreme coverage scenarios.
In satellite communication, the RTT normally exceeds the maximum conventional HARQ timers (after which an ACK
is received) or the maximum possible number of HARQ processes (i.e., a flexible pool of parallel HARQ processes
similar to LTE). This is also true even for LEO constellations, where the RTT varies between 15 to 63 times longer than
that of the terrestrial RTT [28]. Thereby, simply extending the number of HARQ processes linearly to RTT induced by
the satellite channel might not be feasible for some UEs due to memory restrictions and the maximum possible parallel
processing channels [41][42][43][44][45]. Furthermore, gNBs also have to consider this latency impact on the number
of their active HARQ processes. Therefore, it is important to study the impact on NR HARQ operation for the
introduced NTN delays. Figure 7.3.3.1-1 depicts the HARQ feedback RTT, i.e., from data retransmission time up to
acknowledge reception at a bent-pipe satellite. The figure also shows the backhauling delay as well as the
forward/reverse satellite link delay , i.e., where the RTT = .
3GPP
Release 15 106 3GPP TR 38.811 V15.1.0 (2019-06)
Figure 7.3.3.1-1: Bent-pipe Satellite HARQ feedback operation with a maximum RTT (from sending a
NACK until receiving a retransmission redundancy version (RV))

NR has extended the number of HARQ processes in Rel. 15 to 16 processes [29]. For NR NTN satellite transmission,
the number of HARQ processes may need to be further extended flexibly according to the induced RTT delay. Here, the
minimum required number of HARQ processes can be computed directly from the RTT delay of each satellite
constellation, e.g., LEO, MEO and GEO, using the following formula [28]:
(7.3.3.1-1)
where is the minimum required number of HARQ processes, is 1ms assuming a reference numerology
15 kHz subcarrier spacing, and is the time duration between the initial transmission of one transport block (TB)
and the corresponding ACK/NACK complete decoding.
The is depicted in Figure 7.3.3.1.1-1 considering the RTT ( ), transmission time ( ), and processing
time, and , for decoding the TB and the ACK/NACK frame, respectively. is illustrated in details in
Figure 7.3.3.1.1-1.
Figure 7.3.3.1.1-1: Timing diagram of a single HARQ process for a NTN with a single bent-pipe
satellite in the link
Table 7.3.3.1.1-1 gives an overview of the number of HARQ processes, , based on different values
(including the RTT) for different satellite constellations, LEO, MEO, and GEO [5].
3GPP
Release 15 107 3GPP TR 38.811 V15.1.0 (2019-06)
Table 7.3.3.1.1-1: The minimum required number of the HARQ processes, , assuming a 1ms
slot duration for 15 kHz* reference subcarrier-spacing
constellation Max. processes for 1 UE side feasibility

ms slot operation
Terrestrial 16ms 16 Feasible (Rel. 15)
LEO 50ms 50 Feasible (with HARQ extension)
MEO 180ms 180 FFS (impact on TBS/MCS)
GEO/HEO 600ms 600 FFS (impact on TBS/MCS)
NOTE*: For larger subcarrier spacing (SCS), i.e., 2 k * 15 kHz, the min. required number of the HARQ processes
might be scaled by 2k.
7.3.3.1.2 Assessment of conditions for NR operation in non-terrestrial networks

Handling a higher number of parallel HARQ processes and its feasibility for NTN have been addressed in [46] [41] [32].
The impacts of NTN delays on the existing HARQ-supporting mechanisms need to be addressed in a subsequent study.
At least the following principles can be considered in further study:
- Enhancing existing HARQ operation to extend the HARQ processing accommodating low to moderate NTN
RTT delays.
- Limiting HARQ capabilities and/or disabling HARQ for long RTT delays.

It is required to study the impact on NR HARQ operation due to the long RTT delay of a non-terrestrial network. The
impacts should be considered as well for the NTN UEs and serving gNBs, when the number of HARQ processes is
either extended to satisfy high reliability scenarios or limited/disabled for longer NTN delays.
7.3.3.2 MAC/RLC procedures

UE operating in GEO satellite access networks can experience a one-way propagation time of 240 ms at the minimum,
270 ms at the maximum between UE and satellite base station. The base station can observe the acknowledgement of
packets sent to the UE only after the round trip plus some processing time, which is more than ½ second later. Similarly
the UE can observe acknowledgement for its packet sent to the base station in about the same time interval. For the
LEO satellite systems with typical 600 km orbit, the one way propagation delay changes continuously between 2 ms
when the satellite is directly above, and 7 ms when the satellite is seen with 10° elevation.
ARQ requires that the transmitted packets be buffered in anticipation of potential packet loss and released only after the
successful receipt of an acknowledgement, or until a time-out mechanism reinitiating a retransmission. The long round
trip delay requires larger transmission buffer, and potentially limits the number of retransmission allowed for each
transmitted packet in both the forward and return links. Note that in LEO satellite systems, the ARQ transmit buffer size,
and retransmission mechanism must be designed for the longest possible delay, i.e. at the lowest elevation.
Scheduling mechanisms must be able to cope with the long RTT. UL scheduling delay parameters are expected to be
redefined to accommodate the RTT of the associated deployment scenario.

For efficient ARQ operation in GEO or NGSO satellite networks, NR UE and base stations must size their transmission
buffer and the retransmission time-out mechanism according to the longest round-trip delay to be anticipated. The
number of retransmissions allowed before a packet is dropped from the retransmission buffer may also be adjusted.
UL scheduling delay parameters are expected to be redefined to accommodate the RTT of the associated deployment
scenario.
3GPP
Release 15 108 3GPP TR 38.811 V15.1.0 (2019-06)

No impacts on the ARQ protocol itself, except parameter sizing needs to account for the longer delay of the NTN
network.
7.3.3.3 Physical layer procedures (ACM, power control)

As mentioned in section 7.3.3.2.1, UE operating in GEO satellite access networks can experience a one-way
propagation time up to 270 msec. Using LEO satellite access network with 600 km orbit, the one way propagation delay
changes continuously between 2 ms, and 7 ms. The slow reaction time is expected to have performance impact on some
of the physical layer procedures particularly those with close loops such as power control and ACM.

While slower reaction on the control loops affects the performance of all the control loops between UE and base station,
most of them require some adjustments in implementation, but not fundamentally different design.
While the link margin may be different for specific links and systems depending on applications, satellite power is
typically at a premium. Due to the large free space loss and limited EIRP and battery power available at UE, power
margin is also limited for mobile terminals. Thus, a very limited amount of power control, if at all, is available for the
GEO satellite links. Due to the long delay in the loop, the power control is not expected to track fast fading, but may be
used to track slower power variations.
For Ka-band satellites, ACM is an essential tool that maintains connection through rain fades, which typically changes
somewhat slower than the ½ second round trip delay. It generally works well with some hysteresis to avoid excessive
oscillations between two ACM modulation coding modes. But, this reaction time is too slow for ACM to adapt for
changes of signal strength for mobile terminals when line of sight is interrupted by shadowing.
For GEO systems in S-band, the main issue is multipath fading, which can be much faster than ½ second round trip
delay. As such, ACM will not be able to follow it. ACM algorithm typically attempts to settle on a modulation coding
mode that closes the link if possible, by giving up some power to maintain a margin.
For LEO satellites, ACM may also be used to adapt for the large variation of free space loss. The variation is
sufficiently slow compared to the 20 ms worst case round trip delay. It should also be able to react to shadowing fades
to a large extent, but still unable to follow fast fading.

Further studies need to be performed to define the required margin for power and ACM control loops to accommodate
to the long RTT.
7.3.4 Cell size (Beam foot print)
7.3.4.1 PRACH and Random access

The RTT in NTN can be much larger than the RTT in terrestrial network as analyzed in Section 5.3. Therefore, it is
necessary to consider its impact on PRACH and random access procedure. As shown in Figure 7.3.4.1.1-1, for one
given beam covering a cell, there is one common propagation delay for all served UEs and one relative propagation
delay for each served UE. If the common propagation delay can be compensated, then the NTN PRACH signal design
will depend on the relative propagation delay, which is limited up to 200km in current specifications regarding TA
range. However even if only the difference value left, a NTN PRACH signal design and PRACH procedure design are
still needed since the common propagation delay could be thousands of kilometers which cannot be ignored.
3GPP
Release 15 109 3GPP TR 38.811 V15.1.0 (2019-06)
h
Beam
Cell
Smax a1
a2
Figure 7.3.4.1.1-1: NTN systems, new geometry and the cell size

Random Access Response
For non-terrestrial networks, the round-trip time can be much larger than the round-trip time in terrestrial networks (up
to 600 ms for the case of GEO satellites, with bent pipe architecture). However the current window for the PRACH

response in NR, which starts at    N slot
subframe,.μ
   T N
 N symb
slot
Tsf sf
subframe,.μ
slot 
 N symb
slot
symbols after transmitting the last
symbol of the preamble and has the size of "rar-WindowLength", cannot cover this round-trip time [30]. Therefore, the
random access response window length in NR should be revisited to accommodate the round-trip time of NTN.
However, extending the RA response window size in the existing procedure introduces unnecessary UE monitoring
intervals thus more power saving due to large propagation delay in NTNs. Therefore, the solution to handle long
propagation delay with the consideration of power saving at the UE side needs to be further studied.
PRACH Sequence and Format
Considering that the cell size of NTN HAPS is not extremely large, the current NR preamble format design should be
enough to support HAPS. The current NR preamble format design should be enough to support HAPS but for Satellite
it needs to be revisited [30][31][23][39][32]. As shown in Figure 7.3.4.1.1-1, the difference d3 = d2 – d1 depends on the
elevation angle, as listed in Table 7.3.4.1.2-1.
Table 7.3.4.1.2-1 Differential delay d3 = d2 – d1 vs. satellite elevation angle
[degree] cell radius ( ) [km] d3 [km]

10 200 390
20 200 372
30 200 343
40 200 303
50 200 254
60 200 197
70 200 134
80 200 67
3GPP
Release 15 110 3GPP TR 38.811 V15.1.0 (2019-06)
Clearly, the difference d3 exceeds the maximum cell coverage supported by the NR preamble format of 2x 100 km for
satellite elevation angles below 60 degrees. In such a case, NR preamble format needs to be extended for NTN system
considering different footprint of NTN cells. So it should be further studied if new random access preamble format is
needed for spaceborne vehicles.
The effect of the residual differential delay may be mitigated as follows:

- Case 1: NTN differential delay ≤ Max currently specified NR PRACH CP duration.
In this case no change is required to the specified NR RACH procedure.
- Case 2: NTN differential delay > Max currently specified NR PRACH CP duration.
In this case different solutions are possible which are for further study.
RACH Procedure
For PRACH procedure, huge propagation delay [66.7μs, 120ms] [3] corresponding to BS heights [8km, 35786km] will
lead to long RA response procedure [33][32]. As a result, it is necessary to shorten the RA transmission/retransmission
delay and design an NTN-specific RAR window based on the given BS information to make the UE efficient and power
saving.

Regarding NR PRACH waveform/format, no NR spec impacts are identified if the GNSS based techniques are
applicable at the UE side, otherwise, potential enhancements w.r.t PRACH waveform/format may be needed which are
for further study. For RACH procedures, potential enhancements can be studied regardless of whether the GNSS based
techniques are applicable or not.
7.3.4.2 TA in Random access response message

The Timing advance mechanisms ensure that transmissions from all UEs operating in the same cell are synchronized
when received by the same gNB. A TA command is provided to the UE in a RAR message during initial access, as well
as later, to adjust the uplink transmission timing for PUCCH/PUSCH/SRS. TA adjustment is analyzed in Section
7.3.2.2.
The timing advance ( ) is given by:
where and TA is an integer between 0 and 3846.
The maximum value of the TA command sent in RAR message constraints the maximum distance between an UE and
the serving base station, which also defines the maximum allowed cell size, as illustrated in Table 7.3.4.2.1-1.
Table 7.3.4.2.1-1: Terrestrial link distance for 5G NR
Subcarrier spacing 15 kHz 30 kHz 60 kHz 120 kHz 240 kHz 480 kHz
µ= 0 µ= 1 µ= 2 µ= 3 µ= 4 µ= 5
Maximum timing advance 0.67 0.335 0.1675 0.0838 0.0419 0.0209
(ms)
Maximum link distance (TA 300 150 75 37.5 18.75 9.38
= 3846) (km)
Even at 15kHz, the maximum cell size is below the satellite altitude, for both LEO and GEO scenarios. Therefore, it is
not expected that the timing advance mechanism can be reused as is for satellite networks.
For HAPS scenarios, the HAPS altitude is quite negligible compared to the maximum link distance for low numerology.
In these cases, the maximum allowed cell size for terrestrial operations needs to be slightly reduced to ensure the
maximum link distance is never exceeded. For high numerology however, the maximum link distance may be smaller
than the HAPS altitude, leading to the same conclusion as for satellite operations.
3GPP
Release 15 111 3GPP TR 38.811 V15.1.0 (2019-06)

Compensation of entire propagation delay by TA
For satellite operations, it is expected that TA mechanism cannot be reused as is to compensate for the propagation
delay. For HAPS operations, the mechanism can be reused as is in some cases (low numerology and/or low HAPS
altitude). The maximum cell size shall then be redefined to ensure that the maximum allowed link distance is never
exceeded.
Common delay known and handled by the network

In this case, TA compensates only for differential delay/distance, so that the maximum TA value defines the maximum
allowed differential delay/distance. Considering Figure 7.3.4.1.1-1, it can be easily shown that the maximum
differential distance is achieved when is minimum. As shown in Table 7.3.4.2.2-1, the maximum differential
distance is almost equal to the maximum cell diameter for .
Table 7.3.4.2.2-1: Maximum cell size for SCS=15kHz, GEO/MEO/LEO cases
NTN vehicle altitude Min Elevation angle Max Differential Maximum cell
orbit distance d3 = d2 –
(deg) diameter Smax
d1 (km)
(km)
GEO (at 35 786 km) 10° 300 km 304.66
MEO at 10 000 km 10° 300 km 304.73
LEO at 1500 km 10° 300 km 305.05
LEO at 600 km 10° 300 km 305.50
In first approximation, it can be stated that the existing TA mechanism can compensate the differential delay if the cell
diameter does not exceed the maximum link distance given in Table 7.3.4.2.1-1.

In HAPS operations, extension of the maximum TA value could be considered to extend the maximum allowed cell size,
especially for high numerology.
In satellite operations, it is expected that the network can take into account the common propagation delay. To extend
the maximum cell size, different approaches can be studied further.
7.3.5 Propagation channel
7.3.5.1 DMRS frequency density

For PDSCH, the mapping to physical resources of the Demodulation reference signals is described in chapter 7.4.1.1.2
of [21] recalled below. There is either one DM-RS symbol per slot or 2 in consecutive OFDM symbols within a slot.
The resource elements of this DM-RS symbol are scattered over different subcarriers according to two configuration
types.
In configuration type 1, DM-RS symbols are inserted every 2 subcarriers in PDSCH. In configuration type 2, DM-RS
symbols are inserted every 5 subcarriers in PDSCH.
Table 7.3.5.1.1-1: Minimum coherence bandwidth of PDSCH supported for a given SCS value
Minimum supported coherence bandwidth

(assumed no orthogonally required) (kHz)
SCS (kHz) Configuration type 1 Configuration type 2
15 30 75
30 60 150
60 120 300
120 240 600
3GPP
Release 15 112 3GPP TR 38.811 V15.1.0 (2019-06)
For PBCH, the mapping to physical resources of the Demodulation reference signals is specified in Table 7.4.3.1-1 of
[21]. DM-RS are inserted every four subcarriers in PBCH on all OFDM symbols.
For PDCCH, the DM-RS symbols are inserted every 4 subcarriers as for PBCH (chapter 7.4.1.3.2 in [21]).
Table 7.3.5.1.1-2: Minimum coherence bandwidth of PBCH/PDCCH supported for a given SCS value
SCS (kHz) Minimum supported

coherence bandwidth
(kHz)
15 60
30 120
60 240
120 480
The minimum coherence bandwidth for each NTN deployment scenario can be computed with the following formula:
Minimum coherence bandwidth = 1/(alpha *delay spread)
with alpha being a constant between 1 and 50. As a worst case, it is set to 50 for the DM-RS mapping selection. The
maximum delay spread depends on the UE antenna directivity.
Table 7.3.5.1.1-3: Maximum delay spread and minimum coherence bandwidth for each deployment
scenario
D1, GEO, Ka D2, GEO, S D3, LEO, S D4, LEO, Ka D5, HAPS, S
band band band band band
Maximum Delay 10 100 100 10 150
spread (ns)
Min coherence >> MHz 200 kHz 200 kHz >> MHz 133 kHz
bandwidth
(NOTE 1, NOTE 2)
NOTE 1: In Ka band, typical antenna directivity is taken into account in the delay spead estimate
NOTE 2: In S band, delay spread in satellite scenarios is lower than in HAPS scenario, because the min operating
SNR is lower which leads to discard largest delay spread.

For S band, the minimal coherence bandwidth has been computed with the formula in the previous clause and an alpha
value corresponding to a theoretical worst case (Alpha set to 50). This may put some constraints on the SCS value
selection which needs to be further studied.
Enhancement to NR specifications with respect to DMRS frequency diversity may not be needed depending on possible
SCS choice constraints that needs to be further studied.
7.3.5.2 Cyclic prefix

Originally, the NR CP has been dimensioned for terrestrial transmissions characterized by high delay spread due to
multi path propagation. This clause describes selection criteria of the CP that need to be addressed when considering
using NR in Non-Terrestrial Network deployment scenarios.
3GPP
Release 15 113 3GPP TR 38.811 V15.1.0 (2019-06)

Delay spread in satellite propagation channels
Signal echoes are associated to the presence of indirect rays that reach the receiver antenna and carry a significant
energy with respect to the energy of the direct ray.
ITU-R recommendation [34] defines for the 2 GHz band three parameter sets of wideband models, including LOS and
NLOS cases, applicable for an elevation range from 15 to 55° and for urban, suburban and rural environments. The
delay spread of these three parameter sets ranges between 180 ns to 250 ns, whereas the 250 ns are stated to cover 90%
of the cases.
For higher elevations than 55°, we assume that the delay spread of the satellite channel will be in the same range or
even lower due to the traveling distances of the echoes arriving at a receiver.
Few papers are available on delay spread measurements in Ka-Band. Reference [35] is stating the coherence bandwidth
to be 30 MHz at 40 GHz with omnidirectional antennas. According to [36], the coherence bandwidth (Δ f)c of a channel
with maximum delay spread Tm is
(Δ f)c ≈ 1 / (5T)
For the stated coherence bandwidth in [35], this results in a maximum delay spread of T m = 25 ns for omni-directional
antennas. For directional antennas, echoes with significant delay are normally filtered out by the antenna radiation
pattern, so flat fading can be assumed for Ka-band signals.
CP length defined in NR
The possible CP lengths currently defined for New Radio [21] are summarized in the following table:
Table 7.3.5.2.2-1: CP lengths and minimum/maximum RF bandwidth of NR as currently defined in [21]
Subcarrier spacing (SCS) SCS [kHz] normal CP extended CP length

configuration parameter µ length [µs] [µs]
0 15 4,688 Not defined
2 60 1,172 16,67

It is observed that lower numerologies (µ= 0, 1) are associated with CP lengths exceeding the requirement of NTN,
resulting in a slightly reduced spectral efficiency due to the over-dimensioned CP (e.g. overhead is for µ= 0: (4.688µs -
0.25µs) / 66.67 µs = 6.7 %).
High numerologies (µ= 3, 4) result in CP lengths which are well matching to propagation characteristics in Ka-Band.
The extended CP for a SCS of 60 kHz is not required, because it is significantly larger than required for satellite
applications.
Enhancement to NR specification is not expected for NTN applications due to the NTN channel model delay spread
compatible with the existing specified CP values.
7.3.6 Duplex mode
7.3.6.1 FDD/TDD Duplexing mode

Most of the existing satellite systems operate in the frequency bands designated for FDD mode, with defined transmit
direction. For some frequency bands, TDD mode is possible.
3GPP
Release 15 114 3GPP TR 38.811 V15.1.0 (2019-06)
When considering TDD mode, a guard time is necessary to prevent UE to simultaneously Transmit and reception. This
guard time directly depends on the propagation delay between UE and gNB. This guard time will directly impact the
useful throughput and hence the spectral efficiency.
When considering Non-Terrestrial networks, this guard time should commensurate the round trip delay.
Guard time would range between 2 x 7 ms for LEO at 600 km and 2 x 270 ms for GEO satellite access networks since
NTN terminals can experience a one-way propagation time of
- 240 ms at minimum and 270 ms at maximum between UE and satellite base station for GEO
- 2 ms at minimum and 7 ms at maximum between UE and satellite base station for LEO at 600 km altitude
Such excessive guard time would lead to a very inefficient radio interface especially in GEO or even MEO based access.
It may be acceptable in the case of LEO access system with the need to deal with the variable delay.

The applicability of the duplexing mode TDD or FDD depends on the regulations (ITU-R and/or national) associated to
the targeted spectrum.

FDD is preferred duplexing mode for most NR based NTN access network.
In case the regulations allow it, TDD mode can be considered for both HAPS and LEO satellite based access network
with potential NR impacts if required.
7.3.7 Satellite or aerial Payload performance
7.3.7.1 PT-RS

Phase variations in time domain can be caused by different phenomena: presence of phase noise, frequency drifts due to
Doppler shift, or due to insufficient frequency synchronization (e.g. residual CFO), etc. Phase noise, caused by
imperfect oscillator implementation technology destroys the orthogonality of subcarriers in OFDM-based systems
especially in FR2. Phase noise causes CPE, resulting in a constant rotation angle of the modulation constellation, and
ICI, resulting in scattering of the constellation points in OFDM based systems. Doppler effects and/or residual CFO
cause frequency shifts translated into time domain phase ramps at OFDM symbol level.
The phase variations translating all the above effects may significantly degrade the performance and need compensation.
When mild, these variations can be absorbed by the DM-RS in the channel estimation process. But when strong phase
variations are present, which is particularly the case at carrier frequencies above 6GHz, more frequent support than the
one provided by DM-RS is necessary in order to track the remaining variations.
In NR, PT-RS has been introduced in [21], [23] to compensate for phase errors. PT-RS configuration in NR is very
flexible and allows user-specific configuration depending on scheduled MCS/bandwidth, UE RF characteristics, DM-
RS configuration, waveform, etc.

It has been proposed that eMBB should be supported for NTN and the modulation order might not always be low. PT-
RS is needed in NTN at high carrier frequencies, where the maximum received SNR is limited by the phase noise.
Considering the high speed of airborne/spaceborne vehicles, CFO and Doppler shift/spread might be worse in NTN and
therefore CFO and Doppler estimation need to be considered.
A number of phase noise masks such as DTH or VSAT use cases considered in [37], or other state of the art bent pipe
of satellite or HAPS payloads as described in Annex B are considered as pertinent for non-terrestrial communications.
Phase noise masks pertinent for terrestrial equipment operating in mmWave have been defined in NR in [38].
3GPP
Release 15 115 3GPP TR 38.811 V15.1.0 (2019-06)
The performance of phase tracking algorithms is highly dependent on a large set of variables such as carrier frequency,
subcarrier spacing, UE specific RF characteristics, configured MCS/bandwidth, frame length, DMRS configuration,
Doppler, receiver implementation options, etc.

PTRS is needed in NR supporting NTN for phase error compensation. PTRS configuration in NR is very flexible.
Typical phase noise masks of state of the art bent pipe of satellite or HAPS payloads in Non-Terrestrial networks can be
efficiently compensated by the current NR design in the absence of important Doppler shifts and/or residual CFO at a
carrier frequency of up to 30 GHz. Some further investigation is needed in the case of important Doppler shifts and/or
residual CFO, or in the presence of the specific phase noise masks of on board payloads significantly different from the
ones considered in cellular network so far, or with very large channel bandwidths.
7.3.7.2 PAPR

A key component in the satellite payload architecture is a PA. It exhibits nonlinear behavior when driven near saturation
in an effort to increase power efficiency. Nonlinear distortion causing constellation warping and clustering, thus
complicating signal reception. One measurement that determines the vulnerability of the transmitted signal to nonlinear
distortion is the PAPR, with higher values associated with worse impact.
In the NR downlink, CP-OFDM is used. CP-OFDM is composed of superposition of narrowband orthogonal subcarriers,
resulting in higher PAPR values compared with the underlying modulation in a single carrier. The PAPR is a function
of the number of subcarriers in the signal, as well as the underlining modulation.
The amount of distortion caused by nonlinear amplifier characteristics is a function of the PAPR of the signal. The
greater the PAPR, the greater the distortion can be. The distortion can be reduced by increasing the backoff of the
amplifier operating point. But this reduces the amplifier efficiency accordingly. Tradeoffs are often made between the
amplifier OBO and the nonlinear signal distortion, such that the total degradation of output power and signal distortion
is minimized.
As an example, the figures (NOTE 1) below show the total degradation of CP-OFDM with over a typical satellite
transponder is about 6 dB for QPSK modulation and 7.6 dB for 16-QAM, whereas that for DFT-spread-OFDM is about
4 dB for QPSK and 6 dB for 16-QAM. The difference between CP-OFDM and DFT-spread-OFDM is 1.6-2 dB.
NOTE 1: See ETSI TR 103 297 "Satellite Earth Stations and Systems (SES), SC-FDMA based radio waveform
technology for Ku/Ka band satellite service", , V.1.1.1., (2017-07). Pages 29-30.
QPSK Modulation (b) 16-QAM modulation

Figure 7.3.7.2.1-1: DFT-S-OFDM vs CP-OFDM total degradation
The additional 2 dB output backoff may represent 20 to 40 percent reduction of link capacity compared to State of the
art satellite radio interface, depending on whether the satellite is bandwidth limited or power limited.
As for HAPS, the payload architecture may be very similar to satellite, and power amplifier remains to be a key
component, although SSPA is likely to be used instead of traveling wave tube. Regardless, the PAPR difference
3GPP
Release 15 116 3GPP TR 38.811 V15.1.0 (2019-06)
between CP-OFDM and DFT-spread-OFDM also requires different OBO which translates to different total distortion.
The higher total distortion of CP-OFDM ultimately means lower capacity. In this case, however, the problem may be
less acute, since the payload power efficiency may not be as critical.

The use of CP-OFDM in the downlink does not restrict NR operation in non-terrestrial networks, but it may affect the
system performance:
- In one case, a satellite transponder is sufficiently wide and sufficiently powerful to accommodate more than one
FDM carriers, each of which is a different NR CP-OFDM signal. The satellite amplifier is backed off to
minimize the intermodulation between these FDM carriers within the same transponder. For such cases, the
distortion introduced by the amplifier nonlinearity is small, with little impact.
- In second and much more common case for communicating to small UEs, the satellite amplifier is used to send
only one NR CP-OFDM downlink. It is highly desirable to operate the amplifier with as small OBO as possible.
But, due to the higher PAPR of CP-OFDM signal, sufficient OBO is necessary. To close the link, it may be
necessary to reduce the size of CP-OFDM carrier or operate the CP-OFDM carrier with a lower modulation and
coding mode. Either way, the forward link capacity is reduced significantly.
On the uplink, DFT-spread-OFDM might be beneficial.

PAPR reduction techniques of CP-OFDM signal on the downlink would be beneficial to optimize the capacity of non-
terrestrial networks and therefore could be considered in future studies.
7.3.8 Network architecture
7.3.8.1 Protocols

Non-Terrestrial networks differ from typical cellular networks in terms of network architecture, deployment scenarios
as well as coverage which may span across several countries.
Potential impacts on the NR specifications could be minimized through suitable mapping options of the Logical NG
RAN architecture onto the NTN network architecture.
The respective NG-RAN logical architecture and the NTN network architecture are depicted below. The area of impacts
associated to the NG-RAN mapping options need to be identified.
UE gNB-DU gNB-CU 5GCN DN

NR-Uu F1 NG (=N2/N3) N6
Figure 7.3.8.1.1-1: 3GPP NG-RAN (or NR radio access) architecture defined in TS 38.401
3GPP
Release 15 117 3GPP TR 38.811 V15.1.0 (2019-06)
NTN Satellite
terminal or HAPS
Hub
Data
network
Service link Feeder link
Figure 7.3.8.1.1-2: Typical Non-Terrestrial network physical architecture
Furthermore, when dealing with mobility of UEs, one should distinguish:
- The mobility induced by the motion of the NTN platform, in particular when considering Non-Geostationary
Satellites;
- The mobility of the terrestrial equipment:
- when moving from one beam to another beam generated by a satellite or a HAPS
- The mobility of UE between beams generated by different satellites or HAPS
- The mobility of UE between NTN (Satellite or HAPS) and cellular access
Location update, paging and hand-over RAN related protocols may need to be adapted to accommodate
- The extended delay of NTN for intra NTN access mobility
- The differential delay when mobility between satellite access and cellular access
- The mobility of the cell pattern generated by non geo-stationary satellites
7.3.8.1.2 Assessment of conditions for NR operation in Non-Terrestrial Networks

RAN architecture mapping
There are several options to instantiate the RAN architecture in Non-Terrestrial networks. It depends on whether the
satellite or HAPS implement a bent pipe payload or a processed payload. This leads to the following example mapping
options that prevent the need to create new interfaces or reference points. Note that there may exist other mapping
options not listed here.
Table 7.3.8.1.2-1: RAN architecture example mapping options in Non-Terrestrial networks
Satellite/HAPS payload NG RAN architecture

Bent pipe Mapping option 1
Processed – low Mapping option 2
Processed - high Mapping option 3
A bent pipe satellite/HAPS payload implements Radio frequency conversion, analogue filtering and amplification. A
processed payload implements some RAN functions. The type of gNB functions that can be embarked will be
constrained by the available power and mass on board satellite/HAPS.
Here under are depicted different mapping options of the logical NG RAN architecture onto the physical architecture of
Non-Terrestrial networks which prevents to modify the NG RAN architecture.
3GPP
Release 15 118 3GPP TR 38.811 V15.1.0 (2019-06)
NTN NG Radio access network

NTN
UE Radio Frequency
Remote
Bent pipe
Radio
Unit gNB 5G CN
NR-Uu NR-Uu N1/2/3 N6 Data
network
Figure 7.3.8.1.2-1: Mapping option 1 - NG RAN architecture in Non Terrestrial network with bent pipe
payload

NTN
UE Remote
gNB-DU
Radio gNB-
CU 5G CN
F1 over Unit
NR-Uu SRI N1/2/3 N6 Data
network
Figure 7.3.8.1.2-2: Mapping option 2 - NG RAN architecture in Non Terrestrial network with gNB-DU
processed payload
NOTE: SRI refers to Satellite Radio Interface

NTN
UE Remote
gNB
Radio 5G CN
N1/2/3 over Unit
NR-Uu SRI N1/2/3 N6 Data
network
Figure 7.3.8.1.2-3: Mapping option 3 - NG RAN architecture in Non Terrestrial network with gNB
processed payload
3GPP
Release 15 119 3GPP TR 38.811 V15.1.0 (2019-06)
Table 7.3.8.1.2-2 highlights the difference in terms of interfaces transported by the satellite links and NG RAN
functions implemented in the NTN physical entities.
Table 7.3.8.1.2-2: Mapping options of the logical architecture of NR RAN onto the physical
architecture of Non-Terrestrial networks
Mapping Satellite/HAPS Service Satellite/HAPS Terminal side Network side

options link feeder link
1 –Bentpipe NR-Uu radio interface NR-Uu radio interface UE NTN Remote Radio
payload Unit + gNB
2 –Processed NR-Uu radio interface F1 over Satellite UE NTN Remote Radio
payload (gNB- radio interface (SRI) Unit + SRI modems +
DU) gNB-CU
3 –Processed NR-Uu radio interface N1, N2, N3 over UE NTN Remote Radio
payload (gNB) Satellite radio Unit + SRI modems
interface (SRI)
Given that Non-terrestrial networks feature longer propagation delays, the timers associated to the protocols transported
over the feeder and service links may require to be extended. This applies for example to F1 as well as N1, N2 and N3
reference points.
In mapping option 1, the Satellite or HAPS implements an RF repeater (with frequency conversion), whereas in
mapping option 2, they implement some kind of "intermediate" node which could be based on the outcomes of
3GPP TR 38.874: "Study on Integrated Access and Backhaul".
It is recommended to consider the outcome of this TR 38.874 and see how it can be applicable to Non-Terrestrial
networks (especially for NTN deployment scenarios 2 and 3) and what areas of impact it may create. Other mapping
options may be considered depending on different CU-DU lower layer split for NR that may be defined (See TR
38.816).
Location update, paging and hand-over RAN related protocols
Table 7.3.8.1.2-3 identifies the possible areas of impact on NR associated to Location update, paging and hand-over
RAN related protocols in different types of non-terrestrial networks.
Table 7.3.8.1.2-3: Potential areas of impact on NR specifications associated to Location update,

paging and hand-over RAN related protocols for different types of non-terrestrial networks
NTN Deployment scenarios GEO Non GEO HAPS

Cell pattern Earth fixed: same as in Motion over earth or Motion over earth (e.g.
cellular earth fixed rotating) &
Possibly UE altitude
dependent
Tracking area Tracking area congruent For further study Same as for GEO and
with cell pattern: same as cellular networks
in cellular
Potential areas of impact on NR Mobility management Location update, paging Case 1: tracking area
specifications procedures (Paging, hand- and hand-over RAN corresponds to HAPS
over, location update): related protocols to be coverage: no impact
Potential extension of adapted depending on Case 2: Tracking area
some timers tracking area design smaller than HAPS total
Handling of network coverage, same impact as
identities to be adapted for Non GEO are expected

Thanks to suitable mapping options of the logical architecture of NG RAN onto the physical architecture of non
terrestrial network, modifications of the NG RAN architecture are not needed in terms of reference points, interfaces or
functions.
However the N1/N2/3 and GTP based F1 interface protocols may need to be adapted to accommodate the non terrestrial
networks feeder link characteristics (long delay, BER).Depending on the non-terrestrial network deployment scenarios,
3GPP
Release 15 120 3GPP TR 38.811 V15.1.0 (2019-06)
other impacts on NR specification may have to be considered among which, location update, paging and hand-over
RAN related protocols and handling of network identifies.
Last, it may be considered how the architecture being defined in 3GPP TR38.874 "Study on Integrated Access and
Backhaul" can be used in the context of various Non-Terrestrial networks deployment scenarios.
3GPP
Release 15 121 3GPP TR 38.811 V15.1.0 (2019-06)
8 Recommendations on the way forward
8.1 General outcomes

This report has
- provided detailed descriptions of some deployment scenarios for non-terrestrial networks and the related system
parameters such as altitude, orbit, etc.
- defined a way to generate Non-Terrestrial network channel models leveraging on the models defined in TR
38.901 "Study on channel model for frequencies from 0.5 to 100 GHz (Release 14)"
- identified potential key impact areas on the NR to support Non-Terrestrial networks
In line with 3GPP TR 38.901, the study considered channel models in Non-Terrestrial networks operating in any
frequency bands between 0.5 and 100 GHz.
8.2 Reference deployment scenarios

A number of scenarios is considered in clause 5.1 "Scenarios overview" including GEO, Non-GEO as well as HAPS
based access networks. The intent is to assess relatively worst-case scenarios with respect to Doppler shift, Doppler
variation rate, Delay and Delay variation rate. For non-GEO deployment scenarios, the altitude has been set to 600 km,
which is pessimistic with respect to the Doppler, and Doppler rate, related effects.
Non-Terrestrial access networks can be used to serve directly User Equipment or indirectly via a relay node.
8.3 Non-Terrestrial network channel modelling

Leveraging on the models defined in TR 38.901 "Study on channel model for frequencies from 0.5 to 100 GHz (Release
14)", complementary features and specific parameters have been identified and proposed in chapter 6 of this report to
generate Non-Terrestrial network channel models for all possible deployment scenarios characterised by altitude/orbit
of space/aerial vehicles, user environments (rural, sub-urban, urban and dense urban), targeted user equipment type
(omni or directive antennas, LOS/NLOS conditions) as well as frequency bands.
8.4 NR impacts to support Non-Terrestrial networks

8.4.1 Type of potential NR impacts
The potential NR impacts are characterised according to the classification in Table 8.4.1-1:
Table 8.4.1-1: Type of Potential area of NR impacts to support Non-Terrestrial networks
Type of Potential area of impact Modifications of 3GPP NR specifications

Higher layers impact Layer 2/3 protocols/architecture
Physical layer impact Parameters settings, procedures or physical channels and
signals
8.4.2 Assessment of potential NR impacts to support Non-Terrestrial

networks
Table 8.4.2-1 identifies the potential area of impacts on the NR specification. This will depend on the deployment
scenarios. The intent is to identify the main studies to be carried out taking into account the defined NTN channel model
3GPP
Release 15 122 3GPP TR 38.811 V15.1.0 (2019-06)
to enable operation of NR in Non-Terrestrial networks. Other areas of impact are identified which could lead to
performance improvement.
Table 8.4.2-1: Evaluation of NR impacts to support Non-Terrestrial networks
Non-Terrestrial Potential areas

Impacted NR
network Effects of impact to be Comment
features
specifics further studied
Paging and Hand-over procedures should
be adapted to take into account the relative
motion of the cell pattern with respect to the
Moving cell Hand- Higher layers
tracking area. Further analysis on tracking
pattern over/paging impact
area design may need to be carried out.
Mobility management is also to be
considered (NOTE 1)
Motion of the Delay Physical layer Alignment of uplink signals may need to be
TA adjustment
space/aerial variation impact considered
vehicles The preferred SCS values for Non-
Terrestrial Networks may be respectively 60
Initial downlink KHz for frequency bands lower than 6 GHz
No impact
synchronization and 240 KHz for frequency bands above 6
Doppler
GHz. However it can also operate with lower
SCS value
DMRS time The preferred DM-RS configuration may be
No impact
density type 1 to cope with Doppler variation rate
Higher Layers & Need to adapt the HARQ specification.
HARQ
physical layer Deactivation and/or enhancements of NR
Impact HARQ can be considered
Physical layer
The operation/configuration of Adaptive
Procedures Physical layer
power and coding/modulation control loop
Long (ACM, power impact
Altitude protocols may have to be adapted.
latency control)
MAC/RLC Higher layers Timers limit of MAC/RLC and higher layers

Procedures impact loop protocols may have to be extended
TA in Random Doppler/Delay (NOTE 2) compensation

access Physical layer technique (NOTE 3) can be implemented.
Differential response impact Further analysis/simulations using the NTN
Cell size
delay message channel model is needed. Adaptations of
Physical layer PRACH format and random access
Random access
impact procedure may have to be considered.
Non terrestrial network propagation channel
may feature a frequency selective at most
DMRS comparable with cellular channel
Impairments frequency No impact
Propagation density
channel
Non terrestrial network propagation channel

Impairments Cyclic prefix No impact may feature a worse delay spread at most
comparable to cellular channels.
FDD is preferred especially for most satellite
Regulatory Duplexing mode Higher layers
Duplex scheme systems. TDD can be considered for HAPS
constraints (TDD/FDD) impact
and for LEO
Potential Satellite Radio links are typically operated
Phase noise constraint on the with relatively low order modulation scheme,
PT-RS
impairment operation to be in most of the cases up to 16QAM
Satellite or aerial further studied
Payload Uplink: It is recommended to use DFT-S
performance OFDM
Physical layer
Back-off PAPR Downlink: Low PAPR scheme may improve
impact
performance. Howver not mandatory to
support non-terrestrial networks
3GPP
Release 15 123 3GPP TR 38.811 V15.1.0 (2019-06)
NOTE 1: Some of the considered deployment scenarios assume indirect access based on fixed or even mobile
relays. For NGSO-based deployment scenarios NTN and/or relays mounted on a moving platform such as
a train, mobility management also needs further studies.
NOTE 2: The common delay, the Differential delay as well as their time variation may need to be compensated.
NOTE 3: Doppler/Delay compensation techniques can be implemented especially for Non GEO satellites:
- GNSS based techniques: The User Equipment equipped with a GNSS receiver determines its position
and the universal time. Thanks to pre-loaded / updated ephemeris of the satellite constellation which
can be theoretical or actual, the UE is able to compute the position and motion of the possible serving
satellites enabling to determine the Doppler shift and variation rate as well as the absolute Delay and
Delay variation.
- Non GNSS based techniques can also be envisaged and implemented.
3GPP
Release 15 124 3GPP TR 38.811 V15.1.0 (2019-06)
Annex A: Example of reference scenario for calibration of

large scale parameters
For large scale calibration, fast fading is not modelled. The calibration parameters can be found in Table A-1.
Table A-1: Simulation assumptions for large scale calibration
Parameter Values
Scenarios NGSO satellite at 1500km, suburban environment
Satellite position (7871,0,0)
Number of beams 4
Beam layout (as seen from satellite)
Beam centre B1: (6371,0,0), elevation = 90°

B2: (6347, 481,278), elevation = 65°
B3: (6347, 481,-278), elevation = 65°
B4: (5242,3622,0), elevation = 45°
Satellite antenna pattern parameter ka = 10
Satellite EIRP 36dBW/MHz
Bandwidth 5MHz
Co-phased array – Dual linear polarization
UE antenna configurations (M, N, P) = (1, 2, 2), d H  0.5 , Polarization: 0°, 90°
UE orientation TBD
Handover margin (for calibration) 0dB
UE distribution Uniform dropping, 25 users per beam
UE attachment Geometric
UE noise figure 7 dB
Fast fading channel Fast fading channel is not modelled
Carrier Frequency 2 GHz
Metrics 1) Coupling loss – serving cell (based on LOS pathloss)
2) Geometry (based on LOS pathloss) with and without white noise
3GPP
Release 15 125 3GPP TR 38.811 V15.1.0 (2019-06)
Annex B: Non Terrestrial network characteristics
B.1 NTN Phase noise masks

DTH P1 and P2 and VSAT P1 and P2 phase noise masks were previously defined by DVB [37]. in the context of
satellite communications.
3GPP
Release 15 126 3GPP TR 38.811 V15.1.0 (2019-06)
Annex C: Change History
Change history
Date Meeting TDoc CR Rev Cat Subject/Comment New
version
2017-06 RP-76 RP-170983 Skeleton report provided as input to RAN #76 0.0.0
2017-06 RP-76 RP-171453 Agreed version as output of RAN #76 including pCRs of 0.1.0
RP-170917, RP-170918, RP-171447, RP-171448.
RP-171578, RP-171579, RP-171580, RP-172075.
2017-11 RP-78 RP-172179 Cleanup of v0.2.0 to align with 3GPP drafting rules 0.2.1
RP-172274, RP-172768, RP-172769
RP-180036, RP-180135, RP-180180, RP-180543
RP-180661, RP-181381, RP-181382, RP-181383, RP-
181392, RP-181394
2018-06 RP-80 Approved TR version (including editorial cleanup of v1.0.0) 15.0.0
2019-06 RP-84 RP-190842 0001 - F Corrections for TR 38.811 Chapter 6 Non-Terrestrial 15.1.0
Networks channel models
Note: Some renumbering of clauses/figures/equations,
reuse of reference numbers could not be implemented.
3GPP

3GPP TR 38.811

Uploaded by

Copyright:

Available Formats

3GPP TR 38.811

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

3GPP TR 38.811

Uploaded by

Copyright:

Available Formats

3GPP TR 38.811 V15.1.

3rd Generation Partnership Project;

3GPP support office address

No part may be reproduced except as authorized by written permission.

6.6.2 Path loss and Shadow fading ...................................................................................................................... 48

7.3.4.1.2 Assessment of conditions for NR operation in Non-Terrestrial networks .................................... 109

x the first digit:

1 presented to TSG for information;

2 presented to TSG for approval;

3 or greater indicates TSG approved document under change control.

The objectives for the study are the following

- For a specific reference, subsequent revisions do not apply.

[1] 3GPP TR 21.905: "Vocabulary for 3GPP Specifications".

[15] F. P. Fontan, M. Vazquez-Castro, C. E. Cabado, J. P. Garcia and E. Kubista, "Statistical modeling

[28] O. Kodheli, A. Guidotti, and A. Vanelli-Coralli. "Integration of Satellites in 5G through LEO

[36] T. Rappaport, "Wireless Communications", New Jersey: Prentice Hall, 1996

[46] O. Kodheli, A. Guidotti, and A. Vanelli-Coralli. "Integration of Satellites in 5G through LEO

3 Definitions, symbols and abbreviations

Beam throughput: data rate provided in a beam

User Throughput: data rate provided to a terminal

ICI Inter-Carrier Interference

4 Non-Terrestrial Networks Overview – background

4.1 Roles for Non-Terrestrial Networks in 5G system

4.2 5G Use Cases wherein Non-Terrestrial Network

Table 4.2.1-1: 5G use cases for Satellite access networks

5G 5G Use case 5G Use case description Satellite service 3GPP

The intent is to off load popular content from

5G 5G Use case 5G Use case description Satellite service 3GPP

5G 5G Use case 5G Use case description Satellite service 3GPP

eMBB Direct to Public safety authorities want to be able to Broadcast/Multicast TR 22.864,

Table 4.2.1-2: 5G use cases for Aerial access networks

5G 5G Use 5G Use case description Aerial access service 3GPP

4.3 Satellite and aerial access network architecture principles

Very Small Spaceborne Spaceborne

Handheld or IoT device Airborne

Handheld Airborne Airborne

Very Small Airborne Airborne

4.4 Characteristics of NTN Terminals for satellite / aerial access

Very Small Aperture Terminal Handheld or IoT devices (3GPP

G/T = Ga – NF – 10*LOG (To+(Ta-To)/(100.1*NF))

- Antenna Gain : Ga in dBi

- Equivalent EIRP of 20 dBm (-10 dBW) under satellite coverage.

Note that other performance may be considered.

4.5 Air/Space borne vehicle characteristics

Table 4.5-1: Typical characteristics of Airborne or Space-borne vehicles

Characteristics Geostationary satellites Non-Geostationary Airborne platforms

Satellite & aerial Typical minimum Elevation Rationale/remarks

4.6 Coverage pattern of NTN

Table 4.6-1: Typical beam foot print size

Attributes GEO Non-GEO Aerial

Transparent (bent-pipe) Non Transparent (On Board processor)

Figure 4.6-1: NTN Beam patterns

4.7 Non-Terrestrial Network architecture options

Table 4.7-1: 5G system elements mapping in NTN architecture

5G elements - NTN elements mapping

5 Non-Terrestrial Networks deployment scenarios

5.1 Scenarios overview

Main attributes Deployment-D1 Deployment-D2 Deployment-D3 Deployment-D4 Deployment-D5

Carrier frequency between air / space-borne platform and UE

G/T = Ga – NF – 10LOG (To+(Ta-To)/(100.1NF))

Doppler shift formula: ∆F = FoV cos (θ)/c