LTE Cell Planning

LTE RNP

1/16/17

2010-06

Content
Process for Planning the LTE Network
Frequency Planning
Cell ID Planning
TA Planning
PCI Planning
Neighbor Cell Planning
X2 Planning
PRACH Planning

Page 2

Process of LTE Network Design

The general process includes
information collection, network
dimensioning, network detailed
design and cell planning. For cell
planning, it contains frequency
planning, cell ID planning, TA
planning, PCI planning, neighbor
cell planning, X2 interface
planning, and PRACH planning.

Inform
ation
Collecti
on
Network
Dimensio
ning
Network
Detailed
Design
Cell
Plannin
g

Freque Cell
TA
PCI NB
X2 PRAC
ncy
ID
Planni
Cell
Planni H
Planni Planni ng
Planni Planni ng
Planni
ng
ng
ng
ng
ng
Page 3

Content
Process for Planning the LTE Network
Frequency Planning
Cell ID Planning
TA Planning
PCI Planning
Neighbor Cell Planning
X2 Planning
PRACH Planning

Page 4

Frequency Planning
Why and when perform frequency planning?

When the LTE system works on the same frequency band,

interference occurs between the UEs on the edge of cell because
they are close to each other and use the same resources. In this
case, the performance of the UEs deteriorates. The inter-cell
interference coordination (ICIC) technology can be used to change
interference distribution to improve the throughput on the edge of
cell.
When static DL ICIC is used, the entire
bandwidth is divided into three parts, each of
which serves as the edge band of a cell for
reuse. In this case, network planning
engineers need to perform frequency
planning.
Page 5

Content
Process for Planning the LTE Network
Frequency Planning
Cell ID Planning
TA Planning
PCI Planning
Neighbor Cell Planning
X2 Planning
PRACH Planning

Page 6

Cell ID Planning

Page 7

Different from WCDMA cell ID, LTE cell ID consists of 20-bit eNB ID and
8-bit cell ID, which ensures that the LTE cell ID is unique in the entire
network. If the PLMN (MCC + MNC) is used, the LTE cell ID is unique
worldwide. The WCDMA cell ID is unique on each RNC, the GSM and
CDMA cell ID also is similar to the WCDMA cell ID.
The eNB involves the local cell ID, sector ID, and cell ID. It is advised to
plan the three IDs starting from 0, which ensures that they are consistent.

Considerations for Actual Planning

Page 8

Operator may provide numbering rules for different areas and cities. If
the operator have no additional requirements, the only principle of eNB
ID planning is to confirm it s unique in the whole network.

Content
Process for Planning the LTE Network
Frequency Planning
Cell ID Planning
TA Planning
PCI Planning
neighbor Cell Planning
X2 Planning
PRACH Planning

Page 9

TA Planning

TA Concept

In LTE system, TAC is similar to the location area and routing area ( LAC ) in 2G/3G
networks, the tracking area (TA) is used for paging and route update. TA planning
aims to reduce location update signaling caused by location changes in LTE system.

TAI list

TA list identify the tracking areas that the UE can enter without performing a tracking
area updating procedure. The TAIs in a TAI list assigned by an MME to a UE pertain to
the same MME area. Additionally, the TAIs in a TAI list assigned by an MME to a CS
fallback capable UE pertain to the same location area. In this case, the defining of the
relationship between the tracking area(s) and the location area(s) is operator specific.

In LTE system, if an UE changes the TAs in the TAI list, TA update wont be triggered.

Page 10

TA Planning Principle

A TA should be medium. The limitations by the EPC must be considered. (Currently,

each TA supports a maximum of 30 eNBs in the EPC. USN9810
V900R001C03SPC300 100 eNB )

When the suburban area and urban area are covered discontinuously, an
independent TA is used for the suburban area.

TA should be planned for a continuous geographical area to avoid TA discrete

distribution.

The paging area cannot be located in different MMEs.

The mountain or river in the planned area can be used as TA boundary to reduce the
overlapping depth of two TAs. In this way, fewer location updates are performed on
the edge of TA.

The LAC planning in the existing 2G/3G networks can serve as a reference for
planning TAC.

Page 11

Content
Process for Planning the LTE Network
Frequency Planning
Cell ID Planning
TA Planning
PCI Planning
Neighbor Cell Planning
X2 Planning
PRACH Planning

Page 12

PCI Planning
In LTE system, the physical cell identifier (PCI) is used to differentiate radio signals of different
cells. PCI can be reused in the whole network, but should be unique in one coverage area to
ensure UE can distinguish different cells.
PCI= 3 * PCI Group ID+ ID within PCI Group
UE captures ID within PCI Group through demodulating PSCH, and captures PCI Group ID
through demodulating SSCH.
The function of PCIs in the LTE system is similar to that of scrambling codes in WCDMA
system. PCI planning also aims to ensure the reuse distance.
Differences between a scrambling code and a PCI: The scrambling code ranges from 0 to 511
whereas the PCI ranges from 0 to 503. In addition, the protocols do not have specific
requirements for scrambling code planning. Therefore, only the reuse distance needs to be
ensured in scrambling code planning. For PCI planning, however, 3GPP protocols require that
the value of PCI/3 should be 0, 1, or 2 in each eNB.
UENT 3.6 can be used for PCI planning.
Page 13

Actual Considerations
PCIs should be reserved for indoor coverage.
For multiple cities, PCIs should be reserved for border coverage.
For a high site that may lead to cross-cell coverage, a large reuse
distance needs to be set independently.

Page 14

Content
Process for Planning the LTE Network
Frequency Planning
Cell ID Planning
TA Planning
PCI Planning
Neighbor Cell Planning
X2 Planning
PRACH Planning

Page 15

Neighbor Cell Planning

The method of LTE neighbor cell planning is similar to neighbor planning of

GSM/WCDMA/CDMA. Currently, the planning method and tool (UNET 3.6) for LTE are
available.
The configuration is different from GSM/WCDMA/CDMA . There is no BSC/RNC in the LTE
system. When an eNB cell is configured as neighbor cells of other eNBs, external cells must
be added first, which is similar to the scenario where inter-BSC/RNC neighbor cells are
configured on the BSC. That is, neighbor cells can be configured only after the corresponding
cell information is added.

ADD EUTRANINTRAFREQNCELL; DSP EUTRANEXTENALCELL

Page 16

ANR and Neighbor Cell Planning

Automatic Neighbor Relation (ANR) can automatically add and maintain

neighbor relations. However, ANR function is based on UEs measurements ,
so it is closely related to the traffic load in the entire network and additional
time delay is introduced to the handover process.
Therefore, ANR is not recommended in initial phase of network construction.
After initial neighbor relations configured and the number of UEs increasing,
ANR can be used to detect missing neighbor cells and add neighbor
relations, to improve handover and other performance.

Page 17

Content
Process for Planning the LTE Network
Frequency Planning
Cell ID Planning
TA Planning
PCI Planning
Neighbor Cell Planning
X2 Planning
PRACH Planning

Page 18

X2 Interface Planning

X2 interface planning is based on neighbor relations, that is, obtain the

neighbor relations belonging to different eNBs, and configure X2
interface for these two eNBs.

In eRAN1.1 and eRAN2.0, each eNB can support 32 X2 interfaces.

The later version of the ANR can automatically maintain X2 interfaces to

solve the problems with missing X2 interfaces or configuration errors.

Page 19

Content
Process for Planning the LTE Network
Frequency Planning
Cell ID Planning
TA Planning
PCI Planning
Neighbor Cell Planning
X2 Planning
PRACH Planning

Page 20

PRACH Planning - Logical Root Sequence

Indexes

What is the logical root sequence index?

The random access preambles are generated from ZadoffChu sequences that with zero correlation zone.
There are 64 available preamble sequences in each cell.
The 64 preamble sequences are first generated from a
root Zadoff-Chu sequence using cyclic shift. If less than 64
preamble sequences are generated, the remaining are
generated from the next root Zadoff-Chu sequence
corresponding to the logical index.
The logical root sequence index is sent
to the UE through the SIB2.

Page 21

PRACH Planning - Logical Root Sequence

Indexes

Zadoff-Chu sequence
A Zadoff-Chu sequence has good self-correlation and crosscorrelation and is defined as follows:

xu n e
N ZC

un ( n 1)
N ZC

, 0 n N ZC 1

indicates the length of the Zadoff-Chu sequence, and u indicates

the physical root sequence index. The relation between the logical
root sequence index and physical root sequence
index is defined in
u th
protocols.
The preamble sequences are generated from the

root Zadoff-Chu

sequence through the following cyclic shift.

vN CS
v 0,1,...,
x
xu ((ncyclic
Cv ) modshift
N ZC ) value is
defined as follows:
u ,v (n) The

Cv 0

RA
RA

dstart v nshift (v mod nshift ) N CS

Page 22

N ZC N CS 1, N CS 0 for unrestricted sets

N CS 0

v 0,1,..., n

RA RA
shift group

RA
shift

for unrestricted sets

for restricted sets

Causes for Planning the Root Sequence Index

There are 64 preamble sequences in each cell. The preamble sequence

is selected randomly or assigned by the eNB. To reduce interference of

preamble sequences between neighbor cells, the root Zadoff-Chu
sequence index need to be planned properly.
The planning aims to assign the root Zadoff-Chu sequence index for cells

to ensure that different preamble sequences are generated for neighbor

cells through the index. In this way, interference of preamble sequences
between neighbor cells can be reduced.

Page 23

Factors Affecting the Access Radius

Preamble format

Preamble Format

Ncs

Maximum Cell Radius

0

14.5 km

77.3 km

29.5 km

100 km

N CS 1.04875 (6.67 r TMD 2)

The unit of r is km. The unit of TMD is sec. The value of

cell radius and maximum time delay extension.

Page 24

N CS is subject to the

How to Plan (Take Low Speed Cell as an Example)

Step 1: The Ncs value is determined by the cell radius. If the cell radius is
10 km, the Ncs value is 77.

Step 2: The value of 839/77 is rounded down to 10, that is, each index
can generate 10 preamble sequences. In this case, 7 root sequence
indexes are required to generate 64 preamble sequences.

Step 3: The number of available root sequence indexes is 119 (0, 7, 14

833).

Step 4: The available root sequence indexes are assigned to cells. The
assignment principles are similar to those for PCIs.
The planning method of high speed cell is similar to that of a low
speed cell. The algorithm for determining available root sequence
indexes, however, is more complex.
Its recommended to use UNET 3.6 for PRACH planning.

Page 25