Impacts of PCI Conflict

Computer Networks 76 (2015) 4254
Contents lists available at ScienceDirect
Computer Networks
journal homepage: www.elsevier.com/locate/comnet
Analysis of the impact of PCI planning on downlink throughput

performance in LTE
R. Acedo-Hernndez a,, M. Toril a, S. Luna-Ramrez a, I. de la Bandera a, N. Faour b
a
b
Dpt. Ingeniera de Comunicaciones, ETSI Telecomunicacin, Universidad de Mlaga, Campus Universitario de Teatinos, s/n, E-29071 Mlaga, Spain
Ericsson, Severo Ochoa 55, 29590 Mlaga, Spain
a r t i c l e
i n f o
Article history:
Received 26 November 2013
Received in revised form 9 April 2014
Accepted 22 October 2014
Available online 4 November 2014
Keywords:
Mobile network
Long Term Evolution
Physical cell identity
Reference signal
Throughput
a b s t r a c t
The planning of Physical Cell Identities (PCI) has a strong impact on the performance of
Long Term Evolution cellular networks. Although several PCI planning schemes have been
proposed in the literature, no study has quantied the performance gains obtained by these
schemes. In this paper, a comprehensive performance analysis is carried out to quantify the
impact of PCI planning on user quality of service and network capacity in the downlink of
LTE. First, an analytical model for the inuence of PCI planning on reference signal
collisions is developed. Based on this model, different PCI planning schemes are tested
on a dynamic system-level simulator implementing a macrocellular scenario. During the
analysis, both Voice-over-IP and full buffer services over time-synchronized and non-time
synchronized networks are considered. Results show that call blocking and dropping for
real-time services and user throughput for non-real time services can be signicantly
improved by a proper PCI plan.
2014 Elsevier B.V. All rights reserved.
1. Introduction
The size and complexity of current mobile networks
make it very difcult for operators to manage them. Thus,
a huge effort has been made by standardization bodies
and vendors to dene and develop automatic network
operation features [1]. As a result, the Long Term Evolution
(LTE) mobile communication standard includes SelfOrganizing Networks (SON) capabilities [2,3]. SON features
aim to perform planning, optimization and healing tasks
with minimal human intervention.
Physical Cell Identier (PCI) planning in LTE has been
identied as an important use case for self-planning
[4,5]. A PCI (or Layer 1 identity) is a signature assigned during network planning to identify a base station in mobility
Corresponding author. Tel.: +34 952136311.
E-mail addresses: rah@ic.uma.es (R. Acedo-Hernndez), mtoril@ic.
uma.es (M. Toril), sluna@ic.uma.es (S. Luna-Ramrez), ibanderac@ic.uma.
es (I. de la Bandera), nizar.faour@ericsson.com (N. Faour).
http://dx.doi.org/10.1016/j.comnet.2014.10.023
1389-1286/ 2014 Elsevier B.V. All rights reserved.
functions, such as cell reselection or handover [6]. The

number of PCIs is limited, which forces several base
stations (or eNodeBs, eNBs) to share the same PCI. As a
result, a wrong assignment of PCIs may cause that a user
receives the same PCI from two different cells (problem
referred to as collision) or a serving cell has two neighbors
with the same PCI (referred to as confusion) [2]. Both
situations prevent users from detecting cells, causing that
no radio communication is possible.
At the same time, PCI denes the location in time and
frequency of signaling channels, amongst which are
downlink Cell-Specic Reference Signals (CRS). In the time
domain, CRSs (or pilots) are transmitted in the same OFDM
symbol of the frame structure in all cells. However, in the
frequency domain, each cell transmits CRSs in different
subcarriers depending on the value of its PCI. Thus, each
cell has a specic pilot pattern corresponding to its cell
identity. The number of possible pilot patterns depends
on the antenna conguration, but it is always less than 6
43
R. Acedo-Hernndez et al. / Computer Networks 76 (2015) 4254
[7], causing that CRSs of surrounding cells often collide.

CRS collisions degrade Signal-to-Interference-plus-Noise
Ratio (SINR) estimates, reported by the User Equipment
(UE) and later used by the eNB to select the modulation
and coding scheme (MCS) for downlink transmission. Thus,
an improper PCI planning may give inaccurate SINR estimates, which leads to inefcient data transmission in the
downlink [8].
In the literature, most research efforts on PCI planning
have focused on the avoidance of collision and confusion.
For this purpose, PCI planning is formulated as a graph coloring problem, which can be solved by graph-theoretic
algorithms [911] or general-purpose optimization algorithms (e.g., genetic algorithms [12]). Preliminary studies
considered centralized schemes [13] and later studies evaluate distributed versions [14,15]. Both types of schemes
are considered by 3GPP for LTE [16]. Distributed PCI
assignment is conceived for self-conguring eNBs to support plug-and-play operation. In this approach, a centralized entity provides a list of possible PCI values, which is
then restricted by the eNB by removing those reported
by terminals or by neighbor eNBs through the X2 interface
[17]. Alternatively, the centralized solution is based on a
central entity that stores location and PCI assignment of
all eNBs in the network [18]. Thus, a collision and
confusion free assignment is guaranteed at the expense
of an increased computational complexity, which makes
it suitable only for greeneld design or infrequent replanning processes. In [19], a decentralized method for
detecting PCI conicts based on user measurements is proposed. Recent works have extended the analysis of PCI
planning to heterogeneous LTE networks, consisting of several layers, by considering collision and confusion between
cells of different layers [2025]. Other studies (e.g., [26])
propose optimal CRS allocation patterns in the time and
frequency domains. However, these schemes do not
consider CRS collisions between neighbor cells nor they
are compliant with LTE standard. In [27], a heuristic PCI
planning algorithm is proposed to keep PCI reuse distance
within certain limits while still avoiding CRS collisions in
sectors of the same site. However, although LTE network
performance has been well documented (e.g., in [28,29]),
to the authors knowledge, no study has quantied the
1
1
3
2
2. Problem formulation
Cell search is the rst process executed by a user connecting to an LTE network. This process requires the synchronization of the radio symbols and frame user timing
with that of the eNB. For this purpose, two synchronization
signals are used, namely the Primary Synchronization Signal
(PSS) and the Secondary Synchronization Signal (SSS), broadcasted by the eNB every 10 ms [30]. PSS is used to detect
the carrier frequency and the SCH (Shared Channel) symbol timing, while SSS is used to align frame timing by identifying slots within the frame. Detection of these signals
not only enables synchronization, but also allows the user
to obtain the PCI of the cell.
Fig. 1 shows the structure of PSS and SS frame in the
time domain in the frequency division duplex (FDD) case.
During cell search, the UE rst nds the PSS, which is
located in the last OFDM symbol of the rst time slot in
4
3
impact of PCI planning on downlink data transmission

due to pilot collisions on a system level. In this work, a
performance analysis is carried out to check the impact
of classical PCI planning schemes on user quality of service
and network capacity in the downlink of LTE. The core of
the analysis is the modeling of the inuence of PCI planning on CRS collisions. The proposed model is included in
a dynamic system-level simulator implementing both VoIP
and full buffer services, with time-synchronized and nontime synchronized schemes, in a macrocellular scenario.
The main contributions of this work are: (a) a simple
interference model to evaluate the impact of PCI planning
on LTE downlink performance, and (b) simulation results
showing how much is network capacity and user throughput inuenced by PCI planning for different services. The
rest of the paper is organized as follows. Section 2 formulates the PCI planning problem explaining how PCI planning determines CRS collisions. Then, Section 3 outlines
classical PCI planning schemes proposed in the literature,
for which a downlink interference model is developed in
Section 4. Section 5 presents the results of simulations
carried out to assess the PCI planning approaches. Finally,
Section 6 presents the concluding remarks.
5
4
6
5
10
SSS
PSS
Fig. 1. PSS and SSS frame and slot structure in the time domain in FDD case.
ID
452857
Title
Analysis of the impact of PCI planning on downlink throughput performance in LTE
http://fulltext.study/article/452857
http://FullText.Study
Pages
13

Impacts of PCI Conflict

Uploaded by

Copyright:

Available Formats

Impacts of PCI Conflict

Uploaded by

Document Information

Original Description:

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

Impacts of PCI Conflict

Uploaded by

Copyright:

Available Formats

How does PCI planning impact LTE network performance?

What are some classical PCI planning schemes discussed in the paper?

Computer Networks 76 (2015) 4254

Contents lists available at ScienceDirect

Analysis of the impact of PCI planning on downlink throughput

functions, such as cell reselection or handover [6]. The

R. Acedo-Hernndez et al. / Computer Networks 76 (2015) 4254

[7], causing that CRSs of surrounding cells often collide.

impact of PCI planning on downlink data transmission

You might also like