Modeling of Tracking Area List-Based Location
Update Scheme in Long Term Evolution
Xian Wang
∗
,XianfuLei
†
, Rose Qingyang Hu
†
,andYiQian
‡
∗
School of Information Science and Technology, Southwest Jiaotong University, Chengdu, Sichuan 610031, China
†
Department of Electrical and Computer Engineering, Utah State University, Logan, UT 84322, USA
‡
Department of Computer and Electronics Engineering, University of Nebraska-Lincoln, Omaha, NE 68588, USA
E-mail: drwangxian@gmail.com, xflei81@yahoo.com.cn, rose.hu@usu.edu, yqian@ieee.org
Abstract—Long Term Evolution uses a new location update
(LU) scheme, called a tracking area list (TAL)-based scheme,
to overcome the defects of the LU scheme used in 2G and 3G
cellular networks. Under the TAL-based LU scheme, each time
a user equipment (UE) performs an LU, it is allocated a group
of tracking areas, referred to as a TAL, within which the UE
can move freely without any LU. The UE performs an LU when
moving out of the TAL. The performance of the TAL-based LU
scheme depends on the allocated TAL. In this paper we develop
a mathematical model to analyze the signaling overhead of the
TAL-based LU scheme for local UEs whose mobility exhibits
strong regularity. We derive formulas for the LU cost and the
paging cost of a TAL allocation strategy. With these formulas
we can find an optimal TAL allocation strategy to minimize the
signaling cost of the TAL-based LU scheme.
I. Introduction
In a cellular n etwork, the location of a user equipment
(UE) must be determined before the successful delivery of an
incoming call to the UE. The network uses a mechanism called
location managemen t to keep track of the UE location. Loca-
tion managem ent consists of two essential op erations, n amely,
location update (LU) and paging. LU is the process through
which the UE dynamically updates its location registration
in network d atabases. Paging is the process through which
when an incoming call arrives, the network pinpoints the area
where the UE currently resides so as to d eliver the call by
broadcasting a paging message in cells close to the cell where
the last LU befo re the call arrival occurs. To facilitate location
management, the coverage of the network is partitioned into
registration areas, each of which comprises an aggregation of
cells. In Global System for Mobile Communication (GSM)
and Universal Mobile Telecommunications System (UMTS), a
registration area is called a location area (LA) in the circuit-
switched service domain, and called a routing area (RA) in
the packet-switched service domain. In Long Term Evolution
(LTE) a registration area is called a tracking area (TA).
In GSM and UMTS, an UE performs an LU whenever
moving into a new LA/RA. Define a paging area as an area
within which the UE can be located upon the arrival of an
incoming call. The above LU scheme used in GSM and UMTS
is in trinsically a static LU scheme, owing to the fact that all
the UEs in the network have paging areas of the same size,
i.e., the size of an LA/RA, irrespective of the diverse mobility
and traffic characteristics of individual UEs. Intuitively, for
UEs with dense call arrivals, frequent LUs are necessary
to reduce the paging area size and accordingly the paging
cost; whereas for UEs with sparse call arrivals, infrequent
LUs are necessary. Imagine that for a UE that never has an
incoming call, LU is n ot needed at all. Hence, the static LU
scheme is cost-ineffective. A cost-effective LU scheme is such
a scheme that is dynamic enough to adapt to diverse mobility
and traffic characteristics. In addition to the prob lem of cost-
ineffectiveness, the static LU scheme is widely criticized for
the following problems, namely, ping-pong LU effect and
uneven distribution of LU signaling. The ping-pong LU effect
arises when a UE that moves back-and-forth among multiple
LAs/RAs performs LUs in the same way as in the case of
straight line movement. Under the static LU scheme, all the
LUs occur in the boundary cells of an LA/RA, causing the
uneven distribution of LU signaling. In an extreme case that
a large number of UEs simultaneously move into a boundary
cell, e.g., all the UEs in a train moving into a new LA/RA,
there is a risk of an LU signaling storm in the boundary cell
due to a large number of LUs launched in a short period. This
is an undesirable case, because it could degrade the quality-
of-service in the boundary cell and even lead to signaling
congestion.
To tackle the problems associated with th e static LU scheme
used in GSM and UMTS, LTE adopts a tracking area list
(TAL)-based LU scheme. Under this scheme, each time the
UE performs an LU in a TA, it is allocated a g roup of TAs,
called a TAL, by the Mobility Management Entity (MME), the
entity responsible for mobility management in LTE. The UE
can move within the TAs contained in the TAL without any
LU, referred to as TA update (TAU) in LTE. Only after the
UE moves into a new TA that does not belong to the TAL,
does it n eed to perform a TAU toward the MME, which will in
turn allocate the UE a new TAL that includes the new TA. In
contrast with the static LU schem e, the TAL-based LU scheme
has the following advantages. (a) It is dynamic enough to adapt
to diverse mobility and traffic characteristics. (b) It can avoid
the ping-pong LU effect through allocating the UE a TAL that
contains all the TAs among which the UE moves back-and-
forth. (c) It can evenly distribute the LU signaling among TAs
by allocating different UEs with different TALs.
The allocation of TALs determines the performance of the
TAL-based LU scheme. References [1]–[5] devote to this issue.
All these references assume that a TA contains only one cell.
IEEE ICC 2014 - Mobile and Wireless Networking Symposium
978-1-4799-2003-7/14/$31.00 ©2014 IEEE 2325