See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/287248743

Configuration of Tracking Area Code (TAC) for Paging Optimization in Mobile

Communication Systems

Article  in  Lecture Notes in Electrical Engineering · January 2014

DOI: 10.1007/978-3-642-41671-2-9

CITATIONS READS

2 10,811

5 authors, including:

Hyung-Woo Kang Seok-Joo Koh

Kyungpook National University Kyungpook National University
11 PUBLICATIONS   56 CITATIONS    163 PUBLICATIONS   1,150 CITATIONS   

SEE PROFILE SEE PROFILE

Some of the authors of this publication are also working on these related projects:

IoT-based IVI Devices and Services Management View project

VLC-based IoT Devices and Services Management View project

All content following this page was uploaded by Seok-Joo Koh on 18 March 2016.

The user has requested enhancement of the downloaded file.

 Configuration of Tracking Area Code (TAC) for Paging
Optimization in Mobile Communication Systems

Hyung-Woo Kang1,*, Woo-Ju Kim1, Seok-Joo Koh1,

Hyon-Goo Kang2, and Jung-Bae Moon2
1
Kyungpook National University, Daegu, Korea
{hwkang0621,kachukun}@gmail.com, sjkoh@knu.ac.kr
2
SK Telecom Acess Network Lab, Seoul, Korea
{hyongoo.kang,jbmoon}@sk.com

Abstract. Recently, the mobile telecommunication traffics have been rapidly

increasing due to the growth of smart phone services. In this paper, we propose
a new scheme for configuration of TAC (Tracking Area Code) to maximize the
paging success rates in the LTE-based mobile communication networks. The
proposed scheme includes the initial configuration of TAC, the local
optimization algorithm, and the re-clustering algorithm for further improvement
of the TAC configuration. From the performance analysis with real traffic data
of service provider, we see that the proposed TAC configuration scheme can
improve the paging success rates in the LTE networks, compared to the existing
TAC configuration scheme.

Keywords: LTE, Tracking Area Code, Paging, Optimization.

1 Introduction

With the popularity of smart phones, mobile communication has been rapidly changes
from 3GPP to Long Term Evolution (LTE). Recently, LTE is emerging as the new
mobile communication technology [1-3].
A mobile user typically moves around in a zone that is composed of many cells in
a mobile communication system. When a call request to a specific user arrives, the
cellular system should page the user in the cells to locate the user in the network. It is
noted that the cellular systems require efficient methods to find a specific mobile user
in the paging process [4-6]. In particular, the paging success rate is a very important
factor in the design of the paging areas. A paging area is defined by a Tracking Area
Code (TAC).
In this paper, we propose a new scheme for configuration of TACs to improve the
paging success rates in the paging process. A TAC consists of a group of cells to
which a paging signal is broadcast in the paging process. The proposed TAC
configuration scheme can be used to increase the paging success rate and to reduce
the overhead of paging traffic in the LTE networks.

*
Corresponding author.

Y.-S. Jeong et al. (eds.), Ubiquitous Information Technologies and Applications, 59

Lecture Notes in Electrical Engineering 280,
DOI: 10.1007/978-3-642-41671-2_9, © Springer-Verlag Berlin Heidelberg 2014
60 H.-W. Kang et al.

This paper is organized as follows. In Section 2, we discuss the TAC optimization

issue and how to configure the TAC in the mobile communication system.
In Section 3, we describe the proposed TAC configuration algorithm. Section 4
describes the experimental analysis and comparison with the existing scheme. Finally,
Section 5 concludes this paper.

2 TAC Optimization in Mobile Communication System

2.1 TAC Configuration

At present, most of the mobile operators configure the TAC in an arbitrary way, in
which only the topological location information of all cells are considered and the
network manager manually configures a group of cells as a TAC. TAC is defined as a
group of cells. The TAC is coded into a 2-bytes hexadecimal digit, in which the first
byte represents the area code, called the Tracking Area List. When a call request is
required for a specific user, if the user was registered with a TAC, then a paging
signal will be broadcast to all of the cells contained in the TAC.
Usually, when a mobile user is connected to a cell in network attachment phase,
the user is assigned to the TAC that the cell is associated with. If the user moves to a
new cell in the dormant mode, the TAC of the user may be changed, if the TAC of the
new cell is different from that of the old cell. Accordingly, the TAC should be
configured by considering the mobility and traffic of users in the mobile network.
The Self-Organizing Network (SON) system is used in the LTE-based mobile
communication in order to configure and manage a mobile network. For TAC
configuration, the TAC optimization S/W provides the information of optimized TAC
configuration to the SON system. Figure 1 describes the overview of SON system [7].

Fig. 1. SON system and TAC Configuration

With such TAC configuration, when a paging request to a specific user arrives, the
paging operations are performed as follows. In the first paging, the paging request
message (or signal) will be broadcast to the cells contained in the TAC. If the paging
request fails (i. e., no response to the paging request from the mobile user), then the
second paging is performed, in which the paging request will be broadcast to all of the
cells in the area. Therefore, it is important to optimize the TAC configuration so as to
 Configuration of TAC for Paging Optimization in Mobile Communication Systems 61

maximize the paging success rate of the first paging and thus to reduce the paging
traffics generated in the network.

3 Proposed TAC Configuration Algorithms

3.1 TAC Optimization Model

For TAC optimization, we first make a mathematical optimization model for TAC
configuration. Given a network area with many cells, the goal is to find an optimal
TAC configuration (mapping from a group of cells to a TAC) by considering traffic
and user mobility between cells. The Paging Success Rate (PSR) is used as the
objective function of the optimization model. The PSR represents the success
probability of the first paging.
Figure 2 shows the mathematical model of TAC optimization.

Objective function: Maximize PSR

PSR = Σk∈M Σi∈N Σj∈N λi × Pij × Xik× Xjk

Constraints: feasibility conditions

Σk∈M Xik = 1, for all i ∈ N (TAC assignment condition)
Σi∈N Xik ≤ STAC, for all k ∈ M (TAC size condition)
Σi∈N λi × Xik ≤ CTAC, for all k ∈ M (PTL condition)
dij × Xik × Xjk ≤ DTAC for all i, j ∈ N, k ∈ M (distance condition)
Xik = 1 or 0, for all i ∈ N, k ∈ M

Fig. 2. TAC Optimization Model

In the optimization model, we use the following variables and parameters:

• Pij: mobility ratio that user moves from cell i to cell j;

• λi: sum of paging traffics in cell i;
• dij: distance between cell i and j;
• N: a group of cells in TAL, with the size of n;
• M: a group of TAC in TAL, with the size of m;
• STAC: the maximum number of cells that a TAC can contain;
• CPTL: the maximum paging traffic load allowable for a TAC; and
• DTAC: the maximum distance between two cells allowable for a TAC.

On the other hand, we consider the following feasibility conditions:

• TAC assignment condition: each cell should be assigned to at least one TAC;
• TAC size condition: the number of cells assigned to a TAC cannot exceed STAC;
• Paging Traffic Load (PTL) condition: the paging traffic load for a TAC cannot
exceeds CPTL;
• Distance condition: the distance between two cells in TAC cannot exceed DTAC.
62 H.-W. Kang et al.

3.2 Proposed TAC Configuration Schemes

Based on the TAC optimization model described in Section 2, we propose the TAC
configuration algorithms in this section. The proposed algorithms consist of initial
configuration of TAC, local optimization algorithm, and re-clustering algorithm for
further improvement of the TAC configuration.
Figure 3 shows the overall sequence of the proposed algorithms.

Fig. 3. Overall Algorithm for TAC Configuration

As input to TAC configuration, we consider the mobility ratio and paging traffic
per cell and the distance between cells. The Initial Starting Point (ISP) algorithm is
used to select an initial cell for each TAC as the starting point to overall algorithm.
Based on these initial points, the Greedy Feasible Solution (GFS) algorithm is used to
obtain a feasible TAC configuration, in which all of the cells are assigned to one of
the TACs in the network. In the GFS algorithm, each cell will be assigned to a TAC
in a stepwise way so as to maximize the PSR objective function while satisfying the
feasibility conditions.
After the GFS algorithm is performed, we apply the Local Optimization Module
(LOM) algorithm to find a more optimized solution. In LOM, we try to optimize the
TAC configuration by changing the TAC of a cell to the other TAC. Finally, the
Adaptive Starting Point (ASP) algorithm is used to explore the other solution space
for further optimization, in which an initial stating cell per TAC will be re-calculated
and then the optimization algorithms (GFS and LOM) are performed again. Such
procedures are repeated until the stopping criterion is satisfied. The details of each
algorithm are described below.
 Configuration of TAC for Paging Optimization in Mobile Communication Systems 63

3.2.1 Initial Starting Point (ISP)

The ISP algorithm is the first step of overall optimization algorithm, in which an
initial cell is randomly selected for each TAC.

3.2.2 Greedy Feasible Solution (GFS)

In GFS, a feasible TAC configuration is obtained, based on the initial cells of ISP. For
each TAC, we select the cell with the largest mobility rate moving from the initial cell
among the other cells that have not assigned to any TAC. The selected cell will be
assigned to the TAC, if the feasibility conditions are still satisfied by this assignment.
GFS will be repeated until all cells are assigned to one of the TACs.

3.2.3 Local Optimization Module (LOM)

The LOM algorithm finds more optimized configuration of TAC. Based on the
configuration obtained by GFS, we try to change the TAC of a cell to another TAC. If
the PSR value is improved by the TAC change, we perform the change of TAC for
the cell. Otherwise, if the PSR value is not improved or the feasibility condition is not
satisfied, we do not change the TAC of the cell. These procedures are performed until
no improvement of PSR is made for all cells.

3.2.4 Adaptive Starting Point (ASP)

After LOM is completed, the ASP algorithm is used to explore much more solution
space in the optimization process. For each TAC given by LOM, we find the ‘center’
cell, which is obtained by calculating the distances of the other cells from a candidate
center cell within the TAC. We define the center cell as the cell with the minimum
distance to the other cells.
In ASP, a center cell is calculated for each TAC and it is used as an initial point to
GFS and LOM. This ASP algorithm is performed until a pre-specified stopping rule
(e. g., totally 10 times)

4 Experimental Results and Analysis

4.1 Test Network Environment

In the experiments, we use the real-world data of network topology, user traffic, and
the mobility rates to neighboring cells, which are based on the information of SK
Telecom in Korea. The proposed TAC configuration scheme was applied to a target
area, and we calculate the PSR value.
For all of the experiments, the default values for parameters are set as follows:
N is different for TAL, M = 11, STAC = (N / M) + 10, CTAC = 2100, DTAC = 1/2 * the
maximum distance between cells in TAL.

4.2 Results and Discussion

In experiments, the distance between two cells is calculated by using the Euclidean
distance. The probability of moving from a cell to another is assumed to be 20%,
64 H.-W. Kang et al.

which is represented as α. That is, a mobile user remains within the current cell with a
probability of 0.8. Then, the Paging Success Probability (PSP) can be calculated as
follows, as shown in Figure 4.

Paging Success Probability (PSP)

PSP = (1-α) + α × PSR / λ,
where λ = Σi∈N λi , α = the probability of moving from a cell to another

Fig. 4. Calculation of PSP

Table 1 shows the comparison of the existing and proposed schemes for TAC
configuration, which is applied to the target area that is coded 19.

Table 1. Comparing the Existing and Optimization TAC configuration of target area

Existing TAC Configuration Optimized TAC Configuration

TAC # of cells maximum PTL # of cells maximum PTL
Index distance distance
1900 51 12624 684 64 9565 968
1901 56 4041 829 63 5990 688
1902 48 4629 671 64 9624 264
1903 44 2714 680 64 4489 710
1904 72 6325 1083 62 6436 524
1905 68 6789 943 64 5673 787
1906 53 6138 494 65 6917 904
1907 55 3797 569 61 6236 584
1908 5 4002 60 57 6795 685
1909 5 1152 73 71 8115 401
190A 193 7995 571 5 4002 60

In the table, we can see that the original (existing) TAC configuration is composed
of a total of 642 cells and is made up with a number of 11 TACs. It is noted that the
TAC 190A contains 193 cells, which violates the feasibility condition of STAC,
whereas the TAC 1908 and 1909 have only 5 cells. This implies that the existing TAC
configuration is not balanced in the viewpoint of the TAC size. This leads to a lower
Paging Success Rate. The standard deviation of PTL is 91268. Overall, the existing
TAC configuration gives the Paging Success Probability (PSP) of 89.39.
On the other hand, the optimized TAC configuration assigns the total 642 cells to
11 TACs. Most of the TACs have the TAC size of a minimum 61 cells and up to 71
cells, except for TAC 190A. This is because the TAC 190A is a subway TAC, and
thus this TAC cannot be further optimized. Note that all TACs give the nearly equal
size, except the subway TAC, in the proposed optimization scheme. The standard
deviation of PTL is 67268, which is lower than the existing configuration. The PSP of
the proposed scheme is calculated as 91.58, which is greater than the existing scheme.
 Configuration of TAC for Paging Optimization in Mobile Communication Systems 65

Figure 5 compares the PSP values of target area for the existing and proposed
schemes as phases of optimization. From the figure, we can see that the proposed
scheme provides higher PSP than the existing scheme. That is, the proposed
optimization scheme improves the PSP compared with the original configuration.

92
Existing TAC Configuration
Proposed TAC Configuration
91.5

91
PSP(%)

90.5

90

89.5

89
0 2 4 6 8 10 12
Phases of Optimization

Fig. 5. Paging Success Probability of target area

Figure 6 and 7 shows the distribution of TACs for the target area, in the viewpoint
of physical location, for the original and optimized TAC configuration schemes.
5
x 10
4.2
TAC0
TAC1
4
TAC2
TAC3
3.8 TAC4
TAC5
3.6 TAC6
TAC7
TAC8
3.4
TAC9
TAC10
3.2

2.8
4.16 4.17 4.18 4.19 4.2 4.21 4.22 4.23 4.24
6
x 10

Fig. 6. Existing TAC configuration

5
x 10
4.2
TAC0
TAC1
4
TAC2
TAC3
3.8 TAC4
TAC5
TAC6
3.6 TAC7
TAC8
3.4 TAC9
TAC10

3.2

4.16 4.17 4.18 4.19 4.2 4.21 4.22 4.23 4.24

6
x 10

Fig. 7. Proposed TAC configuration

From the figure, we can see that the proposed scheme gives more compact and
balanced TAC configurations that the existing scheme. This is because the proposed
scheme applies the feasibility conditions for TAC configuration, and the PSP is
maximized by including the neighboring cells into the same TAC.
66 H.-W. Kang et al.

5 Conclusions

In this paper, we presented a new scheme for configuration of TAC to maximize the
paging success rates in LTE-based mobile communication networks. The proposed
scheme is composed of a set of sub-algorithms: Initial Starting Point (ISP), Greedy
Feasible Solution (GFS), Local Optimization Module (LOM), and Adaptive Starting
Point (ASP). In the TAC optimization, we also considered the feasibility conditions to
get a balanced configuration of TACs.
By the experimentations for real-world data of network topology and user traffic,
the proposed optimization scheme is compared with the original existing scheme in
the perspective of paging success probability. From the results, we can see that the
proposed scheme provides more optimized TAC configurations than the existing
scheme by maximizing the paging success probability. It is also noted that the
proposed scheme gives a more compact and balanced TAC configuration than the
existing scheme by using the feasibility conditions. It is expected that the proposed
TAC optimization scheme can be used in the real-world mobile communication
networks to maximize the paging success rates and to reduce the paging traffic load.

Acknowledgment. This research was supported by the Basic Science Research

Program of NRF(2010-0020926), and by the MSIP support program of NIPA(NIPA-
2013-H0301-13-2004).

References
1. Astély, D., et al.: LTE: the evolution of mobile broadband. IEEE Communications
Magazine 47(4), 44–51 (2009); 44-51.3GPP TS 36.300, Evolved Universal Terrestrial
Radio Access (E-UTRA) and Evolved Universal Terrestrial Radio Access Network (E-
UTRAN): Overall Description Stage 2 (2008)
2. Ulvan, A., Bestak, R., Ulvan, M.: The study of handover procedure in LTE-based femtocell
network. In: Wireless and Mobile Networking Conference, WMNC (2010)
3. Modarres Razavi, S.: Tracking Area Planning in Cellular Networks: Optimization and
Performance Evaluation. Dissertation of Linköping (2011)
4. Bar-Noy, A., et al.: Paging mobile users in cellular networks: Optimality versus complexity
and simplicity. Theoretical Computer Science (2012)
5. Lyberopoulos, G.L., et al.: Intelligent paging strategies for third generation mobile
telecommunication systems. IEEE Transactions on Vehicular Technology 44(3), 543–554
(1995)
6. 3GPP TS 36.902, Evolved Universal Terrestrial Radio Access Network (E-UTRAN): Self-
configuring and self-optimizing network (SON) use cases and solutions (2009)

View publication stats