10

International Journal of Future Generation Communication and Networking
Vol. 4, No. 4, December, 2011

Improved Vertical Handover Decision Algorithm for UMTS-
WLAN

N. P. Singh and Brahmjit Singh
Electronics and Communication Engineering Department
National Institute of Technology, Kurukshetra – 136119
niraj_pratap_singh@yahoo.com, brahmjit.s@gmail.com

Abstract

Next generation wireless systems consist of heterogeneous wireless networks. Wireless
LAN (WLAN) and cellular mobile radio systems will be important part of heterogeneous
networks. Universal mobile telecommunication system (UMTS) provides wide-area coverage,
high mobility but relatively low data rate. WLAN is designed for low-range, high/medium
data rate access, low mobility and can be used as a complement to larger cellular mobile
radio systems. In this paper, we consider a simple heterogeneous system of WLAN and UMTS
cell. Vertical handover plays a vital role in ensuring quality of service in heterogeneous
networks. In this paper, we proposed a vertical handover algorithm for an integrated
cellular/wireless LAN system. The design goal is to provide the user the best available quality
of service (QoS) at any time. Moreover, connections are handed over independently between
UMTS and WLAN, aiming at seamless service continuity. Here we have taken handover rate,
probability of outage and probability of handover as the performance metrics. Simulation
results show that adaptive vertical handover algorithm improves the performance.

Keywords: Vertical handover algorithm, Heterogeneous wireless network, Probability of
handover and probability of outage

1. Introduction
In recent years, wireless networks have emerged and played key roles in modern
telecommunications. General Packet Radio Service (GPRS) enables Global System for
Mobile Communications (GSM) users to access IP networks, such as the internet, and third-
generation (3G) cellular systems, such as Universal Mobile Telecommunications System
(UMTS) and cdma2000. In the evolution of cellular networks, one problem is that the data
rates in these networks are limited. The data rate in GPRS is up to 144 kb/s, in UMTS it is
from 384 kb/s to 2 Mb/s. One widely acceptable way to deal with this limitation is the
complementary use of wireless local area network (WLAN) [1,2] technology in hot spot areas
(e.g., business centers, airports, hotels, and campuses) since WLAN can provide up to 54
Mb/s data rate with coverage of a few thousand square meters around a single access point.
Numerous interworking architectures between 3G cellular and WLAN systems have been
proposed in the technical literature. The most important objective is to enable 3GPP cellular
network subscribers to access WLAN service. A WLAN terminal equipped with a 3GPP
subscriber identity module (SIM)/universal SIM (USIM) smart card can access both the 3GPP
subscriber database and WLAN. In this article, we focus on handover between cellular system
and WLAN. The major challenge in selecting a network is to decide the most favorable trade-
off among user preference, service application, and network condition. Network selection is

113


typically carried out in three steps. The first step is collecting the necessary information that
has some impact on the final decision. The information might be user preference, service
application, and network condition. The second step is using the collected information as
inputs to a certain handover algorithm that aims to keep the user always best connected
(ABC) [3]. The meaning of ABC is that the user is not only connected but also enjoys the best
achievable QoS at any time and any place. The last step is making a decision according to the
algorithm’s output. A number of researchers proposed network selection algorithms in the
literature [4, 5]. The most important is received signal strength (RSS) based algorithm that
uses the threshold and hysteresis values as input parameters [6-10]. ABC is the design goal,
which means the best network is always selected. One of the major challenges for seamless
mobility is the availability of simple and efficient vertical handover scheme, which is the
decision for a mobile node to handover between different types of networks, cellular wireless
network and wireless local area networks (WLANs). Vertical handover schemes will play a
main role in the IEEE 802.21 standard and shall pave the road for emergence of 4G overlay
multi-network atmosphere.
In this article, we consider UMTS and WLAN as network alternatives. The proposed
technique would, however, be applicable to systems with more heterogeneity (e.g.,
cdma2000-WLAN, GPRS, WLAN), and this will be the subject of future work.
The remainder of this paper is organized as follows: In Sect. 2, we briefly introduce the
vertical handover. System model for UMTS-WLAN Handover is given in Sect. 3. The
proposed handover decision algorithm is given in Sect. 4 followed by performance metrics in
sect.5 and result and discussion in Sect. 6. Finally, we draw our conclusions in Sect. 7.

2. Overview of Vertical Handover
For ubiquitous coverage, trend is to integrate complementary wireless technologies with
overlapping coverage. The state of the art mobile devices in beyond third generation (B3G)
networks will be equipped with multiple network interfaces for accessing different networks.
Users will expect to continue their connections without any disruption when they move across
networks using different technologies. This process in wireless technology is known as
vertical handover. In heterogeneous wireless networks, cellular networks using one access
technology and WLANs using different technology will coexist. In such situation, transfer of
ongoing connection from base station (BS) of cellular network to access point (AP) of WLAN
is possible only by vertical handover mechanism. Vertical handover is more complex process
than horizontal handover due to additional parameters. A decision for vertical handover may
depend on several factors (such as network bandwidth, load, coverage, cost, security, price,
speed, power consumption and QoS) related to the network to which the mobile node is
already connected and to the one that it is going to handover. Handover is needed or not is
indicated by handover metrics. New handover metrics should be considered along with signal
strength for next generation heterogeneous wireless environment.
Several parameters have been proposed in the research literature for use in the vertical
handover algorithms. Some of them have been briefly explained below [11, 12, 13, 14, 15,
and 16].
Received Signal Strength (RSS) is generally used for handover because it is easy to
measure and is directly related to the quality of service. RSS is directly associated with the
distance between mobile station (MS) and its point of attachment. Nearly all horizontal
handover algorithms use RSS as the main decision criterion, and RSS is an important
criterion for vertical handover decision (VHD) algorithms.

114


Network Conditions related parameters like traffic, available bandwidth, network latency,
and packet loss may need to be considered for effective network usage. Networks information
can be useful for load balancing across different networks.
Data Rate is a measure of quality of service of the network. It is a good indicator of the
traffic conditions in the access network and is especially important for delay-sensitive
applications.
Power Consumption is very important criterion because wireless devices operate on
limited battery power. In such situations, it would be preferable to hand over to a point of
attachment with low power consumption. If battery of MS is low then it is advisable to hand
over to UMTS from WLAN.
Cost of Service: For different networks, there would be different charging policies;
therefore, the cost of a network service will probably influence the handover decisions.
Security: Confidentiality or integrity of the transmitted data is very important for some
applications. Therefore, a network with higher security level may be chosen over another one,
which would provide lower level of data security.
User Preferences: A user’s personal preference is one of the most essential criteria in
vertical handover algorithm. For high data rate application, user will like to handover to the
network with high data rate and if user can not afford high cost then certainly user will move
to the cheaper network.

3. System Model
We consider a UMTS and WLAN cells adjacent to each other as shown in fig.1. The
simplest mobility model is the user traveling in a straight line at a constant speed have been
considered and assumed that MS moves with constant speed from the UMTS BS along a
straight line towards WLAN access point. D is the distance between the BS and AP. MS
samples the pilot signal strength measurements at regular distance intervals d = kds, where ds
is the sampling distance (ds = 1m) and k, an integer with k ∈ [0, D/ds]. Both, base station
and access point are assumed located in the center of the respective cell and operating at the
different transmitting powers.

UMTS Cell
WLAN Cell
MS
BS AP

Fig. 1 Vertical Handover

Received Signal Strength (RSS) at MS consists of path loss attenuation with respect to
distance, Shadow fading, and fast fading [17, 18]. The propagation attenuation can be
expressed as the product of the  th power of distance and a lognormal component
representing shadow fading losses [19]. These represent slowly varying variations even for

115


users in motion and apply to both reverse and forward links. For MS at a distance d from BS
or AP, attenuation can be expressed as.

 (d ,  )  d  1010 (1)
Where  is the dB attenuation due to shadowing, with zero mean and standard deviation  .
Alternatively, the losses in dB are
 (d ,  )[dB]  10 log d   (2)
Where,  (eta) is path loss exponent and d represents BS to MS or AP to MS separation in
kilometers. The autocorrelation function between two adjacent shadow-fading samples is
shown in [20]. Let the transmitted power of BS or AP is Pt, the signal strength received by
MS can be written as
S ( d )  P   ( di ,  )
t (3)

The rectangular averaging window can be used to alleviate the effect of shadow
fading according to the following formula [21, 22]
1 N 1
ˆ
S (k )   S (k  n)Wn
N w n 0
(4)

Where ˆ
is the averaged signal strength and S i is the signal strength before
Si
averaging process. W n is the weight assigned to the sample taken at the end of (k  n) th
interval. N is the number of samples in the averaging window
N 1
Nw = W
n 0
n . In the case of rectangular window W n = 1 for all n.

Shadow fading in the present work is modeled as follows [23, 24]
 (k )   (k  1)   i (1   2 )W (0,1) (5)

Where  (rho) is correlation coefficient,  is standard deviation of shadow fading and
W (0,1) represents truncated normal random variable.
Shannon's Theorem gives an upper bound to the data rate of a link, in bits per second (bps),
as a function of the available bandwidth and the signal-to-noise ratio of the link. The Theorem
can be stated as [25]:
 S
D  W  log 2 1   (6)
 N
Where D is the maximum theoretical channel data rate, W is the bandwidth of the link,
S
is the ratio of average signal power to average noise power.
N

4. Vertical Handover Algorithm
The integration of cellular network and WLAN is the trend of the beyond 3G mobile
communication systems, and mobile station will handover between the two kinds of

116


networks. The received signal strength (RSS) is the dominant factor considered in the case of
handover. However, signal power level of different network is dissimilar. Therefore, in the
case of vertical handover signals from different networks can not be compared like horizontal
handover. Hence, horizontal handover algorithms can not be used in heterogeneous network.
In this paper, new vertical handover algorithm based upon normalized power has been
proposed. MS will be handed over to WLAN if SWn exceeds SU n in following way.

SWn  SUn (7)

Where, SU n and SWn are the normalized signal power received by the MS from the UMTS
and the WLAN respectively.
Handover to UMTS will be made if SU n exceeds SWn in the following way.

SUn  SWn  hys margin (8)

Performance of vertical handover can be enhanced by using distance based adaptive
hysteresis margin shown below.
   X 4  
 
hys margin  max 20 1     , 0  (9)
 R 
 
   
Where X is the distance of MS from WLAN access point and R is the radius of WLAN
cell.
In heterogeneous networks, data rate is very important factor for selection of network. So
depending upon user preference UMTS to WLAN handover may take place if DWLAN exceeds
DUMTS in following way.
DWLAN  DUMTS (10)

Where DWLAN and DUMTS are the maximum achievable theoretical data rate in WLAN and
UMTS respectively.
Vertical handover decision is shown in given in Fig.2

5. Performance Metrics
This section discusses performance metrics used to evaluate soft handover algorithm:
Handover rate: Reducing the number of handovers is usually preferred, as frequent
handovers would cause wastage of network resources. Ping-pong effect in handovers should
be minimized.
Probability of Handover: Probability of handover at the boundary should be very high for
improving the performance of the system. In ideal case, probability of handover should be
one at the boundary and handover area should be small.
Probability of outage: Here probability of outage refers to the fact that MS is not
connected to any network. Probability of outage is used to describe the quality of service.

117


6. Results and Discussion
In this section, handover rate, probability of outage and probability of handover have been
computed as the function of different system parameters and characteristic parameters of
radio propagation environment. Numerical results for this performance metric are obtained
via computer simulation for the system parameters indicated in Table 1. System simulation is
performed in Matlab 7.0. To obtain probability of outage for each system parameter setting,
10000 runs of the simulation program are performed.

Fig.2 Vertical Handover between UMTS and WLAN

Fig. 3 shows the plot of received signal strength from UMTS base station and WLAN
access point with respect to distance. Here same environmental condition has been considered
for both networks.

118


Fig. 4 shows the handover rate versus hysteresis margin plot. This plot shows that
handover rate decreases with the increase of hysteresis margin. Unnecessary handovers can
be reduced by increasing hysteresis margin. Nevertheless, large hysteresis margin increases
the probability of outage and handover delay. Handover delay refers to the duration between
the initiation and completion of the handover process. Handover delay is related to the
hysteresis margin, and reduction of the handover delay is especially important for delay
sensitive voice or multimedia sessions.
Therefore, hysteresis margin should not be very high nor it should be very low. Simulation
with adaptive handover margin shows handover rate equal to 2.26. This value is neither very
low nor very high. This result shows the advantage of adaptive hysteresis margin over fixed
hysteresis margin.

Table 1.System Parameters for System Simulation
D = 1100 m Distance between UMTS base station and WLAN access point

Smin=-90dBm Minimum acceptable signal to maintain the call
ds = 1m Sampling distance
PtUMTS = 30dBm UMTS Base station transmitter power
PtWLAN = 0dBm WLAN access point transmitter power
RUMTS=1000m Radius of UMTS cell
RW=100m Radius of WLAN cell
WUMTS=5MHz Bandwidth of UMTS
Ww=20MHz Bandwidth of WLAN

40
Signal from UMTS
20 Signal fromWLAN

0
Signal power (dBm)

-20

-40

-60

-80

-100

-120
0 100 200 300 400 500 600 700 800 900 1000 1100
Distance between UMTS BS and WLAN access point (m)

Fig.3 Received Signal Power versus Distance for given   0.5,   8dB,  3.6

Probability of handover is very important factor for seamless handover. Fig. 5 shows the
probability of handover for different hysteresis margin in the case of MS moving from UMTS
to WLAN. Handover should occur near the boundary for better performance of network.
Ideally, Probability of handover should be one at the boundary and zero when MS is away

119


from the boundary for given simulation model. However, practically it is not possible. Hence
vertical handover should be designed in such a way that rate of increase of probability of
handover should be very high near the boundary. Plot shows that adaptive hysteresis margin
gives best performance compared to fixed hysteresis margin. Adaptive handover algorithm
will decrease unnecessary handover when MS is away prom boundary and handover will take
place only when MS is very near to boundary.

4.5

4

3.5
Handover rate

3

2.5

2

1.5
0 2 4 6 8 10 12 14 16 18 20
Hysteresis margin (dB)

Fig.4 Handover Rate versus Hysteresis Margin for given   0.5,   8dB,  3.6 .

1

0.9 hys margin=Adaptive
Probability of handoff from UMTS to WLAN

hys margin=0dBm
0.8 hys margin=6dBm
hys margin=12dBm
0.7

0.6

0.5

0.4

0.3

0.2

0.1

0
0 100 200 300 400 500 600 700 800 900 1000 1100
Distance between UMTS BS to WLAN access point (m)

Fig.5 Probability of Handover from UMTS to WLAN versus Distance for
given   0.5,   8dB,  3.6

Fig. 6 shows the probability of outage versus distance plot. This plot shows that probability
of outage increases with the increase of hysteresis margin. With large hysteresis margin,
probability of outage is high but at the same time, probability of handover is very high near
the boundary. In the case of adaptive handover algorithm, probability of outage is not very
high but probability of handover at boundary is high. Therefore, it is better to use adaptive
handover algorithm for enhancing vertical handover performance.

120


Fig. 7 shows the variation of maximum theoretical data rate with distance for both
networks UMTS and WLAN. This plot shows that data rate of WLAN is very high near the
access point and decreases with distance. Data rate in UMTS is low compared to WLAN but
coverage area is more. Hence, for low data rate application and high mobility user UMTS is
better network. High mobility user will not like to be connected with WLAN due to its low
coverage area and frequent handover will increase network load. 4G networks will use
multimodal MS and in such case, handover will be based upon user preference. Hence, for
high data rate application and low cost MS will handover to WLAN.

0.09
hys=Adaptive
0.08 hys=6dBm
hys=12dBm
0.07

0.06
Probability of outage

0.05

0.04

0.03

0.02

0.01

0
0 100 200 300 400 500 600 700 800 900 1000 1100
Distanc between UMTS base station to WLAN access point (m)

Fig.6 Probability of Outage versus Distance.   0.5,   8dB, N  20

9
x 10
3.5
UMTS Data rate
WLAN Data rate
3
Maximum Theoretical Data rate (bps)

2.5

2

1.5

1

0.5

0
0 100 200 300 400 500 600 700 800 900 1000 1100
Distance btween UMTS base station to WLAN access point (m)

Fig.7 Data Rates of UMTS and WLAN versus Distance

7. Conclusion
In this paper, a novel handover scheme has been presented for the vertical handoff between
UMTS and WLAN for guaranteeing the best QoS by selecting the most suitable network
whereas preventing frequent handovers. It is based on absolute signal strength, distance and
data rate. Results show that proposed algorithm outperforms the conventional algorithm based

121


on signal strength measurements. Simulation result shows that adaptive handover algorithm
outperforms the handover algorithms with fixed hysteresis margin. The present study may be
significant in enabling the deployment of seamless networks. The approach presented in this
paper may be applied for traffic load sharing between UMTS and WLAN. This can be
achieved by adapting the threshold parameter settings on distance and signal strength
measurements as functions of traffic load conditions. The MS can be directed from serving
network to another if the first network becomes heavily loaded and vice -versa. This will
increase the overall traffic carrying capacity of the integrated network. Simultaneously, it will
improve the QoS by making more radio and network resources available to the users
irrespective of the network, it is initially associated with. Another issue deserving further
explanation is the adaptation of the algorithm design parameters to varying system parameters
and their optimization.

References
[1] Y. Kim, B. J. Jeong, J. Chung, et al., “Beyond 3G: vision, requirements, and enabling technologies,” IEEE
Communications Magazine, vol. 41, no. 3, pp. 120–124, 2003.
[2] Qingyang Song & Abbas Jamalipour, “Network Selection in an Integrated Wireless LAN and UMTS
Environment using Mathematical Modeling and Computing Techniques”, IEEE Wireless Communications •
June 2005, pp 42-48.
[3] Eva Gustafsson and Annika Jonsson, “ALWAYS BEST CONNECTED” IEEE Wireless Communications •
February 2003, pp 49-55.
[4] W.K. Liao, Y.C. Chen, Supporting vertical Handover between universal mobile telecommunications system
and wireless LAN for real-time services, IEEE IET Communications 2 (1) (2008) 75–81.
[5] Dina S. Deif, Hesham EI-Badawy, and Hadya EI-Hennawy , “Toplogy Based Modeling and Simulation of
UMTS-WLAN Wireless Heterogeneous Network” IEEE international conference on wireless and optical
communication network (WOCN) 2010,pp. 1-5.
[6] H.S. Park, H.S. Yoon, T.H. Kim, J.S. Park, M.S. Duo, J.Y. Lee, Vertical Handover procedure and algorithm
between IEEE802.11 WLAN and CDMA cellular network, Mobile Communications (2003) 103–112.
[7] M. Liu, Z.-C. Li, X.-B. Guo, H.-Y. Lach, Design and evaluation of vertical Handover decision algorithm in
heterogeneous wireless networks, in: Proceedings of the 2006 IEEE International Conference on Networks
(ICON’06), Singapore, September 2006, pp. 1–6.
[8] S. Mohanty, I.F. Akyildiz, A cross-layer (layer 2 + 3) Handover management protocol for next-generation
wireless systems, IEEE Transactions on Mobile Computing 5 (10) (2006) 1347–1360.
[9] X. Yan, N. Mani, Y.A. S_ekerciog˘lu, A traveling distance prediction based method to minimize unnecessary
Handovers from cellular networks to WLANs, IEEE Communications Letters 12 (1) (2008) 14– 16.
[10] B.-J. Chang, J.-F. Chen, Cross-layer-based adaptive vertical Handover with predictive RSS in heterogeneous
wireless networks, IEEE Transactions on Vehicular Technology 57 (6) (2008) 3679–3692.
[11] Xiaohuan Yan, Y. Ahmet Şekercioğlu, Sathya Narayanan “A survey of vertical Handover decision algorithms
in Fourth Generation heterogeneous wireless networks” computer Networks, Volume 54, Issue 11, 2 August
2010, Pages 1848-1863
[12] J. McNair, F. Zhu, Vertical Handovers in fourth-generation multinetwork environments, IEEE Wireless
Communications 11 (3) (2004) 8–15.
[13] Meriem Kassar, Brigitte Kervella, Guy Pujolle “An overview of vertical Handover decision strategies in
heterogeneous wireless networks” Computer Communications, Volume 31, Issue 10, 25 June 2008, Pages
2607-2620.
[14] Johann Márquez-Barja, Carlos T. Calafate, Juan-Carlos Cano, Pietro Manzoni “An overview of vertical
Handover techniques: Algorithms, protocols and tools” Computer Communications, In Press, Corrected
Proof, Available online 4 December 2010
[15] Mark Stemm and Randy H. Katz “Vertical Handovers in wireless overlay networks” Mobile Networks and
Applications, 1998, Volume 3, Number 4, Pages 335-350
[16] Nasser, N.; Hasswa, A.; Hassanein, H. “Handovers in fourth generation heterogeneous networks”
Communications Magazine, IEEE Volume: 44, Issue: 10, Publication Year: 2006, Page(s): 96 – 103

122


[17] N. Zhang, J.M. Holtzman, “Analysis of CDMA Soft-Handover Algorithm”, IEEE Transactions on Vehicular
Technology, Vol. 47,No. 2 May 1998, pp. 710-714.
[18] B. Homnan, V. Kunsriruksakul, W. Benjapolakul, “A Comparative Evaluation of Soft Hnadoff between S-
95A and IS-95B/cdma2000”, Proc. IEEE Vehicular Technology, 2000, pp. 34-37.
[19] V.K.Garg, J.E. Wilkes, “Principles and Applications of GSM”, Prentice Hall PTR, Upper Saddle River, NJ
07458.
[20] M.Gudmundson, “Correlation Model for Shadow Fading in Mobile Radio Communication System,”
Electronics Letters., vol.27, November 1991, pp.2145-2146.
[21] G.E. Corazza D. Giancristofaso, F. Santucci, “Characterization of Handover Initiation in Cellular
Mobile Radio Networks”, IEEE Technology Conference, 1994, pp. 1869 -1872.
[22] Brahmjit Singh “An improved Handover Algorithm based on signal strength plus distance for
interoperability in mobile cellular networks” Wireless Personal Communication (2007), 43:879-887.
[23] Celal Çeken , Serhan Yarkan , and Hüseyin Arslan “Interference aware vertical Handover decision algorithm
for quality of service support in wireless heterogeneous networks” Computer networks 54, 2010,726-740
[24] N.P.Singh & Brahmjit Singh “Performance Enhancement of Cellular Network Using Adaptive Soft
Handover Algorithm” Wireless Personal Communications, Online First™, 26 May 2010.
[25] Theodore S. Rappaport, “Wireless Communications: Principles & Practice” Prentice Hall, 2002. 2nd edition.

Authors

N. P. Singh received B.E. & M.E. degree in Electronics and
Communication Engineering from Birla Institute of Technology, Mesra
Ranchi, India in 1991 & 1994 respectively and presently pursuing PhD
degree in the area of Handover algorithms for 3G and B3G mobile
networks. Currently he is Assistant professor in the department of
Electronics and Communication at National Institute of technology
Kurukshetra, India. The current focus of his research interests is on radio
resource management, interworking architectures design, mobility
management for next generation wireless network

Brahmjit Singh received B.E. degree in Electronics Engineering from
Malviya National Institute of Technology, Jaipur in 1988, M.E. degree in
Electronics and Communication Engineering from Indian Institute of
Technology, Roorkee in 1995 and Ph.D. degree from Guru Gobind Singh
Indraprastha University, Delhi in 2005 (India). Currently, he is Professor,
Electronics and Communication Engineering Department, National
Institute of Technology, Kurukshetra, India. He teaches post-graduate
and graduate level courses on wireless Communication and CDMA
systems. His research interests include mobile communications and
wireless networks with emphasis on radio resource and mobility
management. He has published 65 research papers in
International/National journals and Conferences. Dr. Brahmjit Singh
received the Best Research Paper Award from The Institution of
Engineers (India) in 2006.

123


124

10

More Related Content

What's hot (17)

Viewers also liked (15)

Similar to 10 (20)

10