International Journal of Scientific & Engineering Research Volume 2, Issue 10, October-2011 1
ISSN 2229-5518
Drop Call Probability Factors in Cellular
Networks
Nathaniel S. Tarkaa, Joseph M. Mom, Cosmas I. Ani
Abstract— Of the numerous performance metrics applied to cellular telephone systems, probably none is more important for customer satisfaction than the system drop call rate. Customers are more sensitive to call dropping than to call blocking at initiation. Proper system design and operation involve keeping the drop call rate as low as possible. Call/packet dropping refers to the event described as the termination of calls in progress before either involved party intentionally ends the call. There are numerous drop call causes in cellular networks with majority of them occurring in the Um interfaces mainly due to lack of radio resources created by electromagnetic causes and user mobility (i.e. handover). Another important contributor of drop call rate is the traffic load in which, the call arrival rate and holding time play significant roles. Drop call probability is defined as the probability that a call is terminated due to one or all of the above-mentioned causes and is basically estimated from drop call rate by applying the Poisson probability distribution function. Drop- call probability has been the subject of several network performance studies and a major contributor to service optimization in established cellular networks. In this paper, we present an overview of drop-call probability factors in cellular networks. Moreover, some of the factors have been analyzed to study the trends in relation to an operative GSM network and the results are discussed.
Index Terms— drop call probability, holding time, packet dropping, quality of service (QoS), traffic load.
—————————— ——————————
rop-call probability is one of the key performance indicators (KPI) used by various mobile phone service operators for measuring quality of service
(QoS). It generally refers to the phenomenon of call/packet dropping in both voice and data networks. Call/packet dropping refers to the event described as the termination of calls in progress before either involved party intentionally ends the call.
Wireless networks involve radio and wire-line links as well as switching hardware and software, and data base operations. However, drop call rate is mainly determined by the radio resources in the network. These resources translate mainly into the plethora of radio channels all of which share a common bandwidth through a process known as frequency reuse.
Call dropping is caused by lack of available radio channels which in turn may be caused by propagation factors such as distance losses, path loss, multipath fading, shadowing and RF interference [3], [4], [5], [6]. Other channel capacity varying factors include handover and service prioritization [7], [11].
As the signal travels from the transmitting antenna to the receiving antenna, it loses strength. This may be due to the phenomenon of path loss, or it may be due to the Rayleigh effect [4]. Rayleigh (or Rician) effect is due to the
————————————————
Nathaniel S. Tarka is a lecturer with Dept of Electrical and Electronics
Engineering,University of Agriculture, Makurdi-Nigeria.
Joseph M. Mom is is a lecturer with Dept of Electrical and Electronics
Engineering,University of Agriculture, Makurdi-Nigeria
fast variation of the signal level both in terms of
amplitude and phase between the transmitting and receiving antennas when there is no line of sight. Rayleigh fading can be divided into two kinds: multipath fading and frequency-selective fading [4], [5], [6].
Shadowing is caused by diffraction which is a phenomenon that takes place when a radio wave strikes a surface and changes its direction of propagation owing to the inability of the surface to absorb it [4]. The loss due to diffraction depends upon the kind of obstruction in the path which may be high buildings or hills. It is known as shadowing because the mobile receiver is in the shadow of these structures.
Handover or handoff is the mechanism that transfers an ongoing call from one cell to another as a user moves through the coverage area of a cellular system. If and when a Mobile Terminal (MT) moves, it is quite possible that the currently serving Base Station Subsystem (BSS) may no longer be able to provide reasonable quality of service as compared to some other BSS. Rather than dropping the service to this MT, the currently serving Mobile Switching Centre (MSC) may decide to hand over this service to some other better serving BSS or in some cases to another MSC. Occasionally, this handover process fails and the call drops [8, 10]. The minimization of drop call rate needs efficient schemes for making handover requests at the right place at the right time based on the propagation environment.
Drop-call probability is also influenced by traffic intensity parameters such as call arrival rate and call duration [1], [
2]. Theoretically, drop-call probability is defined as traffic lost/traffic offered [12].
The rest of the paper is organized as follows: Section 2
.+2348059121411. E-mail: josseffmom@yahoo.com
Cosmas I. Ani is a lecturer with Dept of Electronic Engineering,
University of Nigeria, Nsukka.
IJSER © 2011
International Journal of Scientific & Engineering Research Volume 2, Issue 10, October-2011 2
ISSN 2229-5518
discusses the influence of radio resources availability and utilization on drop-call probability. This is followed by a discussion about the effects of traffic parameters in section 3. Section 4 is about estimation of drop-call probability. In section 5, some experimental results and analysis with data obtained from an operating GSM network are presented. Lastly in section 6 is the conclusion.
Erlangs, N = available number of channels
This shows that drop-call probability decreases as the
number of channels increases.
The channel utilization depends on the total traffic, t and is given by [12]:
In many network environments, the available network capacity varies unpredictably with time [11]. For
p = Traf ic Intensity
Number of channels
(2)
example, in a reservation-based network with multiple priority levels, high priority calls such as video conferences or emergency services may take precedence over ordinary traffic. The network capacity available for low priority traffic thus varies with time based on high priority traffic demands. In wireless networks, capacity variation arises from the mobility of users (e.g., handovers) and the time-varying characteristics of the wireless propagation environment [7]. The patterns of
In this section, the basic principles of some of the key
drop-call probability related traffic parameters will be discussed. They are as follows:
Call arrival rate, t , refers to the traffic offered expressed as the number of call attempts per unit time [12], which in this case is given as:
wireless interference for the active connections may dynamically change the available capacity for these connections [11]. Such networks are generally referred to
λt = Number of Call Attempts/busy hour
14400 seconds/busy hour
(3)
as stochastic capacity networks.
Cellular systems use one or more of four different techniques of access (TDMA, FDMA, CDMA, and SDMA) [11]. In TDMA/FDMA cellular radio systems, Fixed Channel Allocation (FCA) is used to allocate channels to
customers. In FCA, the number of channels in the cell remains constant irrespective of the number of customers in that cell. This, results in traffic congestion and some calls being lost when traffic gets heavy [8], [11]. A better way of channel allocation in cellular systems is Dynamic Channel Allocation (DCA) which is supported by the Digital Cellular System (DCS) and other systems. DCA is a better way not only for handling bursty cell traffic but also in efficiently utilizing the cellular radio resources. DCA allows the number of channels in a cell to vary with the traffic load, hence increasing channel capacity with little costs [11]. Since a cell is allocated a group of frequency carriers (e.g. f1-f7) for each user, this range of frequencies is the bandwidth of that cell [11].
The number of channels is calculated using the Erlang B formula for loss probability (assuming Mobile Assisted Handover (MAHO) Scheme) [9], [12]. The calculation is done with Erlang B calculator.
To relate call arrival rate to the performance of a network,
the term grade of service (GOS) denoted by B is used. The
GOS can be mean proportion of time for which
congestion exists, or probability of congestion or blocking
probability, or probability that a call will be dropped due to congestion. It is defined in [12] as:
B = traffic lost/traffic offered (4)
From (4), it is obvious that drop call-probability varies
inversely with call arrival rate, that is, drop-call probability decreases as call arrival rate increases. This leads to the deduction that system performance improves as the traffic entering the system increases [1], [2].
Call duration is another parameter that can affect the quality of service in a cellular network, hence it is considered when planning the network [1], [2]. Call duration or mean call holding time is defined as the time a mobile station takes to complete a call connection. Mathematically, call duration is given by [12]:
ℎ A (5)
λ
The formula is given as follows:
AN
B N!
k
(1)
Where A = traffic intensity in Erlangs, = call arrival rate
Thus call arrival rate varies with call duration the same way it varies with drop-call probability. Thus drop-call probability increases with a decrease in call duration.
∑N A
k=0 k!
Where B = loss probability, A = offered traffic intensity in
IJSER © 2011 http://www.ijser.org
International Journal of Scientific & Engineering Research Volume 2, Issue 10, October-2011 3
ISSN 2229-5518
Drop-call probability is given by [1]:
n
cellular radio networks. In this case, the available radio resources and their utilization as well as traffic intensity related parameters play significant roles. More work has been carried out to discover the trends of these drop-call
P(Y = n) = (vd t)
n!
e , n ≥ 0 (6)
probability factors in a live GSM network. Starting with data obtained from the network, mean values were
Here, vd is the drop-call rate, t the call duration, while Y is a random variable that counts the number of drops and n is the confirmed calls dropped. This is a Poisson Probability function with a discrete variable which counts the number of dropped calls [1], [2], [13], [14].
The number of dropped calls is calculated from the relation:
computed for the number of channels, utilization factor, and then the call arrival rate and call duration. At least, the results show a general trend that is in line with the basic principles. The result can be used as a good guide for evaluating and optimizing an operating GSM network. In this case, the results show the network under study as a well established network. The result can further be used for drop-call probability model
simulation for further confirmation.
Drop call rate = No. of dropped calls
No. of call attempts
(7)
TABLE 1
The probability of occurrence of the call dropping event
NETWORK DATA
(drop-call probability) based on the above formula is thus
calculated using the Poisson mass probability function from Microsoft Excel.
In this section, the analysis of the above mentioned drop- call probability factors with measured data obtained from an operative GSM network are presented. The data were collected over a period of four months for each of six
MSC Numb
er of Active Base Statio ns
Perio
d
Traffic
Intensi ty
Dro
p- call Rate in busy hour (%)
Drop-
call rate with handov er (%)
Numbe
r of Call Attemp ts in Busy Hour
MSCs and comprised of the traffic intensity, drop-call rate, and number of call attempts as shown in Table 1. From these, the drop-call probability and the required radio channel and traffic parameters were estimated. The radio channel related parameters are: number of channels and channel utilization factor, while the traffic parameters are: call arrival rate and call duration. The mean values of these parameters were estimated per cell for each of the six MSCs.
The results indicate that on average, drop-call probability decreases as the number of radio channels, utilization factor, call arrival rate, and call duration increase, all in line with the above stated basic principles. These trends are depicted in Table 2 showing mean values of these parameters and the drop-call probability. These are further represented graphically in Fig. 1, 2, 3, 4.
The interpretation of these results is that the network in question was well established and optimized and operating optimally during the period the measurements were taken. The kinks noticed in the curves are not unexpected because of the constantly varying and unpredictable propagation characteristics of the radio channel.
MSC
1
MSC
2
MSC
3
MSC
4
MSC
5
MSC
6
65 Mar Apr May June
89 Mar Apr May June
60 Mar Apr May June
54 Mar Apr May June
51 Mar Apr May June
46 Mar Apr May
5211
5190
4933
5098
3988
4315
4591
3977
2611
2499
2598
2535
1836
2071
1920
1848
1718
1801
1698
1607
1391
1438
1298
2.10
2.38
3.17
2.80
1.35
2.35
1.92
1.99
3.10
2.60
2.80
2.96
3.17
2.69
2.85
2.34
1.99
2.20
2.32
1.98
1.51
1.71
1.81
0.05
0.06
0.06
0.08
0.07
0.056
0.05
0.04
0.07
0.07
0.08
0.06
0.04
0.06
0.06
0.09
0.09
0.065
0.08
0.08
0.08
0.08
0.08
250014
251185
248991
249492
228825
220306
231222
201650
148700
149881
151596
157018
116561
129988
113818
134853
117876
115928
112131
113339
95213
94388
91889
In this paper, a wide variety of factors influencing drop- call probability have been identified and firstly explained on basic principles and axioms available in literature for
June 1319 1.91 0.08 90866
IJSER © 2011 http://www.ijser.org
International Journal of Scientific & Engineering Research Volume 2, Issue 10, October-2011 4
ISSN 2229-5518
6.5 | |
TABLE 2 | 6 |
MEAN VALUES OF COMPUTED TRAFFIC/CHANNEL PARAMETERS
AND DROP-CALL PROBABILITY
5.5
5
MSC Call Arrival Rate
Call Duration (s)
Number of Channels
Channel Utilization Factor
Drop-call Probabilit y (%)
4.5
4
3.5
(calls/s)
3
1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
Channel Utilization
MSC6 0.14 211 34 0.88 6.1
Fig. 2: Variation of Drop-call Probability with channel utilization
6.5
6
5.5
5
4.5
4
3.5
3
2.5
2
1.5
6.5
6
5.5
5
4.5
4
3.5
3
2.5
2
1.5
0.16 0.18 0.2 0.22 0.24 0.26 0.28
Call Arrival Rate (calls/s)
30 40 50 60 70 80 90
Number of Channels
Fig. 1: Variation of Drop-call Probability with Number of
Channels
Channel Utilization
Fig. 3: Variation of Drop-call Probability with Call-Arrival
Rate
IJSER © 2011 http://www.ijser.org
International Journal of Scientific & Engineering Research Volume 2, Issue 10, October-2011 5
ISSN 2229-5518
6.5
6
5.5
5
4.5
4
3.5
3
2.5
2
1.5
210 220 230 240 250 260 270 280 290 300
Call Duration (secs.)
Figure 4: Variation of Drop-call Probability with Call
Duration
[1] G. Boggia, P. Camarda, and A. D’Alconzo, “Modeling of Call Dropping in Well-Established Cellular Networks”, EURASIP Journal on Wireless Communications and Networking, Vol. 2007, Article ID
17826, October, 2007
[2] G. Boggia, P. Camarda, A. D’Alconzo, A. De Biasi and M. Siviero, “Drop Call Probability in Established Cellular Networks: from Data
Analysis to Modeling”, DEE – Politecnico di Bari, Via E. Orabona, 4 – 70125 Bari (Italy), Proc. IEEE VTC Spring 2005.,Vol. 5, pp2775-
2779, 2005
[3] B.L. Cragin, “Prediction of Seasonal Trends in Cellular Dropped Call Probability”, IEEE, Capital City Applied Research, Research Paper, 2375 Club Mredian Dr. #B7, Okemos,MI 48864, USA, 2006
[4] A. R. Mishra, “Fundamentals of Cellular Network Planning & Optimization”, 2004 John Wiley & Sons, Ltd. ISBN: 0-470-86267-X, PP 21-29
[5] H.M. Aziz Basi, M.B.R. Murthy, “Improved Performance of
Traffic Dependent Outage Rate Cellular System”, Journal of
Computer Science, Jan, 2005.
[6] M.J. Nawrocki, M. Dohler and A. H. Aghvami, “Modern Approaches to Radio Network Modeling and Planning” – “Understanding UMTS Radio Network, Modeling, Planning and Automated Optimisation”, 2006 John Wiley & Sons Ltd.
[7] S. Ganguly, B. Nath, and N. Goyal, “Optimal Bandwidth Reservation Schedule in Cellular Networks”, Research Paper, Department of computer Science, Rutgers University, Piscataway, NJ, 08852, IEEE INFOCOM, 2003.
[8] B.B. Madan, S. Dhamaraja and K.S. Trivedi, “Combined Guard Channel and Mobile Assisted Handoff for Cellular Networks” IEEE Transactions on Vehicular Technology, 2006
[9] G. Haring, R. Marie, R. Puigjaner and K. Trivedi, “Loss formulas and their Application to Optimization for Cellular Networks”, IEEE Transactions on Vehicular Technology, Vol. 50, No.3, May,
2001
[10] N. Ekiz, T. Salih, S. Küçüköner, and K. Fidanboylu, “An Overview of Handoff Techniques in Cellular Networks”, Proceedings of World Academy of Science, Engineering and Technology, Volume 6, June 2005, ISSN 1307- 6884.
[11] H. Sun and C. Williamson, “Simulation Evaluation of Call Dropping Policies for Stochastic Capacity Networks”, Department of Computer Science University of Calgary, Calgary, AB, Canada T2N
1N4, 2005.
[12] C.Y. Lee, “Mobile Cellular Telecommunications Systems”, McGraw-Hill, 2nd Edition, 1995.
[13] D.D. Dajab, S.N. Tarka, and B.G. Bajoga, “Simulation and Analysis of Drop-call Probability Model: A Case Study of MTEL”, Nigerian Journal of Engineering, Vol. 16, No. 1, 2009
[14] S.N. Tarka, “Effects of Network Characteristics on Conditional Drop Probability: A Case Study of MTEL”, MSc Thesis, Department of Electrical Engineering, Ahmadu Bello University, Zaria, June,
2009.
IJSER © 2011 http://www.ijser.org