Lakhasly

The importance of performance evaluation of mobile networks is inspired by the need to
utilize the limited radio resource in the most efficient manner. The continuous increase in
demand, for high-quality service, coupled with the future telecommunication network target of
providing integrated services with multimedia applications, over low-powered mobile
computing devices, has further increased the need for evaluating the performance of mobile
networks [1 - 4]. From the users’ perspective, the performance of a mobile network is measured
in subscribers’ satisfaction index, network availability, and efficiency [5 - 8]. However, from
technical and regulatory perspectives, network performance evaluation is necessary for assessing
and measuring the coverage, capacity, and quality of mobile networks with respect to the service
level agreement (SLA) signed with the regulators.
While the operators focus on metrics like the bit error rate (BER), frame erasure rate (FER),
and mean opinion score (MOS), the Nigerian Communications Commission (NCC) has
identified some measurable key performance indicators (KPIs) to evaluate the success of
network operators in Nigeria vis-à-vis subscribers’ perception. These KPIs include dropped
calls, handover success, blocked calls, interference, coverage, and quality, amongst others.
Notably, the KPI results are obtained from the voice and data channels logs on the network [9,
10]. However, to meet the quality of service (QoS) target set by NCC, the following KPI target
must be met by all network operators – accessibility (call setup success rate ≥ 98%), retainability
(call drop rate ≤ 1%), mobility (handover success rate ≥ 98%) and service integrity [11, 12].
As a national regulatory body vested to establish minimum QoS standards in service delivery
for mobile network operators (MNOs) and internet service providers (ISPs), the NCC requires
that each network provider reports quarterly, on customer satisfaction in (rural, urban and suburban) Nigeria [13]. This was motivated by the seeming dissatisfaction with the QoS by the
rapidly increasing number of information and communications technology (ICT) users across
the country [6, 13]. It has, increasingly, become evident and pronounced that the Nigerian
telecommunication sector has been plagued with problems, causing frustration among
consumers. In recent studies, operators have attributed the poor QoS to the vandalism of network
equipment, lack of roads, epileptic power supply, and security breaches in the country. However,
mobile network users expect that the increase in technological innovations will foster good
speech quality and uninterrupted services [5, 6]. This expectation is based on the perceived
increase in the penetration of long-term evolution (LTE) networks in the country and the
anticipation of the deployment of 5G networks. The August 2019 to July 2020 report in [13],
revealed that for the GSM network, all, but one of the MNOs satisfied the minimum requirement
for network accessibility (≥ 98%). This justifies previous statistics by the NCC in May 2016 to
April 2017, that amongst the four major MNOs, only one performed outstandingly (satisfying
the minimum requirement and beyond) in network accessibility, while one of the MNOs
performed poorly in network retainability [14].
The authors in [15] conducted a drive test to evaluate and compare the network performance
of four networks (W, X, Y, and Z) in Minna, Nigeria, using Test Mobile System (TEMS). The
study validated the claim by NCC that X network had the best network quality in Nigeria. A
related study in [5] employed the drive test method with statistical analysis models (chi-square
and Fisher’s tests) for the performance benchmarking of GSM and Universal Mobile
Telecommunications System (UMTS) network operators in Ankara, Turkey. Another study in
[4], revealed an analytical survey on the performance evaluation of wireless cellular networks
under more realistic assumptions. This study explained that the premature termination of calls
(call drop) is possible due to unsuccessful handover of calls when a user moves out of the
coverage area of a serving cell, and the target cell has no or limited resources to serve the call
connection. A call admission control (CAC) scheme was proposed to reduce the call drop rate
[4] [16 - 18]. Results in [5, 15] emphasize the role of regulatory organizations in monitoring and
administering wide-scale benchmarking of operators.
The exponential increase in the penetration of broadband communication requires evaluating
its performance from an end-user perspective [5, 19]. The authors in [19], conducted a pilot study
to benchmark and evaluate the QoS experienced by mobile wireless users in Pakistan by
deploying an android application (My Speed Test PK) for mobile customers to measure the
performance of the five broadband operators (Ufone, Telenor, Zong, Warid, and Mobilink) in
Pakistan. The analysis of the performance obtained by mobile wirel meticulous measurement for the mobile network operators (EE, O2, Vodafone, and Three) in the
United Kingdom [23]. The measurements carried out were based on the drive tests and walk tests
conducted across 20 large cities and 11 cities in the UK respectively. EE, Vodafone, O2, and
Three were rated very good, good, satisfactory, and good respectively.
To effectively cater to subscribers’ demand, [24] carried out a research on maximizing
coverage, capacity, and quality with guaranteed QoS using a cell cluster optimization approach.
Two types of drive tests were conducted to optimize the network – initial drive test and postdrive test. The initial drive test was conducted to serve as a reference level for verifying the
performance of the network. The study area was divided into three clusters with each of the
clusters having 26 to 40 sites on air. From the drive test analysis, the problem identified across
the three GSM clusters were – reduced power output, poor transmission line, increased
interference, coverage hole, cross feeder, overshooting, amongst others. After the optimization,
the coverage reliability results revealed better coverage performance as well as better quality
performance.
While past studies have critically evaluated the performance of GSM and WCDMA/UMTS
networks from customers’ viewpoint, the study in [25] demystifies the root cause of poor
network signal strength and its impact on the QoS provisioning by MNOs. Specifically, the
handover procedure and power control were observed in a dense (saturated) network
environment using drive test methods. Results from the study emphasized the need for optimal
coverage on all the mobile networks, as well as to have the radio frequency (RF) air interface
parameters configured to avoid problems of ping-pong handovers, and poor power control,
which causes interference and low RSSI among mobile stations (MSs).
In a populated urban area with a literate population such as the University of Ilorin, radio
resource availability and reliability is a factor that determines the success of mobile
communication. Students and staff, each, complain of poor network accessibility mostly in the
academic areas of the University. This incessant complaint which has made many staff and
students, each, to subscribe to more than one network operators’ service, is an indicator of poor
QoS in the University of Ilorin, and by extension most urban centers in Nigeria. On the evidence
of this, the networks need to be under continuous monitoring and control, to maintain and
improve the performance of the system – an approach that this study seeks to evaluate.
As increase in population and infrastructural development is continuous in any higher
institution of learning, the need for improvement in telecommunication services is necessary.
This study forms a basis for the regular inspection and optimization of mobile networks in the
University of Ilorin, as well as creates a guide for path loss modeling suitable for the campus
terrain. Real-time monitoring of radio frequency in the GSM and WCDMA band was conducted.
Three drive test locations were selected in the University of Ilorin - University main campus,
senior staff quarters, and Jalala junior staff quarters. Possible solutions to improve the network
in the selected locations are recommended.