Quality of Service (QoS) Evaluation in Wideband Code Division Multiplexing Access (WCDMA) Networks Using Empirical Analysis Methods

Performance evaluation is essential in maintaining the Quality of Service (QOS) of the Wideband Code Division Multiplexing Access (WCDMA). This work was motivated by the reception of the poor signals, increase call drop, failure rate which was a poor QoS Reception. The aim is to survey WCDMA services in Owerri environs and establish that there are degradation and the level of debasement in the network. The methodology involved an Empirical Analysis through Drive Test across Owerri City in Imo State. The work adopted the empirical approach and deduction of some Standard performance metrics like call drop rate, failure rate, call success rate, call completion rate, Handover success rate and handover Failure Rate, compare with expected KPI(key performance indicator) threshold. From the assessment, it was found that only one out Four Networks (“GLO”) met the target Call Drop Rate (CDR), Call Completion Success rate (CCSR), Call Setup Success Rate (CSSR) and Call Blocked Rate (CBR) and the Handover was better in “GLO” and 9 mobile than in the “MTN” and Airtel.


Introduction
The telecommunication industry around the world is facing notable challenges from the unending customer demands and competition from other operators. market [1]. The increasing need for efficient high-speed internet and multimedia application further resulted in the invention of Universal Mobile Telecommunication Services (UMTS), a 3rd Generation network (3G), widely known as Wideband Code Division Multiple Access (WCDMA) operating at 2.1 GHz [2].
WCDMA covers a larger area and provides greater capacity and a greater number of variable parameters to monitor. Information obtained from Nigeria Communication Commission (NCC) report: 2012, 2017, 2018 on the four networks (MTN, GLO, 9 Mobile, and Airtel) established the fact that the WCDMA network is currently experiencing some challenges such as poor accessibility and retain ability. In the telecommunication industry, network performance is measured using some key performance indicators (KPI). KPI is the pressure unit metrics commonly used to evaluate and define how good/successful a Network is. KPI is used to consider the QoS structure of network through some metrics such as call drop rate (CDR), Call Competition Success Rate (CCSR), Call Success Set up Ratio (CSSR), Network Accessibility Ratio (NAR), Network Retainability Ratio (NRR), recall termination rate (CTR), etc. The performance of network and service quality as frequently accessed with the drive test method [1].
In this research work, the drive test was conducted for the four (4)  This is the probability that a mobile user will establish successful voice communications between the two ends of the network within a given condition. It is also expressed as the call setup rate. It is expressed as: It is the ratio between the number of successful calls and the number of normally terminated calls. It is also the probability that an active call comes to an end successfully in a network.
( ) where, CDR is defined in Equation ( Table 2. smmation of all data that falls in 11 dBm Coverage quality 100 Smmation of data Table 3 shows the NCC Threshold for all the Network Performance KPI's.  Their results showed that operator B had the best performance for accessibility and retainability. The authors did a further analysis of measurement conducted on speech quality using Perceptual Evaluation Speech Quality (PESQ).
They analyzed the speech quality (SQ) of the Network operators using Probability Density Function and Cumulative Density Function (PDF/CDF) analysis, which according to them was the most popular KPI. They used PESQ to measure the SQ and evaluated with Chi-Square and fisher's exact tests. The researchers, therefore, were able to determine which network operator provided the highest Speech quality in Ankara. Their results after comparing UMTS and GSM networks showed that UMTS had better signal levels and voice quality than GSM.
The work of [4], conducted an experimental measurement campaign and drill down the analysis. The aggregate of RSCP and Ec/No, MOS and PESQ was used to deduce the mean, min, max and STD of the UMTS and the aggregate of Rxlev Sub, RxQual Sub, MOS (LQ) and PESQ for GSM in their work on Speech and Video Quality Assessment of GSM and WCDMA Rollout Mobile Radio Access Networks in a Regulated and Competitive Market. The five selected KPI's are Call Setup time (CST), DCR, Call Success Rate (CSR), Access Failure Rate (AFR), Service Success Rate (SSR), voice and voice quality for GSM and UMTS. Finally, they evaluated Video quality by an assessments algorithm with a range of MOS, PSNR parameter which has a range from 0 to 110 dB, Jerkiness/Pausing, Freezing due to video framed skipping, Jitter which is computed as the standard deviation of all frame intervals according to the typical frame interval and Matched frames.
The authors in [5] employed a data mining process which is also referred to as the nontrivial technique of extracting implicit and potentially useful information from existing data sets in their work, titled "End-User Satisfaction Assessment Approach for Efficient Network Performance Monitoring in Wireless Communication Systems". The performance statistical data was obtained from the NCC QoS KPIs database. A total of eight months data within the periods of February to September 2013 was obtained and the selected Network was named A, B, C, and D. The researchers analyzed the two key performance measures which are QoS and Grade of Service (GoS). The KPIs used were CSSR, CDR. From the results obtained it none of the operators except operator D in March to September met the NCC target of 98% of CSSR standard within the evaluation period.
The work of [6] studied the "Data on the key performance indicators for quality of service of GSM Networks in Nigeria" The authors empirically assessed and analyzed the Comprehensive data obtained from Nigeria Communication Commission (NCC) The KPIs used are CDR, CSSR, SDCCH Congestion and traffic channel TCH congestion for the GSM Network operators in Nigeria (Airtel, Etisalat, Glo, and MTN).
(FP) to solve the problem of congestion in WCDMA Networks. The experimental approach was conducted locally through empirical analysis from six selected locations in Owerri metropolis Imo State From the review of literatures, [2] observed that among the two prominent Network assessed, none of the networks met the NCC threshold based on the KPIs selected, [1] and [5] showed that one out of the Networks assed met the threshold, [1] and [4] assessed voice quality as a measure for Network integrity using MOS(LQ) and PESQ for GSM. [3] observed that there was minimal congestion in the four Network studied, none of the operators met the NCC threshold for HOSR, most of the operators met the NCC target for Call Completion Rate (CCR) and in the overall performance, "GLO" was found to be the best.
The software engineering approach was used after evaluation and analysis to optimize the raw data to improve the QoS thresholds by service providers. [7] after the Analysis used fuzzy logic to improve the performance of WCDMA Network caused by congestion of the Network.
This paper study will tend to investigate and evaluate the performance of the four Networks using drive test in Owerri Municipal and Owerri North of Imo State. The experiment will involve the collection of data, analyzing the data to asses Network Availability, Accessibility and integrity for the four networks

Methodology
In this performance evaluation assessment, two approaches were used which include the Empirical and Analytical Approach. The Empirical Approach for QoS Analysis was through drive test macro mobility. Drive test was carried out in Owerri Municipal and Owerri North to evaluate the WCDMA Network performance in the four Networks (MTN, Airtel, 9mobile and "GLO").

Measurement Environment
The measurement was carried in some parts of Owerri Municipal and Owerri North Zone. The Primary Measurement Areas are Imo State University Owerri, Cementary Road, Bishop Court, Ikenegbu, Okigwe Road and Orji Road. The Experiment focuses on the densely populated area and it avoided disjointed data, avoiding traffic jams and with complete information regarding the four Networks. The test was carried out within the busy hours, between 3.00 pm to 5 pm.
Data obtained from the mobile phone (log file) was transferred for postprocessing. Evaluation is then carried out using the Tem development Software. The drive test route that was used in the analysis was clearly shown in Figure 2 and Figure 3. The site identification numbers, cell identification codes, antenna

Result
The data were obtained from TEM discovery software during the drive test include Received signal strength for WCDMA Network, Azimuth position of the Antenna, Neighboring cell information and other statically and pictorial information of the WCDMA parameters.

Discussion
From the Average received Signal Strength of the four Networks, the Signal received from "GLO" was satisfactory compared to the other Networks. It was found that in WCDMA all the Networks expect "GLO" did not reach the NCC threshold of 98% and 96% for CSSR and CCSR respectively. The statistical values of the experiment conducted were used to deduce the KPI's. These standard KPIs were used for benchmarking. From the Statically Analysis in Table 4 it was found that only "GLO" met the target CDR, CCSR, CSSR, and CBR.
In Figures 4-7 and from Table 5, it was found that the handover was smooth in "GLO" and 9 mobile than in "MTN" and Airtel (Figure 8 and Figure 9).

Conclusion and Future Work
This study investigated the Network performance and the Quality of Service delivered by the Radio Networks of WCDMA deployed in Owerri City. The findings of this research work showed that both radio network technologies investigated should be optimized for the better quality of service delivery in some parts of the city. The poor radio coverage and degradation in the quality of service identified in the affected areas of the city were traced to resource utilization, code management and other physical factors like topology and the presence of physical obstructions present in the propagation environment. Also, poor quali-ty samples collected were due to poor coverage in the area. The base stations should be Regularly Monitored and Optimized by Radio Engineers and vendors to enhance radio coverage in the affected areas. The Hand-Over (HO) failures were mainly due to Base Station Controller (BSC) harmonization problems and some black spots observed. This should be resolved appropriately and within three days of occurrences according to the NCC requirement. If the Radio Network is constantly monitored and optimization is successfully performed, the Network Integrity, Availability, Retain ability will be enhanced and that means more customers and more Gain to the Mobile Operators.