Hybrid intelligent model for real time assessment of voice quality of service

Mobile telecommunication has made an unprecedented positive impact on every nation's economy and social life. It has allowed users to save time and money in their daily transactions and improved their quality of life. However, network inaccessibility, unreliability, and unsatisfactory voice service have led to many complaints from clients. One of the significant weaknesses inferred from the literature review is the lack of reliable user-based information for monitoring and performance evaluation. This work is aimed at developing a hybrid model for the assessment of voice calls quality of service (QoS) offered by Mobile Network Operators (MNOs) based on standard performance metrics, and to carry out the analyses on developed models. The experimental study is carried out using the network data collected from volunteer's mobile phones using the host-based crowdsourced technique. The collected user-based data for quality of service metrics is modeled based on the combination of the Fuzzy Logic algorithm and Takagi Sugeno Kang inference mechanism of the Neuro-Fuzzy model. The modeling was implemented using MATLAB and Weka analytics as front end and MySQL as backend on Windows 10 operational environment. The performance evaluation of the hybrid model shows the high accuracy of 97.10%; the true positive rate and precision rates were 95.5% and 97.47%, respectively.