هدي السليماوي

Abstract Network performance optimization is one of the most important topics in the field of communications. Due to the existence of millions of mobile phone network users worldwide, the great economic value of mobile networks, an‎d the provision of complete mobile services, this subject has become critical. Optimizing network performance in this context, using different methods, is of significant importance in order to ensure quality, stability, an‎d reliability for users. Key performance indicators (KPIs) an‎d key quality indicators (KQIs) are the basis for network optimization, leading to improved performance through different methods. Traditionally, optimization techniques relied on data collection through physical presence, measurements, an‎d drive tests. However, with the advent of new approaches such as MDT (Minimization of Drive Test), eva‎luation can be done dynamically using user data, reducing the need for extensive drive tests. Nevertheless, MDT-based methods face various challenges: user equipment limitations, measurement accuracy issues, high costs, time-consuming processes, an‎d geographical coverage restrictions. Therefore, predictive methods for optimization have been proposed, where data collection is used indirectly, allowing optimization to be performed without complete drive testing. The main goal of this thesis is to present a predictive method to optimize this process by reducing drive tests. The method consists of several key steps: identifying measurement points based on signal reception, selec‎ting can‎didate points for prediction, modeling using regression an‎d channel parameters, grouping an‎d classification, an‎d finally estimating the received signal strength (RSRP). In this work, artificial intelligence techniques are employed to enhance prediction accuracy, correct potential errors in estimated RSRP values, an‎d reduce the dependency on costly an‎d time-intensive drive tests. Ultimately, this leads to more efficient network optimization, improved decision-making, an‎d reduced operational costs while maintaining service quality.

(به ين هسازي شبكه، شبك ههاي تلفن همراه، تست درايو، MDT ، هوش مصنوعي، يادگيري عمي ق)

(Network Optimization, Mobile Networks, Drive Test, MDT, Artificial Intelligence Deep Learning)

Huda alsulaimawi

Abolfazl Diyanat