Predicting Quality of Service on Cellular Networks Using Artificial Intelligence

The purpose of this research is to explore the application of artificial intelligence (AI) techniques, particularly Machine Learning, in predicting quality of service (QoS) on mobile networks, with the main focus being to test the ability of AI models to predict several QoS parameters, involving several important stages that reflect best practices in the development of artificial intelligence (AI)-based predictive systems for mobile networks. The dataset used in this study consists of data collected from simulations of mobile networks with various load and latency conditions. The parameters measured include Throughput, Latency and Packet Loss. Model evaluation was carried out to measure prediction performance using Mean Absolute Error (MAE), Root Mean Square Error (RMSE), and Mean Absolute Percentage Error (MAPE) measurements. The AI models used include machine learning algorithms consisting of K-Nearest Neighbors (KNN) for classification and regression on QoS datasets, Support Vector Machine (SVM) to model non-linear relationships between QoS parameters and input variables, and Deep Learning (LSTM=Long Short-Term Memory) used to predict QoS based on time sequence data. This study found that LSTM-based deep learning models have the lowest prediction error rate in estimating packet loss, so they can provide the most accurate results in predicting QoS on mobile networks. This approach is capable of handling data that is sequential and has significant time dependence, making it more suitable for dynamic mobile network applications.