5G Traffic Prediction with Time Series Analysis

In today’s day and age, a mobile phone has become basic requirement needed for anyone to thrive. With the cellular traffic demand increasing so dramatically, it is now necessary accurately predict user in networks, improve performance terms of resource allocation utilization. Since learning prediction classical appealing field, which still yields many meaningful results, there been an interest leveraging Machine Learning tools analyze total served each region, optimize operation network. help this project, we seek exploit history by using nature occurrence future traffic. Furthermore, classify into application types, increase our understanding By power machine identifying its usefulness field networks try achieve three main objectives - classification generating traffic, packet arrival intensity burst occurrence. The design system done Long Short Term Memory (LSTM) model. LSTM predictor developed experiment would return number uplink packets estimate probability specified time interval. For purpose classification, regression layer model replaced SoftMax classifier used one four applications including surfing, video calling, voice streaming.