Network Traffic Prediction in 4G- LTE Using Machine Learning Techniques

Machine-to-machine (M2M) technologies are making a huge impact on the communication network which will create new opportunities and values for various sectors. As part of the Next Generation Networks (NGN), Long-term evolution (LTE), LTE-advanced (LTE-A), and LTE-machines (LTE-M) play an influential role in the M2M communications due to their native IP connectivity and scalability to connect a massive number of devices. A quality of service (QoS) framework is a fundamental component of LTE for providing quality service to machines, end users and managing the network resources. Currently, in internet-based applications, including Real-Time (RT) and Non-Real-Time (NRT), there is a huge traffic characteristics/patterns which consequently affects QoS and user experience. In order to effectively deliver QoS traffic from different applications, machine learning (ML) techniques can be applied to the LTE Networks with the focus on classifying, predicting and optimizing the traffic. In this work, we aim to study the effect of machine learning techniques for predicting traffic flow in the LTE network and its impact on QoS.An extensive experimental evaluation is carried out using a real-time network traffic dataset. In this analysis, the effectiveness of the predictions using various machine learning techniques are explored.