Mobile Network Traffic Forecasting Using Artificial Neural Networks

Mobile communication systems need to adapt to temporally and spatially changing mobile network traffic, due to dynamic characteristic of mobile users, in order to provide high quality of service. Considering these changes are not purely random, one can extract the deterministic portion and patterns from the observed network traffic to predict the future network traffic status. This prediction can, then, be utilized for a series of proactive network management procedures including coordinated beam management, beam activation-deactivation, load balancing. To this end, this work studies an intelligent predictor using artificial neural networks, and compares it with a baseline scheme that uses linear predictors. It is shown that the neural network scheme outperforms the baseline scheme for relatively balanced data traffic between random and deterministic mobility patterns. For highly random or highly deterministic mobility patterns, the performances of the two considered schemes are similar to one another.