Spatiotemporal Mobility Prediction in Proactive Self-Organizing Cellular Networks

Mobility prediction, one of the key enablers of proactive self-organizing networks, aims at efficient management of future cellular networks, which are envisaged to be extremely dense and complex due to conglomeration of diverse technologies. This paves the way for resource reservation prior to actual handover for seamless handover experience and for forecasting user traffic distribution. In this letter, we have utilized semi-Markov model for spatiotemporal mobility prediction coupled with steady state and gain analysis in cellular networks. Maximum prediction accuracy of 90% is achieved in the experimental evaluation leveraging on the real network traces generated by smartphone application.