Mobile Network Resource Optimization under Imperfect Prediction

Abstract

A highly interesting trend in mobile network optimization is to exploit knowledge of future network capacity to allow mobile terminals to prefetch data when signal quality is high and to refrain from communication when signal quality is low. While this approach offers remarkable benefits, it relies on the availability of a reliable forecast of system conditions. This paper focuses on the reliability of simple prediction techniques and their impact on resource allocation algorithms. In addition, we propose a resource allocation technique that is robust to prediction uncertainties. The algorithm combines autoregressive filtering and statistical models for short, medium, and long term forecasting. We validate our approach by means of an extensive simulation campaign for different network scenarios. We show that our solution performs close to an omniscient optimizer as well as the simple solution that always maintains a full buffer in terms of prefetching data before it is needed, while at the same time using 20% less network resources than the simple full buffer strategy.