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 ICARO, 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 based on real measurement data collected in Berlin. We show that our solution performs close to an omniscient optimizer and outperforms a limited horizon omniscient optimizer by 10-15%. Our solution provides up to 30% saving of system resources compared to a simple solution that always maintains a full buffer and is close to optimal in terms of buffer under-run time.