Predictive and Adaptive Resource Reservation (PARR) for Cellular Networks

Abstract

In cellular networks, a higher priority can, out of Quality of Service reasons, be given to ongoing connections than new call attempts. This can be achieved by having capacity in each cell (guard channels) exclusively reserved for arriving handover calls. In this paper, an advanced guard channel scheme, called Predictive and Adaptive Resource Reservation (PARR), will be introduced. The PARR scheme determines the amount of guard channels to allocate in a cell through subscriber movement predictions. The novelty of PARR is that both positioning and history-based measurements are used as input for the predictions. To improve the prediction accuracy, different data sets are applied for calls with different mobility and latest visited cells. Simulations will be used to show that significant performance improvements are obtainable with PARR.