Contrasting single-point and multi-point half-rate allocation strategies for AMR-WB in high-capacity cellular networks

In mobile networks a most economic utilization of the channels on the air interface by means of intelligent radio resource management is desired since these resources are the most valuable ones. Half-rate speech codecs introduced in GERAN networks for capacity increase involve quality degradation compared to full-rate codecs especially in error-prone radio channel conditions. To provide best possible speech quality the available resources in the network should be utilized as extensive as possible. In addition flexible mechanisms are required to provide additional capacity whenever needed for obtaining the optimum trade-off between quality expectations and capacity requirements. In this study two different load dependent half-rate allocation algorithms for speech services have been proposed. The performance differences have been evaluated thoroughly by system level simulations and compared in terms of system performance. Applied quality criteria are frame erasure rate, bad speech quality probability, codec mode distribution, speech quality indicator and call drop rate. The results show significant differences in channel utilization, speech quality and achieved network capacity.