Load diversity based optimal processing resource allocation for super base stations in centralized radio access networks

Abstract

To reduce construction cost and energy consumption of cellular networks, the centralized radio access cellular network infrastructure with a Super BS (CSBS) architecture has been proposed. The CSBS architecture has the advantage that the centralized processing resource can be flexibly organized. However, the problem of low processing resource efficiency caused by load discrepancy of base stations still exists and the processing resource allocation methods for physical layer processing are not suitable for protocol processing. In this article, a framework for protocol processing resource management is designed and a load diversity based processing resource allocation (LDA) algorithm is proposed, which employs dynamic programming with the consideration of load discrepancy. Also, to support large mumber of BSs, a suboptimal algorithm hybrid load diversity based processing resource allocation (H-LDA) is also proposed. Simulation results show that LDA and H-LDA algorithm significantly outperforms the conventional methods, respectively, by more than 60% and 55% in the processing resource efficiency and the performance gap grows with the increasing BS number.