Joint optimization of computing and radio resource for cooperative transmission in C-RAN

Abstract

By concentrating the baseband units (BBUs) into a data center, the cloud radio access network (C-RAN) enhances the utilization of the computing and radio resource, promoting the cooperative transmission among remote radio heads (RRHs). A flexible computation provisioning can obtain better network gain due to the strong relationship between the processing and transmitting emphasized in recent research. Inspired by above observation, this paper jointly optimizes the computing and radio resource to maximize the system throughput along with the energy efficiency. The optimization problem is formulated as a mixed integer non-linear problem (MINLP). We develop a two-step approach to reducing the complexity of the problem. First, the computing resource allocation, i.e. virtual machine (VM) and user mapping is solved by the equivalent maximum weight perfect matching problem. Then, to solve the radio resource allocation, i.e. beamforming design, the original MINLP is transformed into a weighted minimum mean-squared error problem which is solved iteratively, and l 1 -norm reweighting is adopted to relax the constraint. A dynamic clustering scheme is further concerned to improve the system performance. Simulation results indicates that our joint optimization scheme can facilitate the throughput improvement and maintain a high energy efficiency meanwhile.