Leakage-based distributed minimum-mean-square error beamforming for relay-assisted cloud radio access networks

In this study, the authors investigate the linear minimum-mean-square-error beamforming design for relay-assisted cloud radio access network (C-RAN). A standard C-RAN architecture separates baseband processing units and wireless radio units in order to save energy cost. To further enhance network coverage, several relay nodes (RNs) are also deployed. Regrading the per-antenna power constraints at both of the remote radio heads (RRHs) and the RNs in the author's work the beamforming matrices at the RRHs and RNs are 'jointly' optimised for the considered relay assisted C-RAN. The considered optimisation problem is a non-convex and multiple variable optimisation problem which is in general very hard to solve. In order to make the design suitable for large scale networks exploiting to the problem structure a novel two stage decomposition algorithms are proposed. Finally, a detailed mean-square-error performance comparison is given by the simulations.