Antennas deployment optimization in coordinated multipoint processing with local beamforming

The coordinated multipoint paradigm is based on the concept of enabling joint processing of the multiple air-link signals received by a group of spatially dislocated base stations (BSs). The degree of cooperation among BSs and optimal partitioning of the baseband processing into centralized (executed at the common control unit) and local functionalities (autonomously executed at each base station) represent powerful degrees of freedom to be explored. In this paper, we investigate this trade-off for the uplink by accounting for different densities of antennas and users. Given a coverage area, K users communicate through M total available antennas that are uniformly grouped into Q radio access entities (RAEs). Each RAE has N = M/Q antennas that are arranged as a uniform linear array with the purpose of minimizing the inter-cell interference by a local (decentralized) beamforming matched to the user of interest. Every RAE is connected with an ideal (infinite-capacity) link to a centralized unit (CU) that combines the multiple signals after the beamforming from the radio entities. The goal of this paper is to get insights into the interplay between efficient interference rejection capability of the local beamforming (for large M/Q) and the multicell processing gains arising from coordinating the receptions of multiple RAEs (for large Q).