A Framework for Decentralising Multi-Cell Cooperative Processing on the Downlink

Abstract

Multi-cell cooperative processing (MCP) is well acknowledged for significantly improving spectral efficiency and fairness amongst users. Nevertheless MCP comes together with some shortcomings which have also been recognised. Cooperating Base Stations (BSs) need to be inter-connected via a Control Unit (CU) which gathers channel state information (CSI). CU is the entity responsible for performing MS scheduling and defining the transmission parameters. Under a practically feasible linear preceding framework, CU performs user scheduling and designs the beamforming matrix for the chosen users. Therefore MCP implementation adds a very significant infrastructure cost as it requires a CU and low latency links connecting cooperating BSs with it. Furthermore this entails an important protocol complexity in order for all this signaling to be coordinated. In this paper a new framework is proposed that allows MCP on the downlink without the need of costly modifications on the infrastructure comparing to the existing cellular networks. The needed operations are performed in a distributed fashion by the cooperating BSs and this is shown to be a good alternative facilitating MCP implementation.