A Low-Complexity Cross-Layer Algorithm for Coordinated Downlink Scheduling and Robust Beamforming Under a Limited Feedback Constraint

Coordinated scheduling/beamforming (CS/CB) substantially mitigates the intercell interference (ICI), hence increasing the cell-edge throughput on the downlink (DL) of coordinated multipoint (CoMP) systems. To maximize the DL throughput, the cooperating base stations (BSs) jointly select the best set of users for DL scheduling and then jointly design a set of beamforming (BF) vectors to approach the throughput limit. However, finding the optimal BF vectors requires an exhaustive search and substantial channel state information (CSI) feedback, hence resulting in high algorithmic complexity and heavy uplink traffic load. Hence, we conceive a new cross-layer algorithm to achieve high performance at a lower feedback amount and at lower algorithmic complexity. Based on the fact that different BSs usually have different traffic loads, we divide the BSs into two different types, i.e., the master BSs (MBSs) and the slave BSs (SBSs), where MBSs have a higher transmission priority than SBSs. The scheduling relies on an interference threshold, whereas our robust BF scheme exploits both the channel direction information (CDI), which is quantized using the technique of limited feedback, and the channel quality information (CQI), which is assumed to be fed back accurately. Our numerical results show that the proposed algorithm does not lose much performance compared with that achieved by an exhaustive search, whereas the algorithmic complexity is as low as that of the solutions operating without CoMP.