Low Complexity Scheduling Algorithms for Downlink Zero-Forcing Beamforming Systems

In this paper, the downlink zero-forcing beamforming strategy under perfect channel state information at the transmitter (CSIT) for the case of random packet arrivals is investigated. Under this setting, the relevant fairness criterion is the stabilization of all buffer queues which guarantees a bounded average delay for all users. In case of queue stability is the goal, it can be shown that allocating resources to maximize a queue-length-weighted sum of the rates which are feasible in the current time slot is a stabilizing policy. However, the high complexity of user selection or the feasible rates determination for optimal scheme may prevent the real-time scheduling operation. Two simplified algorithms were provided taking the channel state, queue state and orthogonality into to account. In particular, we pick the first user with the largest product between channel gain and queueing lengths, and select the remaining N_t-1 users to form the user set based on the greedy user selection method or takeing channel orientations into account as well. Simulation results show that the average delay performance of the simplified algorithms are reduced by less than 10% compared to optimal schemes.