Queue aware precoder design for space frequency resource allocation

The paper considers coordinated multi-cell multi-user multiple-input multiple-output (MU-MIMO) transmission using orthogonal frequency division multiplexing (OFDM) technique for downlink channels. Linear beamforming and receiving are employed at BS and corresponding users, respectively. The design criterion is to minimize the number of queued packets, since the transmission is mainly guided by the backlogged packets. An indirect way to solve this problem is to associate the weights of the traditional weighted sum rate maximization scheme with the current status of the queue size, i.e., the longer queue, the higher priority. We deal with this problem directly by formulating it as a noncovex optimization problem and then applying a sequential convex approximation method to find beamformers. In particular, we propose an efficient resource allocation scheme based on jointly optimizing beamformers over the space and frequency domain. We refer to this scheme as queue minimizing (QM) joint space-frequency resource allocation (JSFRA) scheme. The proposed solutions are compared to the traditional queue weighted sum rate maximization (Q-WSRM) approaches mentioned above in terms of the rate of convergence and the backlogged packets remaining after a scheduling instant.