Generalized iterative water-filling for MIMO MAC with mixed constraints (invited)

Based on the efficient Generalized Water-Filling with Group peak Power constraints (GWFGP), an iterative algorithm to compute the optimal solutions to system throughput (sum-rate) maximization problems for the multi-user multiple input multiple output multiple access channels (MU-MIMO MAC) in the general communication systems or networks is proposed. The proposed iterative GWFGP algorithm (IGWFGP) has two layers of loops. An inner loop aims at computing the solution to each member in the family; while the outer loop aims at computing the solution to the target problem based on the results obtained by the inner loop. Both GWFGP and the convergence theory of an algorithm are used in the inner loop and the outer loop respectively. Furthermore, by exploiting the concept of variable weighting factor for covariance update, IGWFGP owns fast convergence and provides optimal solutions for the sum rate maximization problems. The used convergence theory of IGWFGP and the algorithm of GWFGP are efficient and novel. To the authors´ best knowledge, no algorithms have been reported in the open literature to solve the target problems of this paper. In addition, the proposed algorithm never requires to choose the initial point for iteration, unlike popular algorithms, including the interior point method. This point may avoid the burden to set up an initial point, especially as the system becomes more and more advanced.