Joint pilot assignment and resource allocation in multicell massive MIMO network: Throughput and energy efficiency maximization

Abstract

In this paper, we study the joint pilot assignment and resource allocation for sum-rate (SR) (or throughput) and system energy efficiency (SEE) maximization in the multi-user and multi-cell (MU-MC) massive MIMO network. We consider the pilot contamination effect in the considered problems, which is known to fundamentally limit the network performance, and we jointly optimize the number of activated antennas together with power allocation and pilot assignment. We transform the problems into a subtractive optimization form based on which we develop efficient algorithms to solve both SR and SEE maximization problems. Specifically, we decompose each considered problem into two subproblems, namely optimization of power and number of antennas in the first sub-problem and pilot assignment in the second one, and iteratively solve both sub-problems until convergence. To solve the first sub-problem, we employ a successive convex approximation (SCA) technique to attain a solvable convex problem. In addition, we propose a novel iterative low-complexity algorithm based on the Hungarian method to solve the pilot assignment sub-problem. Numerical studies are conducted to illustrate the convergence of the proposed algorithms, impacts of different parameters on the total SR and SEE, and significant performance gains of the proposed solution compared to the case with conventional pilot assignment.