A High Throughput Pilot Assignment Strategy with Low Complexity in Multi-Cell Massive MIMO Systems

摘要

This paper investigates a massive multiple input multiple output (MIMO) system utilizing time division duplexing (TDD) for transmission and reception. In the training phase, acquisition of the channel state information (CSI) undergoes significant deviation caused by the pilot contamination effect. In order to remedy this impact, a low-overhead iterative pilot assignment strategy is devised in which both overhead order and complexity order are significantly reduced in comparison with the conventional schemes. To this end, initially, the problem of pilot assignment is formulated by an optimization problem whose optimization problems are continuous. In the following, it is attempted to approximate the resultant problem by an optimization problem with discrete variables. Finally, a convex realization of the outgoing optimization problem is presented, where the global optimum is insured. Numerical results reveals that although the proposed pilot assignment strategy undergoes much smaller complexity as well as a lower overhead compared to the exhaustive search algorithm, the system confronts a minor performance loss.