Dynamic Pilot Reuse in Distributed Massive MIMO Systems

摘要

In distributed massive multi-input multi-output (DM-MIMO) system, the number of users simultaneously served is greatly restricted if the pilots allocated for users are orthogonal. In this paper, a dynamic pilot reuse strategy within a single cell DM-MIMO system is proposed in order to reduce pilot overhead. The reuse in this strategy is applied so that maximum average sum-rate is satisfied within limited pilot resources. Specifically, two users in different subcells separated by large distance and meeting a specific data rate level can share the same pilot sequence. To perform this strategy, an expression for signal to interference plus noise ratio (SINR) is first derived for any pair of users who uses the same pilot. Based on this expression, an algorithm is proposed to choose which pairs of users are able to use the same pilot by comparing their potential data rates with a certain threshold. Finally, a method is employed to find the near-optimal threshold that can be utilized in the suggested algorithm to produce the maximum average sum-rate. The simulation results demonstrate that the uplink achievable sum-rate for the proposed strategy is higher than the both cases when no pilot reuse or random pilot reuse are considered.