A Robust Approach to Dynamic Pilot Assignment in Distributed Massive MIMO Networks

We investigate the pilot assignment problem in dynamic distributed massive multiple-input multiple-output (DMIMO) systems, where a large number of access points (APs) are randomly distributed in a large cell and serve a smaller number of users in a coordinated manner. To save resources for channel estimation, a common practice is to consider only the channels with significant channel strengths between users and APs, whilst ignoring those negligible ones. This artificially imposes a graph structure on the network topology for pilot assignment. In a highly dynamic scenario with e.g., high-mobility users, however, the network topology may be varying over time, demanding frequent re-assignments of pilot sequences. In this paper, we consider a topological pilot assignment in a dynamic setting and propose a robust approach leveraging maximum distance separable (MDS) codes on graphs, such that pilots can be assigned once for all, regardless of the time-varying channel strengths. Simulation results of the downlink data rate performance in both static and dynamic DMIMO scenarios demonstrate the superior robustness of our proposed approach.