Pilot Clustering in Asymmetric Massive MIMO Networks

TL;DR

This paper introduces a coalitional game theory approach to optimize pilot allocation in asymmetric massive MIMO networks, improving spectral efficiency by reducing pilot contamination.

Contribution

It proposes a novel distributed algorithm for coalition formation among cells, enhancing pilot reuse strategies in practical deployments.

Findings

Fast convergence of the proposed algorithm

Significant performance gains over traditional methods

Effective coalition structures for pilot allocation

Abstract

We consider the uplink of a cellular massive MIMO network. Since the spectral efficiency of these networks is limited by pilot contamination, the pilot allocation across cells is of paramount importance. However, finding efficient pilot reuse patterns is non-trivial especially in practical asymmetric base station deployments. In this paper, we approach this problem using coalitional game theory. Each cell has its own unique pilots and can form coalitions with other cells to gain access to more pilots. We develop a low-complexity distributed algorithm and prove convergence to an individually stable coalition structure. Simulations reveal fast algorithmic convergence and substantial performance gains over one-cell coalitions and full pilot reuse.