On the Sum Spectral Efficiency of Dynamic TDD-Enabled Cell-Free Massive MIMO Systems

Abstract

We examine the sum spectral efficiency (SE) performance of a cell-free massive multiple-input multiple-output (CF-mMIMO) system, where each access point (AP) can operate either in the uplink or downlink mode in each slot, corresponding to dynamic time-division duplexing (DTDD) across the APs. We derive the sum SE of the system under a weighted combining of the signals received at the central processing unit. We show that the sum SE is a sub-modular function of the subset of active APs. We exploit this to develop a novel, low-complexity, greedy algorithm for choosing the mode of operation of the APs which is guaranteed to achieve within <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">$(1-1/e)$ </tex-math></inline-formula> of the sum SE attained via a full-complexity brute-force search. Our results show that DTDD with greedy AP mode selection can nearly double the sum SE compared to a TDD based CF-system where all APs operate in the uplink or downlink modes simultaneously. Thus, it is a promising duplexing scheme for beyond 5G communications.