Stochastic geometry analysis of multi-user MIMO cellular networks using Zero-Forcing precoding
Résumé
The Equivalent-in-Distribution (EiD)-based approach to the analysis of Single-Input-Single-Output (SISO) cellular networks has recently been introduced [1]. Its rationale relies upon formulating the aggregate other-cell interference in terms of an infinite summation of independent and conditionally distributed Gaussian Random Variables (RVs). This approach leads to exact integral expressions of the error probability for arbitrary bi-dimensional modulations. In this paper, the EiD-based approach is generalized to the performance analysis of multi-user Multiple-Input-Multiple-Output (MIMO) cellular networks, which use Zero-Forcing (ZF) precoding for transmission over Rayleigh fading channels. In the presence of othercell interference and noise, the error probability is formulated in terms of a twofold integral. In interference-limited cellular networks, the mathematical framework simplifies to a single integral expression. The mathematical analysis is substantiated with the aid of extensive Monte Carlo simulations and the performance of multi-user MIMO cellular networks is discussed.