System-Level Performance Analysis of DOA Estimation in Cellular Networks using Stochastic Geometry

Abstract

In this paper, we analyze the error in the direction of arrival (DOA) estimation using the multiple signal classification (MUSIC) algorithm for a cellular network with the random deployment of BSs. The error in DOA estimation is due to multiple factors such as the channel noise, the accuracy of the covariance matrix, the coherency of incident signals and low signal to noise ratio (SINR). We consider a downlink cellular network scenario where base stations (BSs) locations are deployed according to a homogeneous Poisson point process (PPP). We consider a typical user located at the origin. We assume that the user is equipped with a uniform linear array (ULA) of $L$ isotropic antenna elements. We develop a framework to analyze the system level performance of the DOA estimation for the considered model. We also derive simple upper and lower bounds for the mean squared error (MSE) in DOA estimation and termed them as the pessimistic and the optimistic bounds, respectively. We also perform numerical simulations to validate our results.