Comparative Performance Analysis of OMP and SABMP for Massive MIMO OFDM Channel Estimation

摘要

The orthogonal matching pursuit (OMP) algorithm and Support Agnostic Bayesian matching pursuit (SABMP) algorithm for sparse channel estimation are reported in the literature. In this paper, the performance of these algorithms is analyzed for massive MIMO OFDM channel estimation. The greedy approach based SABMP algorithm orders its metric updates recursively to determine pre dominant sparsesupports to find MMSE estimate. Simulations are carried out to evaluate the normalized mean square error(NMSE) of OMP and SAMP for 5×5 MIMO with 32 pilot carriers. Further, the NMSE performance of SABMP algorithm is demonstrated for antenna arrays of size(5 × 5, 10 × 10, 20×20) with 16 pilot carriers and varying sparsities (3,7), and QAMs (4,16).