Regression model based channel estimation for OFDM with multiple transmitters

Abstract

Based on a nonlinear two-dimensional (2-D) regression model, we analyze the OFDM system performance with respect to the operation block size and pilot density. Specifically, we propose a regression model based channel estimation for the OFDM system with multiple transmit antennas. To obtain the proper estimation, we introduce the guidelines for designing pilot values and an expectation-maximization (EM) based solution with lower computation complexity.