Abstract
In this paper, we propose a novel pilot aided joint estimation technique for carrier frequency offset (CFO) and doubly selective (both time and frequency selective) channel (DSC) in high-mobility multiple-input-multiple-output orthogonal frequency-division multiple-access (MIMO-OFDMA) uplink systems. As the maximum-likelihood (ML) solution to this estimation problem is highly computationally tedious due to the requirement of multidimensional optimization, we propose an iterative scheme based on the oblique projection (OP), which separates the user signal in each iteration, resulting in decomposition of highly complex multidimensional search to many 1-D searches. We also derive a low-complexity approximation to the proposed technique, which makes use of precalculated projection matrices stored at the base station. The exact Cramer-Rao bound (CRB) of the joint estimation of CFOs and DSCs in MIMO-OFDMA uplink is also derived. Simulation studies indicate that the mean square error performance of the proposed techniques nearly achieves the CRB even at low signal-to-noise ratios (SNRs).