An efficient maximum likelihood CFO estimation scheme for MIMO-OFDM systems

In this paper, we address the estimation of carrier frequency offset (CFO) in orthogonal frequency division multiplexing (OFDM) based multiple input multiple output (MIMO) communication systems. One of the novelties of the proposed approach is the judicious combination of two methods so as to reduce the bandwidth overhead and computational complexity without compromising the CFO estimation range. Unlike many of the existing techniques, the proposed CFO estimator does not require MIMO channel estimates as a prerequisite, that are difficult to obtain in practise. It employs the cyclic prefix and a few null subcarriers respectively for the fractional and integer frequency offset estimations. We also propose a novel null subcarrier allocation scheme based on Fibonacci series. The range of frequency offset that can be estimated by the proposed technique is equal to the full OFDM bandwidth. Furthermore, performance of the proposed frequency offset estimator is mathematically analyzed by deriving the Cramer-Rao lower bound and compared it with the mean square estimation error to demonstrate its efficiency. It has been shown that our approach greatly outperform a recently published method in terms of BER performance and computational complexity.