Skip to Main Content
In this paper, a joint channel, carrier-frequency-offset (CFO) and noise-variance estimation scheme is proposed for OFDM systems which is based on Expectation and Maximization (EM) algorithm. The channel parameters are estimated using training sequences incorporated at the beginning of each transmission frame. Based on the assumption that the amplitude and CFO of different paths are independent, the received multipath components may be decomposed into $L$ independent data sets of the $L$ resolvable propagation paths. Hence the associated multi-dimensional minimization problem may be decomposed into separate single-dimensional minimization processes, the maximum likelihood and yet, remains capable of approaching performance at a signifucantly reduced complexity.