An EM-based OFDM receiver for time-variant channels

OFDM modulation combines the advantages of high achievable rates and relatively easy implementation. However, for proper recovery of the input, the receiver needs accurate channel information. In this paper, we propose an expectation-maximization (EM) algorithm for joint channel and data recovery. The algorithm makes use of the rich structure of the underlying communication problem - a structure induced by the data and channel constraints. These constraints include pilots, the cyclic prefix (CP), and the finite alphabet constraints on the data, and sparsity, finite delay spread, and the statistical properties of the channel (time and frequency correlation). Channel identification and equalization is performed optimally and recovery is achieved within the same OFDM symbol using an EM based Kalman filter.