Blind and semiblind detection of OFDM signals in the multipath fading channel

This paper considers the problem of blind symbol detection for the orthogonal frequency-division multiplexing (OFDM) system in the multipath fading channel. Unlike [11] that uses a polynomial to model the variation of channel responses (CRs) across subchannels, we employ the model of channel impulse response (CIR) which takes into consideration of the correlation between the real and imaginary parts of CRs across subchannels. The branch-and-bound principle is used to solve the nonlinear integer programming problem associated with the blind detection problem. We further propose a method to obtain a better initial sequence in each tree search based on channel prediction, and this leads to the complexity reduction for the blind detection algorithm. Simulation results show that by using the CIR model in blind detection, we can obtain better performance than [11], and the complexity reduction based on a better initial sequence is effective.