Fine Timing Synchronization Based on Modified Expectation Maximization Clustering Algorithm for OFDM Systems

摘要

A novel fine timing synchronization method based on the modified expectation-maximization (EM) clustering algorithm is proposed for orthogonal frequency-division multiplexing systems. Using the cross-correlation metrics of one preamble symbol, the cross-correlation peaks corresponding to the channel arriving paths are identified by the proposed modified EM clustering algorithm, the position of the first coherent cross-correlation peak is then chosen as the start of the frame. Computer simulations show that the proposed method is robust in multipath dispersive channels and achieves superior performance to existing techniques in terms of timing accuracy.