A Spectrally Efficient ML Estimation Method for OFDM Systems over Sparse Multipath Channels

For getting satisfactory error rate performance, channel estimation is required at receiver. Maximum likelihood (ML) estimation method is preferred due to its low complexity. The conventional ML estimation for sparse multipath channels requires a large number of pilot symbols and thus increases the spectral inefficiency. This paper investigates an efficient way to apply pilot aided ML estimation technique in orthogonal frequency division multiplexing (OFDM) systems to estimate the channel impulse response (CIR) of sparse multipath channels. An algorithm which exploits channel sparsity is proposed to carry out the ML estimation using lesser number of pilot symbols as compared to the conventional method. The symbol error rate performance (SER) of OFDM system assisted by the proposed scheme is analyzed. Simulation results show that significant reduction in spectral efficiency loss can be achieved with small increase in required SNR.