A Semi Blind Joint CFO Estimation, Equalization and Data Detection in Presence of Non-Linearity for mm-Wave Communications

Millimeter wave (mm-Wave) is an emerging paradigm towards 5G technology that can support high data rate. The foreseen potential of mm-Wave is limited by huge path loss incurred due to the high frequency operation which can be alleviated by high emission power at the transmitter. This in concurrence with the enormous bandwidth of mm-Wave and high frequency design limitations of the integrated circuits enforce the power amplifier (PA) into non-linear region. Further, the non-linear distortion in collusion with frequency selective channel and carrier frequency offset (CFO) degrade the signal detection performance. To solve this problem, we propose a semi-blind joint estimation of CFO and frequency selective channel gains followed by data detection in the presence of PA non-linearity. The presence of non-linearity results in the posterior probability distribution of complex data symbol to be non-Gaussian and hence, analytically intractable. Therefore, sequential importance resampling based particle filter (PF) is suggested for approximating the intractable posterior distribution of interest by the weighted random probability samples (particles) to detect the data symbols. The detected symbols are then used to jointly update the channel gains and CFO using a novel sequential maximum likelihood (ML) estimation. Extensive simulation results validate the proposed algorithm. This novel scheme enhances the non-linear signal detection performance in presence of CFO and frequency selective channel at the receiver.