Research On Pre-Equalization Design For Monobit Transmission System

As with the development of mobile communication in multimedia direction, ubiq-uitous and always-on wireless systems in homes and enterprises have received great attention. Great developments have been got in a variety of short distance wireless communication meanwhile. Due to the characteristics of unlicensed and great band-width around60GHz, strong anti-interference ability, and high security, the60GHz millimeter-wave (mm-wave) communication is considered as the ideal choice for the next generation of short distance wireless communication. Meanwhile, the monobit digital receiver can significantly reduce the power consumption and complexity for re-ceiving the large-band signals. Therefore, it is of great practical importance to research on the60GHz mm-wave communication under monobit transmission system.A major aspect that restricts the improvement of the transmission rate is the exis-tence of inter-symbol interference (ISI). One of the effective ways to suppress the effect of ISI is the equalization technology. In this paper, we first investigate the channel es-timation and pre-equalizer design problem in monobit transmission system. With the sparsity of60GHz channel, we use the Orthogonal Matching Pursuit (OMP) in sparse approximation theory to make preliminary estimation for channel state information (C-SI). The feasibility of employing the sparse algorithm is analyzed by study the60GHz channel model. With respect to the estimation error, we propose a joint channel esti-mation and pre-equalization design method based on the Minimum Mean Square Error (MMSE) criteria. Simulation results show that the iterative method can significantly reduce the estimation error and the bit error rate (BER) performance with the derived pre-equalizer is close with that under full CSI.In consideration of the effective realization of equalizer in real systems, we contin-ue to research on the sparse equalizer design problem. With the analysis of the optimal equalizer design, we propose a low complexity algorithm called Forward Selection Al-gorithm (FSA). The algorithm iteratively add the item which maximize the decrease of quadratic constraint into the nonzero set. Then the coefficients in nonzero set are opti-mized. Moreover, we derive the relationship of the iterative number, correlation matrix, signal power and the tolerance of performance degradation. Simulation compares the performance degradation between sparse equalizer and optimal equalizer. Results show that the proposed algorithm has advantage over traditional algorithms in the aspects of complexity and performance.