Research On Blind Equalization Algorithm Based On Multi-Modulus Algorithm

In underwater acoustic communication system, inter-symbol interference caused by signal multipath propagation are the main obstacles in the development of high-speed underwater acoustic communication system, blind equalization technology is developed in order to eliminate inter-symbol interference caused by the channel. Multi-modulus algorithm has a faster convergence rate and lower steady state mean square error compared to the constant modulus algorithm, it has a better equalization performance for high-order QAM signals, and used in high-speed underwater acoustic communication. This paper studies blind equalization algorithm based on multi-modulus algorithm in high-speed underwater acoustic communication, the main work is as follows:(1)By studying the principle of the conjugate gradient algorithm and analyzing the conjugate gradient algorithm’s advantages in improving the algorithm’s convergence speed, we get the SCGMMA by introducing it into the multi-modulus algorithm; Secondly, we research the influence of the momentum term on the convergence performance of the algorithm, introduce it into the SCGMMA and obtain the MSCGMMA; Thirdly, we analysis the influence of the algorithm caused by the momentum factor varyly, constructe a function between the momentum factor and the mean square error, and obtain the DMSCGMMA on this basis.(2)By studying the principle of the weighted multi-modulus algorithm, we constructe a non-linear function between the weighting factor and the mean square error by analyzing the influence of the algorithm’s performance caused by weighting factor, and obtain the IWMMA; introducing the DD into the IWMMA by considering the dual-mode algorithm and the hybrid algorithm, we get the DIWMMA and the HIWMMA respectively, and compare the two new algorithms.(3)By studying the characteristics of the square QAM signal and the law of the Gaussian distribution, we obtain the GMMA by leading the Gaussian distribution into the MMA; analyzing the influence of the algorithm’s performance caused by the Gaussian distribution, we constructe a dynamic function of the Gaussian distribution parameters, and get the DGMMA.