Fractionally Spaced Momentum Blind Equalization Algorithms And Simulink Realization

Training sequence needs to be sent to the traditional adaptive equalization technique to track and gather changing channel’s parameter, reducing the bandwidth efficiency of underwater acoustic communication. However, the blind equalization can greatly improve the efficiency and reliability of data transmission, without sending training sequences. Fractionally spaced blind equalization algorithm uses over-sampling of the signal to obtain detailed channel parameters, so its performance is better than traditional constant modulus algorithm. In this paper, in-depth research and analysis constant modulus blind equalization algorithms and fractionally spaced algorithms is done and then the fractionally spaced momentum blind algorithm is researched within its Simulink realization, the main work is as follows:(1) According to the disadvantages of traditional constant modulus algorithm, slow converges rate and high MSE, CMA equalizer tap weights vector uses momentum algorithm to improve its convergence speed.(2) For constant modulus algorithm’s high MSE, fractionally spaced momentum blind equalization algorithms is proposed. Fractionally speed blind equalization can over-sample the input signal, significantly reducing the possibilities of frequency aliasing and steady-state error caused by under-sampling.; momentum algorithm can accelerate the convergence rate. Based on the previous knowledge, the impact of the tap spacing on the performance of the proposed algorithm has been studied.(3) Fractionally spaced decision feedback momentum algorithm has been proposed according to the disadvantage of fractionally spaced blind equalization algorithms. The slow convergence has been found in the application of fractionally spaced blind equalization momentum algorithm’s equalization on spectrum zero channel. Decision feedback equalization algorithm can send the signal after decision to the feedback filter and eliminate the impact after estimating from the detected symbol estimates on the current symbol, ultimately reducing inter-symbol interference. The new algorithm can effectively reduce the steady-state error of fractionally spaced momentum blind equalization algorithms, improving the algorithm performance.(4) Because of slow converging rate and high MSE of single fractionally spaced momentum bind equalization algorithm, fractionally spaced momentum cascade blind equalization algorithm has been studied The algorithm connects two types of fractionally spaced blind equalization algorithm together. The first stage utilizes T/2fractionally spaced momentum blind equalization algorithm, while the second stage applies T/4fractionally spaced momentum blind equalization algorithm. Compared with single fractionally spaced momentum blind equalization algorithm, the fractionally spaced momentum cascade blind equalization algorithm has significant advantages, such as convergence rate, low mean square error. Even under a complex environment, the signal equalization has good compliance.(5) Established Simulink simulation model of proposed algorithm. Blind equalization algorithm is for the purpose of reducing the signal’s inter-symbol interference and improving the reliability of communication system. However, proposing a theory is not enough to solve practical problems. In order to enhance the effectiveness of the proposed algorithm,a constant modulus algorithm based on CMA Simulink simulation model has been established, which contains T/2fractionally spaced momentum blind equalization algorithm, T/2fractionally spaced decision feedback momentum blind equalization algorithm and fractionally spaced momentum cascade blind equalization algorithm Simulink models.