The Extended Research On Bussgang Blind Equalization Algorithm

Unlike conventional adaptive equalization, Bussgang blind equalization operates without training sequence, which can save bandwidth and means higher date rate, and it has strong robustness and can be easily implemented, so it is widely used. The problems, however, such as slower convergence rate, larger residual error and phase-blind nature may not be sufficient for system to obtain adequate performance. Accordingly, from the aspects of error function, combination algorithm, overcoming the error offset and spatial diversity blind equalization, we make extended research on Bussgang blind equalization algorithm and propose a series of extended algorithms. By. theoretical analysis and computer simulations, the superior performances of proposed algorithms have been proved.The main contributions of thesis can be summarized as follows:1. Analyzing blind equalization algorithms based on different error functions.For overcoming the slower convergence rate and larger steady state error of Constant Modulus Algorithm (CMA), which result from the non-odd symmetry and invariable error function, we make the researches as follows:(1) By defining a hyperbolic tangent error function hasing the characteristic of odd symmetry, a CMA based on the hyperbolic tangent error function is proposed.(2) By defining an error function with variable slope, whose characteristic can be changed according with different values of slope, a CMA based on the variable slope error function is proposed.(3) By defining an odd symmetric and variable segment error function whose characteristic can be changed by adjusting the segment position, a CMA based on the variable segment error function is proposed.2. Analyzing combination blind equalization algorithms.For making full use of the advantage of each Bussgang blind algorithm and eliminating the disadvantage, we do research on three classes combination blind equalization algorithms.(1) Combining the error functions of hyperbolic tangent and Decision Directed (DD) and by means of the decision to control the weigh vector of each part, a blind equalization algorithm based on the combination error function is proposed.(2) Combining the CMA based on the variable slope error function and DD, and by means of the decision to adjust the slope automatically, the dual-mode blind equalization algorithm based on CMA and DD is proposed.(3) By analyzing the resemblance between CMA 1-2 (Sato) and CMA algorithms, we study a new switching method, and then a dual-mode bind equalization algorithm based on CMA 1-2 and DD is proposed.(4) Because the high-order QAM signals are located in different modulus, which results in error signal is not equal to 0 when even the channel is perfectly equalized, a dual-mode bind equalization algorithm based on VSCMA and multi-modulus algorithm is proposed.(5) For eliminating the carrier phase or frequency offset introduced by time-varying channel, a combination blind equalization algorithm with digital phase-locked loop is proposed.3. Analyzing spatial diversity blind equalization algorithms.In case where long-range communication is needed, spatial diversity blind equalizer (SDE) is an effective way to improve SNR and suppress distortion. We study the method of combining Bussgang blind equalization with different diversity combining technique, and proposed a SDE based on modified CMA and multi-modulus algorithm to improve the performance of SDE.