Varying Step Dual-mode Blind Equalization

In wireless digital communication systems, multipath fading often leads to serious inter-symbol interference (ISI) and inter-channel interference (ICI). To reduce the error rate and improve the communication quality, using blind equalization techniques compensates for changes in the channel and eliminates inter-symbol interference. Blind equalization without the aid of a training sequence is the new transmit sequence approach adaptive equalization which output sequence is close to input sequence only using prior information to accept the sequence itself, to balance the channel characteristics. So in the wireless data communication system without having to send a training sequence, we can achieve the purpose of improving channel efficiency, while the blind equalization technique can effectively compensate for the non-ideal characteristics of the channel, reducing the error rate, overcome inter-symbol interference, improve the communication quality, and get better balanced properties. Blind equalization because of its superior performance becomes key technology in the era of digital communications technology, and the technology is the focus and hotspot of research in the field of modern communication. To this end, studying blind equalization deeply is very important theoretical siginificance and practical value. In this paper, The technology of blind equalization, background, history, current situation and its significance are reviewed and summarized. Based on the principle in the analysis and classification of blind equalization, the basic algorithm for blind equalization was studied with the method of optimization theory.First, the basic nature and the basic principles of Bussgang blind equalization algorithms were stated briefly and the four common kinds of Bussgang blind equalization algorithm were introduced. The constant modulus algorithm (CMA) was conducted a study and was analyzed in detail. The main factors that they restrict the convergence performance of constant modulus algorithm to further improve were pointed out by analyzing. Secondly, the influence on the variable step size factor to the convergence properties of CMA was stated. Then, It was suggested that a new variable step-size CMA blind equalization algorithm. Through theoretical analysis and computer simulation, and concluded that the new algorithm eases the contradiction between the convergence speed and the convergence precision in the traditional CMA algorithm with fixed step size.Finally, In order to improve the blind equalization of the steady-state error, convergence rate and convergence precision, the modified dual-mode blind equalization algorithm is put forward here. Dual-mode blind equalization algorithm regards blind equalization algorithm of CMA as a cold start. When the decision error rate achieves low enough level, DD-LMS algorithm was switched to. The variable step algorithm of constant modulus algorithm (CMA) and decision directed (DD) algorithm are combined by the modified dual-mode blind equalization algorithm. Both of the algorithms(CMA and DD) hard switching of blind equalization algorithm were implemented by a ruling circle, while the introduction of variable step size factor. Then the simulation results show that the algorithm has fast convergence rate, and higher convergence precision.