The Study And Improvement On Blind Equalization Algorithm For MIMO System

Now, the MIMO system has been a key technology to the next-generation mobile communication system and has been used widely in the modern telecommunication.The MIMO system has its advantage of making full use of space resources, so it will multiply improve the communication capacity and the spectral efficiency without increasing the bandwidth and the power of transmit antenna. However, because of its special structure, the channel will show a frequency selective fading which will cause the inter-symbol interference at the receiving end. The traditional method of equalization can eliminate this interference. And the blind equalization is a technology which can balance the channel characteristics equalization only depending on the statistical characteristics of the transmitted signal without a training sequence. The aim is to make the output sequence approximate the input sequence. In a word, using the blind equalization technique can improve the communication efficiency and the communication quality in the MIMO system. It has emerged many classical methods and theory in the field of blind equalization algorithm. Though they lay the theoretic foundation for the research of blind equalization algorithm, they have difficult to overcome their own shortcomings. For example, there will be the problem of vulnerable to the curse of dimensionality and the question of local minimum value. The main work of this paper is as follows:(1) This paper focuses on the study of the constant modulus blind equalization for the MIMO systems, and proposes two improved algorithms. These two improvement algorithms are both based on the idea of variable step. One is based on the adaptive variable step size, the other is based on the function of Sigmoid blind equalization. The simulation results show that the two improvement CMA algorithms both enhance the ability of noise-reduction and fast convergence. So these two algorithms are superior to the traditional methods.(2) Support vector machine has been showing a lot of advantages, which is based on the statistical theory. This method can efficiently overcome the curse of dimensionality and the minimum problems. This paper has applied the theory of support vector machine in the blind equalization method and has analyzed the realization of blind equalization in the MIMO system from the points of support vector classification and support vector regression. In the end, the simulation results show that the support vector machine algorithm has lots of advantages when it is applied to the MIMO system.