Blind Equalization For FIR-MIMO Systems With Common Zeros Based On Finite Alphabet

The main interference sources in the wireless communications are Multiple Access Interference(MAI) and Inter-Symbol Interference(ISI), their existence not only restricts the increasing of the transmission rate, but also worses the performance of the system seriously. Multi-user detection can solve the problem of MAI effectively, and using the technology of the equalization can eliminate the ISI. The traditional equalization needs to send training sequence frequently, which will increase the cost of the system. However,the blind equalization can be able to solve this problem effectively without the training sequence, only needing the received signals and the information of the sending signals. Gradually, it has become a key focus in signal and information processing.Based on analysising the blind equalization theory of the FIR-SIMO systems with common zeros, this thesis studies the blind equalization of the FIR-MIMO systems with common zeros based on the finite alphabet. Under the condition that the number of the row of the channel convolution matrix is greater than that of the column, the thesis first gives a sufficient condition for the equalizer's existence for the case that multi-users may have the same delay, and then extends it to the case that multi-users may have possibly different delays. In order to achieve blind equalization, this thesis provides a way to transform the problem of direct blind detection for a sequence into a quadratic programming problem, and then using the genetic algorithm to solve this problem. The simulation results illustrate that the equalizer of this thesis is effective and the direct blind detection algorithm in this thesis is superior than subspace algorithm.The organization of this thesis is as follows. Chapter one briefly describes the background and significance of this theme and main work of this thesis. Chapter two summarizes the correlation knowledge of MIMO systems and multi-user detection, then analyzes the theory of blind equalization for FIR-SIMO systems with common zeros based on finite alphabet. Chapter three studies the blind equalization of FIR-MIMO systems with common zeros besed on the finite alphabet. Chapter four gives the corresponding simulations of FIR-MIMO systems with common zeros. The last chapter presents the summary and prospect of this thesis.