Analysis And Algorithm Study On Equalization Of MIMO Systems

Modern communication system is now developing very fast and has gotgreat success in both technology and commercial field. But there are still manytechnical problems to be solved for the dream of real time transmitting ofmultimedia service composed of texts, voice and video. One of these technicalproblems is high speed wireless access. Many wireless technologies have beenproposed, the most attracting one is MIMO technology, which uses arrayantenna at both the transmitting and receiving end. Researches indicate that theMIMO can improve the performance of wireless system prominently, such assystem capacity, data rate, and frequency efficiency. To develop a MIMOsystem, exact system equalization is necessary. MIMO channel are morecomplex than SISO channel, this leads to the complexity of the receiver.This paper focuses on the research of base-band signal processing inwireless MIMO systems, which include equalization algorithms design and theirapplications in other communication systems. The target of our work is todevelop effective method to overcome MUI and ISI, at the meantime, recovermultiple transmit information.For the MIMO equalization using training sequence, we utilized MMSEcriterion and Kalman filtering scheme. Based on MMSE, a DFE equalizationstructure is modeled into an adaptive filtering structure, then conjugate gradientalgorithm is introduced, which achieve a tradeoff between convergence rate andcomputation complexity. Besides, we modified conventional Kalman method,enhanced the computation accuracy due to computational error, whichimplement more precise channel estimation.For blind equalization technology, we extend CMA from SISO system toMIMO system. Two novel CMA criterions are presented, which overcomes theone-to-many problem effectively. Further more, joint channel estimation anddetection is analyzed, a new method is proposed based on CMA, whichovercome one-to-many problem in another perspective and also achieve afavorable results.MIMO equalization algorithms are also can be applied to other systems.Two other space-time architectures are analyzed and modeled into conventionalMIMO model. In DS-CDMA systems, uplink multi-user detection using singleantennal and multiple antennas are both analyzed, and new multi-user detectionmethods are proposed, which make MIMO equalization algorithm a more broadapplication scenarios. In MMIMO-OFDM system, only the data model isconstructed. Due to the inefficient time, no further algorithm is presented.However, the model makes great sense to take use of mature equalizationmethods.All the developed methods have been simulated using computer. Simulationresults have showed the affectivity of these methods. The equalizationalgorithms can mitigate MUI and ISI, recover multiple antenna signals. Themulti-user detection methods can effectively recover all the users' signals whilecombat other users' interference. All of these show a great promise of the futureof MIMO technology.