Research On Subspace-based Blind Channel Estimation Methods For MIMO System

Multiple-input multiple-output is a technique which is achieved by using multiple antennas at both the transmitter and the receiver to increase the capacity, the range as well as the reliability of wireless systems. As a vital part of wireless communication systems, channel equalization takes advantage of the channel state information to reverse the distortion incurred by a signal transmitted through a channel. Since the wireless channel is unknown and time-varying, it is required to estimate the channel state information beforehand. Traditionally, the channel is estimated by transmitting training sequences periodically, which not only has a penalty in terms of bandwidth efficiency, but also restricts the channel estimation algorithm available to some scenarios unable to send training sequences. Therefore, blind or semi-blind channel estimation methods are of great interest as they save bandwidth by avoiding or reducing the use of training sequences.This thesis focuses on the subspace-based blind channel estimation for MIMO systems. The main works are summarized as follows:1. Existing blind channel estimation methods for MIMO systems are studied and classified according to the statistical properties being used. The advantages and disadvantages of these methods are analyzed as well as their application scenarios. And on this basis, the predominance of subspace-based methods appears.2. Channel order estimation for SIMO systems is studied. This paper firstly summarizes the traditional channel order estimation methods, and then proposes a novel algorithm based on GDE. It is shown that compared with traditional ones, the proposed method is less sensitive to variations in the SNR, and able to work with small data sets and channels with small head and tail taps.3. Subspace-based blind channel estimation method for SIMO systems is studied. In order to overcome the shortcomings of the existing subspace-based methods, a modified method is proposed, which is robust to channel order overestimation.4. Subspace-based blind channel estimation method for MIMO-STBC systems isstudied. We firstly establish conditions for channel identifiability and present a blind channel estimation technique based on a subspace approach. Besides, in order to eliminate the ambiguity, we realize the complete channel identification using semi-blind methods.5. Ambiguity is the inherent problem in blind channel estimation. Since it cannot be solved in a fully blind way, most of the study about blind channel estimation did not pay enough attention to this. However, the remaining ambiguities of the estimated results will lead to the failure of decoding. In this paper, to put emphasis on this issue, we study and analyze the forms of ambiguity in different communication systems.