Research On Joint Reception And Processing Techniques For Multi-Antenna Signals

With the rapid development of modern wireless communications, the requirement of lower transmit power, higher data rates and higher spectral efficiency are fast growing. The varying wireless environment, complex network and continuously reducing of signal energy lead to more severe challenges. In the tradition receiving of a single signal, the processing of the synchronization parameters are disconnected, as a main direction of the research, the joint estimation of multi-parameters are adopted to improve the performance. At the same time, multi-receiver array is an efficient scheme of combining the processing of each data stream, and shows great advantages in the scenarios of linear diversity combining, low earth orbit satellites communications and antenna arraying systems in deep space network.In this paper, the main research is the joint processing of multiple parameters and multiple signals. The analysis of joint estimation of synchronization parameters and symbols, joint synchronization, joint estimation of channel parameters an joint symbol detection of multi-receiver signals are proposed, which can be summarized as follow:1. For the problem of joint estimation of symbol timing and symbol sequence with short data package and low signal-to-noise-ratio, an algorithm of ML symbol detection without timing recovery is proposed. Under the generalized EM scheme, the ML detection is achieved iteratively by direct computation of the matched filter output. Based on the theoretical results, the expression of estimate and a new detection scheme are present, which eliminate the need for optimal samples recovery in traditional schemes. Simulation results shows that the proposed detector enables significant improvement on the bit-error rate (BER) for short data packages and weak received signals.2. For the problem of symbol detection without synchronization information, a ML based algorithm of symbol detection is proposed without the estimation of synchronization parameters. Different with the traditional separated processing of each parameter, all the synchronization parameters such as symbol timing, carrier frequency and carrier phase are trade as the missing information under the generalized EM scheme. Based on the theoretical results, the closed-form expression is proposed, which leads to much decrease of computational complexity compared to the well-known EM based joint estimation of synchronization parameters. Simulation results show obvious improvement in both bit-error rate and convergence.3. Based on the multi-receiver signals, a scheme of joint synchronization and joint ML symbol detection is proposed. In traditional signal combining schemes, array alignment and synchronization stages are achieved separately. In contrast, a new joint ML symbol detection scheme based on multiple receivers without the degradation may be introduced by combining loss and synchronization errors. The transmitted symbol sequence is estimated directly on the oversampled signals from independent receivers, neither signals calibration nor synchronization parameters need to be performed. There are substantial decreases in complexity over implementation and significant improvements in BER performance with low SNR in a large array.4. Based on the receivers with unequal apertures, the joint detection of symbols are achieved by solving the joint ML estimation problem of symbol sequence and multiple SNR parameters of multi-receiver signals. Based on the generalized SAGE (Space-Alternating Generalized EM) scheme, a new iterative scheme is proposed to estimate the multiple SNR parameters and symbol sequence jointly. The joint scheme do not need to establish synchronization first compared to the traditional decision-directed method, and take more advantages by the using of symbol information than the signal-wave correlation based methods. The proposed scheme shows improvements both in SNR estimates and bit-error rate through simulation results.5. For the problem of symbol detection under multi-path fading channel, a new joint scheme of timing synchronization, channel parameters estimation and symbol detection of multi-receiver signals is considered. The detection problem is modeled as joint ML estimation of channel parameters and symbol sequence when the path delay is unknown. An iterative scheme based on EM algorithm is proposed to solve the joint ML estimates through calculating the expecting of path delay by oversampled signals. And the BER performance and that of employing multi-receiver are analyzed through simulation.