Distortion-Tolerant Communications with Correlated Information

This dissertation is devoted to the development of distortion-tolerant communication techniques by exploiting the spatial and/or temporal correlation in a broad range of wireless communication systems under various system configurations. Signals observed in wireless communication systems are often correlated in the spatial and/or temporal domains, and the correlation can be used to facilitate system designs and to improve system performance. First, the optimumnode density, i.e., the optimum number of nodes in a unit area, is identified by utilizing the spatial data correlation in the one- and two-dimensional wireless sensor networks (WSNs), under theconstraint of fixed power per unit area. The WSNs distortion is quantized as the mean square error between the original and the reconstructed signals. Then we extend the analysis into WSNs with spatial-temporally correlated data. The optimum sampling in the space and time domainsis derived. The analytical optimum results can provide insights and guidelines on the design of practical WSNs. Second, distributed source coding schemes are developed by exploiting the data correlation in a wireless network with spatially distributed sources. A new symmetric distributed joint source-channel coding scheme (DJSCC) is proposed by utilizing the spatial source correlation. Then the DJSCC code is applied to spatial-temporally correlated sources. The temporalcorrelated data is modeled as the Markov chain. Correspondingly, two decoding algorithms are proposed. The first multi-codeword message passing algorithm (MCMP) is designed for spatially correlated memoryless sources. In the second algorithm, a hidden Markov decoding process is added to the MCMP decoder to effectively exploit the data correlation in both the space and time domains. Third, we develop distortion-tolerant high mobility wireless communication systems byconsidering correlated channel state information (CSI) in the time domain, and study the optimum designs with imperfect CSI. The pilot-assisted channel estimation mean square error is expressedas a closed-form expression of various system parameters through asymptotic analysis. Based on the statistical properties of the channel estimation error, we quantify the impacts of imperfect CSI on system performance by developing the analytical symbol error rate and a spectral efficiency lower bound of the communication system.