Research On Analog Space-Time Codes For Ultra-Wideband Communications

Impulse radio ultra-wideband (IR-UWB) communications and space-time coding (STC) have drawn a lot of attention in recent years. IR-UWB gains polularity mainly for its attractive features: improved multiple-access capacity, low-power low-complexity basedband operation and the potential for high data throughput over short distances. On the other hand, STC is an effective transmit diversity technique to boost the performance. Some efforts show that, incorporations of space-time block codes (STBC) into UWB transmissions can enhance the performance in bit error rate (BER) and immunity against timing jitter significantly. However, exsiting reseaches are far from complete. Thus, this dissertation is dedicated to the STBC tailored for UWB communications: analog space-time codes (encoding pulses within symbol waveforms). The main contribution is a systematic resolution including timing acquisition, blind channel estimatin, subspace detection and performance enhancement.Due to the fact that the information bearing waveforms are low-power ultra-short pulses, synchronization poses one of the critical challenges in UWB communications. To overcome this difficulty, a low-complexity rapid timing acquisition algorithm based on training scheme is proposed. Exploiting the special correlation pattern among received waveforms, the proposed algorithm entails simple intergrate-dump operations per symbol, thus reduces complexity and improves acquisition speed.The coherent reception of analog space-time codes entails channel information at the receiver. While training data can be used to estimate the channel, this approach will reduce throughput and/or incease the complexity of transceivers. Furthermore, a loss in performance is always incurred by the error of channel estimates. So it is necessary to search for efficient noncoherent reception methods. Firstly, a computationally simple blind channel estimation method is proposed. Specific properties of analog space-time codes are exploited to estimate the channel matrix, which subsequently used in the maximum likelihood (ML) receiver to recover the symbols. Secondly, subspace detection methods of analog space-time codes are presented. Without the need of channel estimation, the proposed approaches yield the estimation of transmitted symbols by minimizing some quadratic form built on the orthogonality between signal and noise subspaces. In order to reduce the computation complexity, the equivalent linear receivers are also given.When the channel state information is available at the transmitter, the pre/post-rake structure can be implemented to enhance the performance in single-input single-output (SISO) systems. Inspired by the above observation, two pre/post-rake schemes (scheme 1,2) wedded with multi-antenna UWB transmissions are proposed. The combing weights of the pre-rake and post-rake are derived by maximizing the SNR at the receiver. Furthermore, sheme 2 are generalized to obtain singular value decomposition (SVD) multi-antenna pre/post-rake, which focuses on maximizing the channel capacity.