| In the information age, modern communication systems and network applications have been growing rapidly to provide us with more versatile and higher bit rate services such as multimedia on demand (MOD), wireless local area network (LAN), 3G networks, and high-speed internet. In order to maintain the qualities of services, how to design a robust transceiver is a crucial issue. Symbol timing offset and channel response are two important parameters for signal reconstruction. A nonzero symbol timing offset at the receiver affects the baseband signal processing unit to process the sequence in a wrong order; therefore, errors occur during the decoding process, and the original information needs to be retransmitted. In this thesis, two timing synchronization and channel estimation methods are first proposed for single-input training-sequence-based wireless communication systems. One is the frequency-domain approach, and the other is the time-domain approach. Then, we extend the time-domain approach to solve the same problem in communication systems with multiple transmit antennas. After these two synchronization methods are discussed, two variable transmission rate (VTR) orthogonal frequency division multiplexing (OFDM) based communication systems using network source coding are presented. First, a new semiblind frequency-domain timing synchronization and channel estimation scheme based on unit vectors is proposed. Compared to conventional methods, the proposed approach has excellent timing synchronization performance under several channel models at signal-to-noise ratio (SNR) smaller than 6dB. In addition, for a low-density parity-check (LDPC) coded single-input single-output (SISO) OFDM-based communication system, our proposed approach has better bit-error-rate (BER) performance than conventional approaches for SNR varying from 5dB to 8dB. Second, a novel joint timing synchronization and channel estimation algorithm for wireless communication systems based on the time-domain training sequence arrangement is proposed. From the simulation results, the proposed approach has excellent timing synchronization performance under several channel models at low SNR which is smaller than 1dB. Moreover, for an LDPC coded 1x2 single-input multiple-output OFDM-based communication system with maximum ratio combining, a comparison BER of less than 10.;{-5} can be achieved usingour proposed approach when SNR exceeds 1dB. Third, a joint timing synchronization and channel estimation scheme for communication systems with multiple transmit antennas based on a well-designed training sequence arrangement is proposed. Simulation results show that the proposed approach has excellent timing synchronization performance under several channel models at SNR smaller than 1dB. Furthermore, the proposed approach has excellent channel estimation performance in 2x2 and 3x3 multiple-input multiple-output systems. Finally, two VTR OFDM-based communication systems that exploit network source coding schemes are proposed, and the system performance characteristics of these two proposed VTR OFDM-based communication systems are evaluated. The proposed 3-stage encoder in the VTR SISO OFDM-based communication system provides three different coding rates from 0.5 to 0.8. As for the proposed VTR multi-band OFDM-based communication system, two correlated sources are encoded by different coding rates from 0.25 to 0.5. Furthermore, compared with a traditional uncoded OFDM system, the proposed VTR OFDM-based communication systems have at least 1 to 4 dB gain in SNR to achieve the same symbol error rate in an additive white Gaussian noise channel. |
