Semester

Fall

Date of Graduation

2004

Document Type

Dissertation

Degree Type

PhD

College

Statler College of Engineering and Mineral Resources

Department

Lane Department of Computer Science and Electrical Engineering

Committee Chair

Matthew C. Valenti.

Abstract

The dissertation concentrates on synchronization of capacity approaching error-correction codes that are deployed in noisy channels with very low signal-to-noise ratio (SNR). The major topics are symbol timing synchronization and frame synchronization.;Capacity-approaching error-correction codes, like turbo codes and low-density parity-check (LDPC) codes, are capable of reaching very low bit error rates and frame error rates in noisy channels by iterative decoding. To fully achieve the potential decoding capability of Turbo codes and LDPC codes, proper symbol timing synchronization, frame synchronization and channel state estimation are required. The dissertation proposes a joint estimator of symbol time delay and channel SNR for symbol timing recovery, and a maximum a posteriori (MAP) frame synchronizer for frame synchronization.;Symbol timing recovery is implemented by sampling and interpolation. The received signal is sampled multiple times per symbol period with unknown delay and unknown SNR. A joint estimator estimates the time delay and the SNR. The signal is rebuilt by interpolating available samples using estimated time delay. The intermediate decoding results enable decision-feedback estimation. The estimates of time delay and SNR are refined by iterative processing. This refinement improves the system performance significantly.;Usually the sampling rate is assumed to be a strict integer multiple of the symbol rate. However, in a practical system the local oscillators in the transmitter and the receiver may have random drifts. Therefore the sampling rate is no longer an exact multiple of the symbol rate, and the sampling time follows a random walk. This random walk may harm the system performance severely. The dissertation analyzes the effect of random time walks and proposes to mitigate the effect by overlapped sliding windows and iterative processing.;Frame synchronization is required to find the correct boundaries of codewords. MAP frame synchronization in the sense of minimizing the frame sync failure rate is investigated. The MAP frame synchronizer explores low-density parity-check attributes of the capacity-approaching codes. The accuracy of frame synchronization is adequate for considered coded systems to work reliably under very low SNR.

Download

Share

COinS