Title:  Detection and Synchronization of Direct-Sequence Spread-Spectrum Signals

Committee:

Dr. John Barry, ECE, Chair, Advisor

Dr. Xiaoli Ma, ECE

Dr. Aaron Lanterman, ECE

Dr. Gordon Stuber, ECE

Dr. Wenjing Liao, Math

Abstract: This thesis proposes strategies for both acquisition of direct-sequence spread-spectrum (DSSS) signals in high-dynamic environments, and noncooperative detection of DSSS signals. In particular, we propose delay-Doppler efficient exhaustive search (DEES), a computationally efficient algorithm that can acquire DSSS signals in the presence of both Doppler rate and large Doppler frequency shifts. DEES combines the second-order keystone transform and the fractional Fourier transform to mitigate the time-varying effects of the channel, before jointly estimating both the code phase offset and the Doppler frequency. Based on numerical simulation results, DEES provides improved acquisition performance over existing FFT-based acquisition algorithms without the computational complexity of a three-dimensional maximum-likelihood exhaustive search. For the noncooperative DSSS detection problem, we explore the advantage of multi-antenna detector that knows the signal has QPSK coded chips, when the spreading sequence is unknown to the detector, and when the spreading code period is longer than the detector observation window. We propose a likelihood-ratio-test detector that takes advantage of knowledge of the signal alphabet, whose complexity grows linearly with the observation window length, and that makes use of multiple antennas. The proposed alphabet-aware detector outperforms multi-antenna alphabet-unaware detectors, especially when the SNR is high, and the observation window is small. However, the performance advantage is not large, it diminishes further when the SNR is low, and it comes at the cost of higher computational complexity. In most cases, simpler alphabet-unaware detectors such as the energy detector provide comparable detection performance with less complexity.