Linear Subspace Approach to Burst Communication Signal Processing

摘要

This dissertation focuses on the topic of burst signal communications in a high interference environment. It derives new signal processing algorithms from a mathematical linear subspace approach instead of the common stationary or cyclostationary approach. The research developed new algorithms that have well- known optimality criteria associated with them. The investigation demonstrated a unique class of multisensor filters having a lower mean square error than all other known filters, a maximum likelihood time difference of arrival estimator that outperformed previously optimal estimators, and a signal presence detector having a selectivity unparalleled in burst interference environments. It was further shown that these improvements resulted in a greater ability to communicate, to locate electronic transmitters, and to mitigate the effects of a growing interference environment.