An analysis of the information content of radar detection

Abstract

The availability of the electromagnetic spectrum (EMS) was an unseen issue in the past, as there was sufficient spectrum access to suit the needs of its’ consumers. Today, the use of the EMS has been become integrated within our daily lives. Applications varying from civil infrastructure to automotive radar has readily consumed the spectrum to communicate, sense, and interpret information. Given the inflation of spectrum use, it is important that we investigate the amount of information and bandwidth that different spectrum-based applications are using and how different parameters can impact spectrum use. Previous work has identified the fundamental decision bound of pulse-Doppler radar, defined by the Rayleigh range-Doppler resolution, of 1 decision per second per Hz of transmitted bandwidth and has identified a Bayesian detection capacity expression. In this thesis, a closed form expression is derived for detection capacity of radar, which does not require a priori probabilities. Furthermore, detection capacity expressions for multiple receivers and the use of M of N integration are also derived. These expressions are used to analyze the information content of different commonly used radar detectors for different assumptions. Another primary focus of this thesis was the analysis of the information content of both fluctuating targets and cell averaging constant false alarm rate (CA-CFAR). The novel analysis of these different radar/target assumptions show the effect that radar detection has on the spectrum. Finally, detection bandwidth and transmit range are connected to determine how signal-to-noise ratio can impact transmit range and spectral usage.