Focus-before-detection radar signal processing: part i—challenges and methods

During the World War II (WWII), radar was invented as an all-day, all-weather, long-range sensor. Over the past 75 years or so, radar has acted as the clairvoyance and clairaudience of humans and has had wide applications in both defense and civilian fields, e.g., surveillance, reconnaissance, fire control, border monitoring, collision avoidance, and traffic control. Along with radar development, radar signal processing (RSP) always plays an important role. In 1943, North pointed out that the matched filter should be the foundation of radar to detect targets from echoes generated by the known transmitting waveform. In 1950, Woodward and Davies introduced the Bayesian statistics theory into the RSP field and proposed the constant false alarm ratio (CFAR) rule for target detection. Based on matched filtering and the CFAR rule, the embryonic framework of classic RSP has been formed and used thus far. In 1953, Woodward further proposed an ambiguity function as a useful tool to evaluate the range and Doppler discrimination abilities of a transmitting waveform, which paves the foundation for radar waveform design. Furthermore, to deal with the tradeoff between large transmitting energy and high range resolution, the early unmodulated rectangular waveform has been replaced by more waveforms with a large time-bandwidth product (TBP), such as linear frequency modulation (LFM) and phase-coded waveforms. These waveforms with large TBP have been widely used to realize the range compression via matched filter.