Radar performance analysis in the presence of sea clutter

Abstract

Radar has been used for decades for surveillance purposes, originally meant for target detection and early warning. During the early days, radar detector has been developed by assuming the radar clutter is Gaussian distributed. However, as modern technology emerges, the radar distribution is seen to deviates from the Gaussian assumption. Thus, detectors designed based on Gaussian assumption are no longer optimum for detection in non-Gaussian nature. Lots of researches have been carried out for optimum target detection in non-Gaussian clutter distributions. Neyman- Pearson detector is proven to be the best detector for radar detection due to the unknown cost and prior probabilities. The theory of target detection in Gaussian distributed clutter has been well established and the closed form of the detection performances can be easily obtained. However, that is not the case in non-Gaussian clutter distributions. Thus, this thesis aims to serve as a basis in understanding performance analysis of target detection in the presence of sea clutter. In the thesis, the performance model in terms of ROC plots of probability of detection against signal to noise ratio for different sea clutter distributions are obtained and analyzed.