MODEL-BASED RAO AND WALD DETECTORS FOR RANGE-SPREAD TARGETS IN HETEROGENEOUS CLUTTER ENVIRONMENT

A significant advantage of the Random Frequency and Pulse Repetition Interval Agile (RFPA) radar system lies in its ability to reduce the correlation of sea clutter, subsequently enhancing the signal-to-clutter ratio (SCR). However, when the high-resolution RFPA radar operates in sea clutter environments, the backscattered energy from the target spreads across multiple range cells within a single pulse, leading to varying power levels among these cells. To address the challenge of detecting range-spread targets in such heterogeneous clutter environments, this paper introduces the Heterogeneous Autoregressive Rao (HTG-AR-Rao) detector and Heterogeneous Autoregressive Wald (HTG-AR-Wald) detector. These detectors are based on the Rao and Wald test, modeling the clutter as a 1st to 4th order AR process, so that no training data are required. The AR coefficients and the variances of the clutter for individual range cells are determined using Maximum Likelihood Estimation (MLE). Importantly, the asymptotic expressions for both the probability of false alarm and detection verify that the proposed detectors achieve an asymptotically constant false alarm rate (CFAR). Simulation results demonstrate that these detectors adeptly adapt to the heterogeneous clutter environment, delivering satisfactory detection performance and reduce computational cost compared to detectors based on the Generalized Likelihood Ratio Test (GLRT).