Comparative Study of Various CFAR Algorithms for Non-Homogenous Environments

Based on the radar field environment, the detection process is generally based on the adaptive threshold required to detect the received radar cell signal. Many algorithms are used to design this adaptive threshold to satisfy a Constant False Alarm Rate (CFAR) in line with detection criteria in non-homogenous environments. Although CFAR algorithms have increasingly become a vital factor in the detection process, the performance of these algorithms differs according to their treatment of the received radar bins. In cases of targets in a clutter edge and multi-targets, the performance of the GO-, VI-, OSSO- and OSGO-CFAR algorithms are better than the performance of the CA-CFAR algorithm, while the CA-CFAR and OS-CFAR algorithms lose such targets. In this context, the SO-CFAR has a higher P_fa than the above algorithms for small targets in a clutter edge. OSSO-, OSGO, VI, and CMLD also have higher P_D S based on their outcomes than the other algorithms in the same P_fa in non-homogenous environments. However, CA-CFAR is simpler and has less complexity than the other algorithms. MATLAB 2015b was used to evaluate the performance of these different types of CFAR algorithm. (paper)