Performance evaluation of CA-, GO- and SO-CFAR processors in a non-centered L´evy-distributed clutter

Constant false alarm rate (CFAR) processors are critical for radar reliable target detection in radar systems. Traditional CFAR designs often assume Gaussian clutter, which may not reflect real-world conditions. L´evy distributions, with heavy tails and a location parameter (δ), provide a more accurate model for non-Gaussian and non-centered clutter in complex environments. This paper presents a comprehensive performance analysis of three widely used CFAR processors-cell-averaging (CA), greatest-of (GO), and smallest-of (SO) in homogeneous L´evy-distributed clutter with an arbitrary δ. We derive integral-form expressions for the probability of false alarm (PFA) for each processor, explicitly incorporating δ. Furthermore, we provide analytical formulations for the probability density function (PDF) of key statistics involving L´evy random variables, such as sums, minima, and maxima. Monte Carlo simulations validate the theoretical results,showing that the PFA performance improves with increasing δ, highlighting the critical impact of clutter location on CFAR detector performance. These findings offer valuable insights for designing robust CFAR detectors in non-Gaussian, non-centered clutter environments.

کلیدواژه ها:

Cell-averaging constant false alarm rate ، Constant false alarm rate processors ، Greatest-of constant false alarm rate ، L´evy-distributed clutter ، Probability of false alarm ، Smallest-of constant false alarm rate