Coherent detection for distributed target in compound-Gaussian clutter with inverse gamma texture had long been researched, since it gained importance in radar community. And there were two fundamental aspects, the target and the clutter, to be regarded. On one hand, whether a target can be seen as a point depended on the range resolution of radar and the range extent of target. On the other hand, choosing a suitable clutter model was principal in designing radar detector and different clutter model lead to different performances. Detection problem under Gaussian assumptions was considered extensively and some classic detectors were given in literature. However, the related detection problem based on compound-Gaussian model has been studied with little considering the texture model.

HAO Chengpeng from the State Key Laboratory of Acoustics, Institute of Acoustics, Chinese Academy of Sciences, SHANG Xiuqin, SONG Hongjun and WANG Yu from the Institute of Electronics, Chinese Academy of Sciences and LEI Chuan from the School of Electronics and Information Engineering, Beihang University studied the problem of coherent detection for distributed target in compound-Gaussian clutter with inverse gamma texture.

In this work, the one-step generalized likelihood ratio test (GLRT), maximum a-posteriori GLRT and two-step GLRT detectors, three related detectors had been derived under Bayesian and non-Bayesian framework. These detectors had similar detection structures with their test statistics modulated by the shape and scale parameters of the texture. The detection structure of the proposed detectors could be written as two forms: one had their test statistics related to the shape parameter and the scale parameter; the other could be reformulated into another form with their test statistics associated with the scale parameter and detection thresholds related with the shape parameter. In addition, the latter detection structure could be reformulated as the matched filter form like the detectors for point target. Subsequently, their detection performances were analyzed via Monte Carlo simulations and compared with those of 2S-GLRT which was the detector for distributed target based on compound-Gaussian model without taking texture model into account. The researchers analyzed the robustness of those Bayesian detectors and find that the detectors behaved well when the shape parameter and the scale parameter suffered from errors. And the scale parameter had less influence on these Bayesian detectors than the shape parameter.

The research results enlightened us that the three Bayesian detectors bore pretty much the same detection performances and the detection performances fluctuated more intensely when the shape parameter or the scale parameter was smaller. Beiseds, the shape parameter had more influences on the detection performances than the scale parameter, as it was an indication of the clutter impulsiveness.

This research results have been published online on Digital Signal Processing (Volume 22, Issue 6, December 2012, Pages 1024–1030).