Adaptive detection of targets embedded in Gaussian disturbance is a classic task in many applications, such as radar, sonar and communications, and has received an increasing attention in recent years in radar community. The assumption of a homogeneous environment is somewhat idealistic. The most frequently used assumption to depart from a homogeneous environment is to assume that the primary and secondary data share the same structure but different power level; this is often referred to as the partially homogeneous environment. Constant false alarm rate (CFAR) detection of distributed targets is a problem of primary concern among radar engineers. It naturally arises when considering detection with high resolution radars (HRRs) capable of resolving a target into a number of scattering centers appearing into different range cells. Nevertheless, the problem jointly attacks the design of adaptive detection schemes for distributed targets and in the presence of partially homogeneous environment has not received much consideration until now.

HAO Chengpeng, MA Xiaochuan and CAI Long from the State Key Laboratory of Acoustics, Institute of Acoustics, Chinese Academy of Sciences and SHANG Xiuqin from Institute of Electronics, Chinese Academy of Sciences deal with the problem of detecting distributed targets in the presence of partially homogeneous Gaussian disturbance with unknown covariance matrix.

Since no uniformly most powerful test exists for the problem at hand, it is thus reasonable to design a different detector with better performance or less complexity than those based on GLRT. To this end, the authors devise two new detectors, resorting to the Rao test and the Wald test. Interestingly, both receivers ensure the CFAR property with respect to unknown quantities. The performance assessment, conducted via Monte Carlo simula- tions, shows that the Rao test and the Wald test have opposite behaviors. More precisely, the Rao test is most selective and, therefore, least tolerant of mismatch, whereas the Wald test is most tolerant of mismatch and, therefore, least selective. The analysis of the pro- posed detectors in a heterogeneous environment is worthwhile topics for future research. Moreover, it would be of interest to investigate under which conditions these design criteria are invariant with respect to different classes of detection problems.

This research result was published on the recently issued Signal Processing (Vol. 92, No. 4, Pages 926-930, 2012).