Statistical analysis of the non-homogeneity detector for STAP applications

Performance Analysis of Projection Statistics through Method of Clutter Covariance Matrix Estimation for STAP

“…따라서 불균일한 클러터 환경에서 STAP의 성능이 크게 감소하는 것 을 극복하기 위하여 불균일한 클러터 신호가 포함 된 range bin(셀)을 제외하는 nonhomogeneity detector (NHD)가 활발히 연구되었다 [5]～ [7] . 그러나 불균일한 클러터 신호의 세기가 강할수록, NHD를 [8] .…”

Statistical analysis of the nonhomogeneity detector for non-Gaussian interference backgrounds

“…Fig. 8 shows the performance of a second goodness-of-fit test We then generate 1024 realizations of a 64 tuple vector from proposed in [10], [11], and [24], which compares the normalthe K-distributed SIRP with as = 0.5 having a prescribed co-ized GIP xHS-lx/K to a theoretically calculated mean value variance matrix according to the physical model described in obtained from [24, eq. (3.6)].…”

“…Fig. 14 shows the results of the NHD test proposed in [10], [11], and Direct maximization of the likelihood function of (19) over R [24] based on comparing the normalized GIP xHS-lx/K with is rendered difficult due to the fact that there is missing inthe threshold setting determined according to [24, eq. (4.2)].…”

“…., K, It is seen from Figs. 14 and 15 that a lot more declarations or xi, Vi, i = 1,2,... , K. However, the observed data xi, of nonhomogeneity result from the NHD methods of [10], i = 1, 2,... , K contains no explicit information about Vi, i = [11], and [24] and [3], [4], and [9], when in fact the MCARM 1,2,.. . , K and, thus, constitutes the incomplete data.…”

“…For the purpose of this SIRVs can be treated in the framework of a complete-incompaper, we assume knowledge of fv (v) and treat the problem of plete data problem and pointed out that the maximum likelinonhomogeneity detection with respect to unknown R. [24] employed the expectation-maximization (EM) algorithm. More precisely, let GIP-based NHD for Gaussian interference scenarios.…”