MIMO Radar Detection in Non-Gaussian and Heterogeneous Clutter

Abstract

Abstract-In this paper, the generalized likelihood ratio test-linear quadratic (GLRT-LQ) has been extended to the multiple-input multiple-output (MIMO) case where all transmit-receive subarrays are considered jointly as a system such that only one detection threshold is used. The GLRT-LQ detector has been derived based on the Spherically Invariant Random Vector (SIRV) model and is constant false alarm rate (CFAR) with respect to the clutter power fluctuations (also known as the texture). The new MIMO detector is then shown to be texture-CFAR as well. The theoretical performance of this new detector is first analytically derived and then validated using Monte Carlo simulations. Its detection performance is then compared to that of the well-known Optimum Gaussian Detector (OGD) under Gaussian and non-Gaussian clutter. Next, the adaptive version of the detector is investigated. The covariance matrix is estimated using the Fixed Point (FP) algorithm which enables the detector to remain texture-and matrix-CFAR. The effects of the estimation of the covariance matrix on the detection performance are also investigated.Index Terms-Detection performance, generalized likelihood ratio test-linear quadratic (GLRT-LQ), multiple-input multiple-output (MIMO) radar, non-Gaussian clutter, Spherically Invariant Random Vector (SIRV).

MIMO Signal Detection Using Neyman Pearson Signal Detection

“…Let T denote a testing procedure based on an observation of X, and let RT denote the subset of the range of X where the test chooses H1. The probability of a false-positive is denoted by (2) The probability of a false-negative is 1 − P1(RT ), where (3) is the probability of correctly deciding H1, often called the probability of detection.Consider likelihood ratio tests of the form (4) The subset of the range of X where this test decides H1 is denoted (5) and therefore the probability of a false-positive decision is (6) This probability is a function of the threshold λ the set RLR (λ) shrinks/grows as λ increases/decreases. We can select λ to achieve a desired probability of error.…”

“…In this system several transmitting antennas and one receiving antenna is present [4]- [6]. This is also known as transmitting diversity.…”

Multistatic adaptive CFAR detection in non-Gaussian clutter

“…This detector has been originally derived in [23] in a different manner. In [23], the authors show that in the case of known covariance matrices, this detector is texture-CFAR and P FA is a function only of the detection threshold λ, the number of pulses L and the number of transmitter-receiver pairs (MN = P),…”

“…In [23], it is also shown that detector (28) outperforms the MIMO-Optimum Gaussian Detector (MIMO-OGD) in non-Gaussian clutter.…”

“…In this work, we consider a multichannel extension of the NMF, introduced with the name of MIMOGeneralized Likelihood Ratio Test-Linear Quadratic (MIMO-GLRT-LQ) in [23]. This detector is the MIMO version of the aforementioned NMF; thus, it can also be termed MIMO-NMF.…”