Performance analysis of the matched subspace detector in the presence of signal-dependent interference for MIMO radar

• Published in: Signal processing Vol. 176; p. 107709
• Main Authors: Ghasemi, Zahra, Derakhtian, Mostafa
• Format: Journal Article
• Language: English
• Published: Elsevier B.V 01.11.2020
• Subjects: Diversity gain; Interference region; MIMO radar; Nulling approach; Probability of miss; Signaling loss; Signaling loss; Probability of miss; Interference region; MIMO radar; Diversity gain; Nulling approach
• ISSN: 0165-1684
• DOI: 10.1016/j.sigpro.2020.107709

•We consider the performance analysis of the subspace detector for MIMO radar.•We derive the conditions for maximum diversity gain of the detector.•We derive the approximated form of the probability of miss at high SNR.•We introduce the signaling loss concept.•We derive analytically the detection performance loss of quasi-optimal signals. For a statistical multiple-input multiple-output (MIMO) radar, we investigate the maximum achievable diversity gain and the minimum possible probability of miss detection and the conditions of achieving those at sufficiently high signal-to-noise ratio (SNR) when the matched subspace detector is employed in the presence of signal-dependent interference. In addition to linear dependence between target returns at each receiver, we show that linear dependence between target and interference subspaces reduces the diversity gain when the matched subspace detector is employed. We derive the approximated form of the probability of miss and introduce the signaling loss concept based on it. We show that the probability of miss is minimized at high SNR when not only is the diversity gain maximized, but the signaling loss is also minimized. Then, we use the signaling loss corresponding to the maximum diversity gain as a novel design metric for designing signals and minimizing the probability of miss at sufficiently high SNR in statistical MIMO radar. Due to the fact that designing the optimal signals is challenging, we derive analytically the detection performance loss of quasi-optimal signals by employing the signaling loss concept.