We gratefully acknowledge support from
the Simons Foundation and member institutions.
Full-text links:

Download:

Current browse context:

eess

Change to browse by:

References & Citations

Bookmark

(what is this?)
CiteULike logo BibSonomy logo Mendeley logo del.icio.us logo Digg logo Reddit logo ScienceWISE logo

Electrical Engineering and Systems Science > Signal Processing

Title: New Findings on GLRT Radar Detection of Nonfluctuating Targets via Phased Arrays

Abstract: This paper addresses the standard generalized likelihood ratio test (GLRT) detection problem of weak signals in background noise. In so doing, we consider a nonfluctuating target embedded in complex white Gaussian noise (CWGN), in which the amplitude of the target echo and the noise power are assumed to be unknown. Important works have analyzed the performance for the referred scenario and proposed GLRT-based detectors. Such detectors are projected at an early stage (i.e., prior to the formation of a post-beamforming scalar waveform), thereby imposing high demands on hardware, processing, and data storage. From a hardware perspective, most radar systems fail to meet these strong requirements. In fact, due to hardware and computational constraints, most radars use a combination of analog and digital beamformers (sums) before any estimation or further pre-processing. The rationale behind this study is to derive a GLRT detector that meets the hardware and system requirements. In this work, we design and analyze a more practical and easy-to-implement GLRT detector, which is projected after the analog beamforming. The performance of the proposed detector is analyzed and the probabilities of detection (PD) and false alarm (PFA) are derived in closed form. Moreover, we show that in the low signal-to-noise ratio (SNR) regime, the post-beamforming GLRT detector performs better than both the classic pre-beamforming GLRT detector and the square-law detector. This finding suggests that if the signals are weak, instead of processing the signals separately, we first must to reinforce the overall signal and then assembling the system's detection statistic. At last, the SNR losses are quantified, in which the superiority of the post-beamforming GLRT detector was evidenced as the number of antennas and samples increase.
Comments: 13 pages, 12 figures, 1 table
Subjects: Signal Processing (eess.SP)
Journal reference: PP. 1-14, 2021
DOI: 10.1109/ACCESS.2021.3095407
Cite as: arXiv:2101.11077 [eess.SP]
  (or arXiv:2101.11077v2 [eess.SP] for this version)

Submission history

From: Fernando Almeida [view email]
[v1] Tue, 26 Jan 2021 20:42:16 GMT (5176kb,D)
[v2] Thu, 8 Jul 2021 23:32:35 GMT (5176kb,D)

Link back to: arXiv, form interface, contact.