<title>Extension of the MUSIC algorithm to millimeter-wave (MMW) real-beam radar scanning antennas</title>

Abstract

Step-scanned radar antennas represent a new application of radar technology for detection of targets and estimation of their locations. In this paper we describe a new development called Scan-MUSIC (SMUSIC), which extends the application of the MUSIC algorithm to improve the cross-range resolution of closely spaced point targets with a step-scanned radar. This paper also demonstrates that SMUSIC can be used with radar data obtained with an experimental Millimeter Wave (MMW) coherent scanning radar. While a mathematical proof of resolvability has not yet been established for the scanning antenna, we have shown that we can apply the spatial smoothing method to the SMUSIC algorithm to estimate the closely spaced point targets that are within the beamwidth of the radar antenna. The results show that the targets that are spaced less than 1/4 of the antenna beamwidth and are interfering can be resolved with SMUSIC in constructive interference case. This paper also presents the performance of the SMUSIC superresolution algorithm for the scanning antenna in terms of probability of successful resolution and the total average mean-squared error of target locations, based on the simulated data generated by using an experimental antenna pattern.

A Bayesian Angular Superresolution Method With Lognormal Constraint for Sea-Surface Target

“…Real-aperture scanning radar imaging is a significant technology to achieve the reconnaissance and detection of sea-surface target because it is suitable for most geometry situation [1]. However, the angular resolution of this radar system is greatly limited by the antenna aperture size, which seriously influences the searching ability and location accuracy [2], [3]. Over the years, the proposed angular superresolution approaches are applied to the ground scenarios mostly.…”

Beam space estimation of the direction‐of‐arrivals of coherent signals

“…Whilst this is true for some radar systems, other systems use the same hardware to form both the transmit and receive beam and then this single transmit (Tx)/receive (Rx) beam is sequentially scanned over the sector of interest. In this case, the standard BS algorithm [6] is not applicable and several approaches [21–24] have been developed, which explicitly take into account the transmit beam pattern. In [21], the various beams (formed by mechanically scanning an antenna over a sector of interest) are split into subsets and the covariance matrices for each subset are then estimated and averaged to form a reduced dimension spatially smoothed BS covariance matrix.…”

“…In this case, the standard BS algorithm [6] is not applicable and several approaches [21–24] have been developed, which explicitly take into account the transmit beam pattern. In [21], the various beams (formed by mechanically scanning an antenna over a sector of interest) are split into subsets and the covariance matrices for each subset are then estimated and averaged to form a reduced dimension spatially smoothed BS covariance matrix. After eigen‐decomposition of this matrix, the noise subspace is used to form a pseudo‐spectrum and the approach is termed SMUSIC.…”

Bayesian Angular Superresolution Algorithm for Real-Aperture Imaging in Forward-Looking Radar

“…Recently, some approaches have been proposed to enhance the spatial resolution [3][4][5][6][7][8][9][10]. In [9,10], the researchers introduced the spectrum estimation methods to enhance the azimuth angular resolution of a real aperture image. However, how to collect enough snapshots to realize the high estimation precision of the covariance matrix is difficult for the mechanical scanning radar system.…”