Search

In this article, we propose a new compressive-sampling-based approach to detect a target and obtain its direction-of-arrival (DOA) by using samples collected by a set of sensor nodes. Since these sensors sample at a rate below Nyquist, instead of the Nyquist rate, the complexity, power consumption, memory requirement, and the volume of data that needs to be exchanged between sensors, are much lower than those of the existing approaches. The proposed scheme jointly estimates the DOA, the variance of the noise, and the covariance matrix of the interference. As a result, simulations have shown that this scheme significantly outperforms techniques that simply combine the noise and the interference.gg [ABSTRACT FROM AUTHOR]

In this paper, we address the problem of radar range-Doppler imaging in the presence of clutter. Specifically, we formulate the range-Doppler imaging problem as that of recovery of a sparse vector contaminated by clutter in addition to noise. We propose a sparse Bayesian learning (SBL)-based algorithm to jointly obtain the range-Doppler image, variance of the noise, and covariance matrix of the clutter. Furthermore, we adapt a simple pruning mechanism that reduces the computational cost and improves the convergence speed. [ABSTRACT FROM AUTHOR]

We apply the technique of compressive sensing (CS) to multiple-input multiple-output (MIMO) radars to estimate the direction of arrival (DOA) of potential targets embedded in cluttered environments using far fewer samples than the Nyquist rate. Specifically, incorporating clutter into the sparse Bayesian learning (SBL) methodology, we devise a novel algorithm for joint estimation of targets’ DOAs, clutter covariance matrix, and noise variance. Furthermore, we develop a low-complexity and fast version of the proposed algorithm, which can be efficiently employed in practice. [ABSTRACT FROM AUTHOR]

This paper employs a special property of the Hadamard matrix to develop a novel compressive sensing (CS) based framework that eliminates the complex reconstruction step required by standard CS methods. In particular, time-difference-of-arrivals (TDOAs) of a signal at different receivers involved in positioning of unknown sources are directly estimated from CS measurements, without any attempt to reconstruct the full signals. Although this approach is designed for TDOA estimation, it is also capable of accommodating other CS applications. [ABSTRACT FROM AUTHOR]

A new compressive-sensing (CS)-based electronic warfare (EW) receiver is designed to estimate the angle-Doppler of adversary targets whose waveforms are unknown. The proposed EW receiver uses a sparse Bayesian learning (SBL) framework, which is blind in the sense that the knowledge of the sparsity basis is not available. Furthermore, a pruning mechanism is proposed to reduce the computational cost and improve convergence speed of the blind-SBL. The convergence of the proposed method is analytically proved. [ABSTRACT FROM AUTHOR]