It is known that there exist two kinds of methods for direction-of-arrival (DOA) estimation in the literature: subspace-based method and sparsity-based method. However, pervious works reveal former cannot address case which number signals larger than sensors, whereas latter one always suffers from influence basis mismatch. In this paper, to overcome these shortcomings, we propose a new called covariance matrix reconstruction approach (CMRA) both uniform linear array sparse array. particular,...

The inherent limitation of the predefined spatial discrete grids greatly restricts precision and feasibility many sparse signal representation (SSR)-based direction-of-arrival (DOA) estimators. In this paper, we first propose a perturbed SSR-based model to alleviate by incorporating bias parameter into DOA estimation framework. Using model, Bayesian learning-based algorithm, named PSBL, is developed solve problem, followed theoretical analysis PSBL. We then present two algorithms based on...

Most existing deep learning (DL) based direction-of-arrival (DOA) estimation methods treat direction finding problem as a multi-label classification task and the output of neural network is probability spectrum where peaks indicate true DOAs. These essentially belong to grid-based may encounter grid mismatch effect. In this paper, we focus on gridless DL DOA under generalized linear array which can be regarded uniform (ULA) with/without "holes". By using Toeplitz structure, convolutional...

The low-rank matrix reconstruction (LRMR) approach is widely used in direction-of-arrival (DOA) estimation. As the rank norm penalty an LRMR NP-hard to compute, nuclear (or trace for a positive semidefinite matrix) has been often employed as convex relaxation of norm. However, solving problem may lead suboptimal solution original problem. In this paper, we propose apply family nonconvex penalties on singular values covariance sparsity metrics approximate particular, formulate minimization...

In this paper, we propose a novel deep learning (DL)-based gridless direction-of-arrival (DOA) estimation method for generalized linear arrays using residual attention network (RAN) and transfer (TL). The proposed can improve the DOA performance in both low high signal-to-noise ratio (SNR) regions by focusing on important features input avoiding problems of gradient vanishing degradation. Moreover, introduce idea TL to reduce complexity costs training. experimental results demonstrate...

The covariance matrix of the array output is widely used for direction-of-arrival (DOA) estimation in signal processing. However, existing methods are only suitable either 1-D DOA or 2-D estimation. In this paper, we present a computationally efficient method both and referred to as fast gridless maximum likelihood method. particular, by exploiting Hermitian-Toeplitz structure matrix, convex optimization problem reconstruction further derive closed-form solution problem. DOAs can then be...

This correspondence proposes a localization method for mixed far-field (FF) and near-field (NF) sources based on cross arrays where two-dimensional (2-D) direction-of-arrival (DOA) range estimation are considered. For 2-D DOA estimation, we first construct range-free cumulant-based vector employ the covariance matching criterion Vandermonde decomposition theorem to retrieve information. We then build sparse model with respect parameter solve an ℓ <sub...

A new extreme learning machine (ELM) localization technique that uses received signal strength indicator fingerprints only is proposed for multifloor environments. This structured scheme forms multiple individual ELMs the floors as well geographically formed data clusters of each floor. Multifloor environments often have huge amount training and online measurement data. To maximize efficiency, we develop a preprocessing algorithm, aiming to: 1) efficiently extract out essential information...

This paper addresses the issue of direction-of-arrival (DOA) estimation with an objective to eliminate off-grid effect sparsity-based methods and enlarge maximum number distinguishable signals in subspace-based methods. We first reconstruct covariance matrix array output Toeplitz structure then employ reconstructed together root-MUSIC estimate DOAs. The proposed reconstruction approach (CMRA) can be used for uniform sparse linear arrays. It also DOAs multiple that are larger than sensors by...

Existing works for 3-D source localization are based on uniform cross arrays. In this correspondence, we propose a single far-field (FF) or near-field (NF) method arrays and then generalize it to the symmetric sparse where each linear array is one. We utilize cross-covariance matrix exploit full aperture employ atomic norm minimization technique retrieve angle information without discretization. The range estimated by MUSIC-like method. Our superior traditional methods exhibits enormous...

This letter develops an extreme learning machine (ELM) and AdaBoost technique for indoor localization using channel state information (CSI) received signal strength indicators (RSSI), aiming to (a) resolve issue with ELM in which weight parameters are generated randomly that leads significant performance variations, (b) maximize given the same set of measurement data, (c) reduce data storage computational needs existing schemes. To this end, we first form fingerprint training dataset...

Modal frequency is an important indicator reflecting the health status of a structure. Numerous investigations have shown that its fluctuations are related to changing environmental factors. Thus, modelling modal frequency–multiple factors relation essential for making reliable inference in structural monitoring. In this study, Bayesian network (BN)-based algorithm developed recognizing pattern between and multiple Different candidates structure BN proposed describe possible statistical...

Direction finding in bistatic multiple-input multiple-output (MIMO) radar system is an important issue target localization. Existing works focus on linear array which can only find 1-D angle information. In this paper, we MIMO where both the transmitter and receiver are equipped with uniform planar arrays (UPAs) full information of targets. To overcome gridding effect, use atomic norm minimization (ANM) to formulate a gridless model equivalent semidefinite programming (SDP). By solving SDP...

A two-phase smartphone localization technique that uses received signal strength indicator (RSSI) fingerprints of long term evolution (LTE) signal, Bluetooth Wi-Fi and the internal camera sensor is proposed for indoor environments. It contains: (1) coarse localization: region determination by LTE (2) refined position estimation image signal. To maximize efficiency we develop data fusion algorithm, aiming to (a) combine RSSI measurement form fingerprint. (b) transform measurements into...

This letter proposes a localization method for mixed far-field (FF) and near-field (NF) sources based on second-order statistics (SOS) with generalized symmetric sparse linear arrays. First, the DOAs of FF are estimated by using MUSIC method. Then, we use oblique projection technique to isolate NF from ones atomic norm minimization is employed retrieve sources. Finally, range information determined one-dimensional searching. To against effect finite measurements, an iterative procedure...

In this article, a new indoor localization technique with extreme learning machine (ELM) is proposed where only small number of received signal strength indicator (RSSI) measurements are labeled. the off-line phase, iterative self organizing data analysis techniques algorithm (ISODATA) used to divide RSSI into some measurement subsets. Then multi-kernel ELM (MK-ELM) method utilized perform classification and obtain function. For each subset, two-stage feature extraction using kernel...

The atomic norm minimization (ANM) has been successfully incorporated into the two-dimensional (2-D) direction-of-arrival (DOA) estimation problem for super-resolution. However, its computational workload might be unaffordable when number of snapshots is large. In this paper, we propose two gridless methods 2-D DOA with L-shaped array based on to improve efficiency. Firstly, by exploiting cross-covariance matrix an ANM-based model proposed. We then prove that can efficiently solved as a...

This paper investigates the secure transmission in satellite communication systems, where a geostationary orbit (GEO) serves an earth station while multiple eavesdroppers attempt to intercept confidential message. Assuming that only imperfect channel state information (CSI) of wiretap channels are available, we propose beamforming scheme maximize secrecy energy efficiency (SEE) satisfying signal-noise-ratio (SNR) requirement at station, constraints eavesdroppers, and per-antenna power...

In this letter, we propose a gridless method for mixed FF and NF sources localization. To establish the Vandermonde structure DOA range estimation, employ fourth-order cumulants to construct two special matrices which are only related or parameters, respectively. Then, build low-rank matrix reconstruction models retrieve DOAs ranges by exhibiting Toeplitz structures of cumulant matrices. We also present an easy-to-implement post-processing pair estimates. Our requires no discretization angle...