Yazar "Elbir, Ahmet Musab" seçeneğine göre listele
Listeleniyor 1 - 17 / 17
Sayfa Başına Sonuç
Sıralama seçenekleri
Öğe Calibration of directional mutual coupling for antenna arrays(Academic Press Inc Elsevier Science, 2017) Elbir, Ahmet MusabMutual coupling (MC) is one of the major error sources in array signal processing. The previous methods mostly assume that the MC is direction-independent and it is modeled by a single MC matrix. However, this is not valid in a practical scenario where the effect of MC differs for the source signals incoming from different directions. In this paper, calibration of directional MC is considered for direction-of-arrival (DOA) estimation problem. An alternating and sectorized parameter estimation (ASPE) algorithm is proposed where the estimates of the source DOA angles and the MC coefficients corresponding to each source direction are found iteratively. A unified approach is introduced so that the proposed algorithm can effectively work for different array geometries regardless of the array geometry and the corresponding MC matrix model. The performance of the proposed method is evaluated by several experiments and it is compared with the conventional calibration techniques as well as the Cramer-Rao lower Bound which is derived for the considered problem. It is shown that the proposed method effectively finds the unknown source and coupling parameters and it has superior performance as compared to the conventional calibration techniques. (C) 2017 Elsevier Inc. All rights reserved.Öğe CNN-based cognitive radar array selection(Institute of Electrical and Electronics Engineers Inc., 2019) Elbir, Ahmet Musab; Mishra, Kumar Vijay; Eldar, YoninaIn cognitive radar, it may be desired to select an optimal subarray from a full antenna array in each scan to reduce the cost and computational complexity. Previous works on antenna selection rely on mostly optimization or greedy search methods. In this paper, we introduce a deep learning approach for antenna selection in a cognitive radar scenario. We design a deep convolutional neural network (CNN) to select the best subarray for direction-of-arrival estimation for each scan. The CNN accepts the array covariance matrix as its input and, unlike previous works, does not require prior knowledge about the target location. The performance of the proposed CNN approach is evaluated through numerical simulations. In particular, we show that it provides more accurate results than conventional support vector machines. © 2019 IEEE.Öğe CNN-Based Precoder and Combiner Design in mmWave MIMO Systems(Ieee-Inst Electrical Electronics Engineers Inc, 2019) Elbir, Ahmet MusabHybrid beamformer design is a crucial stage in millimeter-wave (mmWave) MIMO systems. In this letter, we propose a convolutional neural network (CNN) framework for the joint design of precoder and combiners. The proposed network accepts the input of channel matrix and gives the output of analog and baseband beamformers. Previous works are usually based on the knowledge of steering vectors of array responses which is not always accurately available in practice. The proposed CNN framework does not require such a knowledge, and it provides higher performance in capacity compared with the conventional greedy-and optimization-based algorithms.Öğe Cognitive radar antenna selection via deep learning(Inst Engineering Technology-Iet, 2019) Elbir, Ahmet Musab; Mishra, Kumar Vijay; Eldar, YoninaDirection-of-arrival (DoA) estimation of targets improves with the number of elements employed by a phased array radar antenna. Since larger arrays have high associated cost, area and computational load, there is a recent interest in thinning the antenna arrays without loss of far-field DoA accuracy. In this context, a cognitive radar may deploy a full array and then select an optimal subarray to transmit and receive the signals in response to changes in the target environment. Prior works have used optimisation and greedy search methods to pick the best subarrays cognitively. In this study, deep learning is leveraged to address the antenna selection problem. Specifically, they construct a convolutional neural network (CNN) as a multi-class classification framework, where each class designates a different subarray. The proposed network determines a new array every time data is received by the radar, thereby making antenna selection a cognitive operation. Their numerical experiments show that the proposed CNN structure provides 22% better classification performance than a support vector machine and the resulting subarrays yield 72% more accurate DoA estimates than random array selections.Öğe Coprıme Arrays Wıth Enhanced Degrees Of Freedom(2018) Elbir, Ahmet MusabCoprime array geometries provide robust performance for direction-of-arrival estimation problem with more sources thansensor elements. In previous works it is shown that K source directions can be resolved using only 2M N -1 sensorelements where K is less than or equal to MN for M and N are integer numbers. In this paper we introduce a new approach toenhance the degrees of freedom (DOF) from MN to 2MN by using the same number of sensor elements. The proposedmethod is based on computing the covariance matrix of the observation data multiple times. Hence more DOF can beobtained. The resulting cross terms corresponding to the coherent sources are modeled as interference in a sparse recoveryalgorithm which is solved effectively by an alternating minimization procedure. The theoretical analysis of the proposedmethod is provided and its superior performance is evaluated through numerical simulations.Öğe Deep learning design for joint antenna selection and hybrid beamforming in massive MIMO(Institute of Electrical and Electronics Engineers Inc., 2019) Elbir, Ahmet Musab; Mishra, Kumar VijayIn this paper, we propose a deep-learning-based for joint antenna selection and hybrid beamformer design problem in mmWave massive MIMO systems. In this respect, we treat both problems as a classification problem. We design two convolutional neural networks (CNNs) which accept the input as the channel matrix and it yields the output as the optimum antenna subarray. The selected part of channel matrix is fed to the second CNN which gives the output as the analog and baseband beamformers. We evaluate the performance of the proposed approach through numerical simulations and show that our CNN framework provides significantly better performance as compared to the conventional techniques such as orthogonal matching pursuit. © 2019 IEEE.Öğe A Deep Learning Framework for Hybrid Beamforming Without Instantaneous CSI Feedback(Ieee-Inst Electrical Electronics Engineers Inc, 2020) Elbir, Ahmet MusabHybrid beamformer design plays very crucial role in the next generation millimeter-wave (mm-Wave) massive MIMO (multiple-input multiple-output) systems. Previous works assume the perfect channel state information (CSI) which results heavy feedback overhead. To lower complexity, channel statistics can be utilized such that only infrequent update of the channel information is needed. To reduce the complexity and provide robustness, in this work, we propose a deep learning (DL) framework to deal with both hybrid beamforming and channel estimation. For this purpose, we introduce three deep convolutional neural network (CNN) architectures. We assume that the base station (BS) has the channel statistics only and feeds the channel covariance matrix into a CNN to obtain the hybrid precoders. At the receiver, two CNNs are employed. The first one is used for channel estimation purposes and the another is employed to design the hybrid combiners. The proposed DL framework does not require the instantaneous feedback of the CSI at the BS. We have also investigated the online deployment of DL for channel estimation. We have shown that the proposed approach has higher spectral efficiency with comparison to the conventional techniques. The trained CNN structures do not need to be re-trained due to the changes in the propagation environment such as the deviations in the number of received paths and the fluctuations in the received path angles up to 4 degrees. Also, the proposed DL framework exhibits at least 10 times lower computational complexity as compared to the conventional optimization-based approaches.Öğe Deep-Sparse Array Cognitive Radar(Institute of Electrical and Electronics Engineers Inc., 2019) Elbir, Ahmet Musab; Mulleti, Satish; Cohen, Regev; Fu, Rong; Eldar, Yonina C.In antenna array based radar applications, it is often desirable to choose an optimum subarray from a full array to achieve a balance between hardware cost and resolution. Moreover, in a cognitive radar system, the sparse subarrays are chosen based on the target scenario at that instant. Recently, a deep-learning based antenna selection technique was proposed for a single target scenario. In this paper, we extend this approach to multiple targets and assess the performance of state-of-the-art direction of arrival estimation techniques in conjunction with the proposed antenna selection method. To optimally choose the subarrays based on the target DOAs, we design a convolutional neural network which accepts the array covariance matrix as an input and selects the best sparse subarray that minimizes the error. Once the optimum sparse subarray is obtained, the signals from the selected antennas are used to estimate the DOAs. We provide numerical simulations to validate the performance of the proposed cognitive array selection strategy. We show that the proposed approach outperforms random sparse antenna selection and it leads to a higher DOA estimation accuracy by 6 dB. © 2019 IEEE.Öğe DeepMUSIC: Multiple Signal Classification via Deep Learning(Institute of Electrical and Electronics Engineers Inc., 2020) Elbir, Ahmet MusabThis letter introduces a deep learning (DL) framework for the classification of multiple signals in direction finding (DF) scenario via sensor arrays. Previous works in DL context mostly consider a single or two target scenario, which is a strong limitation in practice. Hence, in this letter, we propose a DL framework called DeepMUSIC for multiple signal classification. We design multiple deep convolutional neural networks (CNNs), each of which is dedicated to a subregion of the angular spectrum. Each CNN learns the MUltiple SIgnal Classification (MUSIC) spectra of the corresponding angular subregion. Hence, it constructs a nonlinear relationship between the received sensor data and the angular spectrum. We have shown, through simulations, that the proposed DeepMUSIC framework has superior estimation accuracy and exhibits less computational complexity in comparison with both DL- and non-DL-based techniques. © 2017 IEEE.Öğe Direction Finding in the Presence of Direction-Dependent Mutual Coupling(Ieee-Inst Electrical Electronics Engineers Inc, 2017) Elbir, Ahmet MusabDirection-of-arrival (DOA) estimation in the presence of mutual coupling (MC) is an important problem in direction-finding applications. Previous methods in the literature assume that the MC in the array is modeled with a single direction-independent MC matrix. However, this assumption is not valid in practice where the effect of MC varies for different directions. In this letter, a new method is proposed in order to estimate both source DOA angles and MC coefficients in the presence of direction-dependent MC. The proposed method iteratively estimates the source DOA angles using the MUSIC algorithm. In order to estimate the direction-dependent MC coefficients, a unified transformation approach is proposed, which can be applied for any array geometry. Then, a convex minimization problem is outlined using the signal-noise subspace orthogonality. The performance of the proposed method is evaluated for uniform linear and circular arrays. It is shown that the proposed method effectively estimates both source and array coupling parameters and it has superior performance than the conventional techniques.Öğe Joint-block-sparsity for efficient 2-D DOA estimation with multiple separable observations(Springer, 2019) Elbir, Ahmet MusabIn sparsity-based optimization problems, one of the major issue is computational complexity, especially when the unknown signal is represented in multi-dimensions such as in the problem of 2-D (azimuth and elevation) direction-of-arrival (DOA) estimation. In order to cope with this issue, this paper introduces a new sparsity structure that can be used to model the optimization problem in case of multiple data snapshots and multiple separable observations where the dictionary can be decomposed into two parts: azimuth and elevation dictionaries. The proposed sparsity structure is called joint-block-sparsity which enforces the sparsity in multiple dimensions, namely azimuth, elevation and data snapshots. In order to model the joint-block-sparsity in the optimization problem, triple mixed norms are used. In the simulations, the proposed method is compared with both sparsity-based techniques and subspace-based methods as well as the Cramer-Rao lower bound. It is shown that the proposed method effectively solves the 2-D DOA estimation problem with significantly low complexity and sufficient accuracy.Öğe L-shaped coprime array structures for DOA estimation(Springer, 2020) Elbir, Ahmet MusabThis paper proposes a new sparse array geometry for 2-D (azimuth and elevation) directionof-arrival (DOA) estimation based on coprime sampling. The proposed array structure is L-shaped coprime array (LCA) whose each portion is one dimensional coprime linear arrays in y- and z-dimensions. Each portion of the array is used separately for 1-D azimuth and elevation angle estimation. In order to obtain the paired DOA estimates the cross-covariance matrix of two portion of the array is utilized and the paired DOA angles are estimated. LCA provides to estimate K <= MN source directions with 2M+ N-1 sensors in each portion and totally 4M + 2N - 3 sensor elements. The proposed method is evaluated through numerical simulations and its performance is compared with other coprime planar array structures. It is shown that LCA has less computational complexity together with less real sensor elements and it provides superior performance as compared to the conventional 2-D coprime planar arrays.Öğe A Novel Data Transformation Approach for DOA Estimation With 3-D Antenna Arrays in the Presence of Mutual Coupling(Ieee-Inst Electrical Electronics Engineers Inc, 2017) Elbir, Ahmet MusabIn antenna array applications, mutual coupling (MC) is an important error source that should be corrected. In previous works, MC parameters are estimated by using a data transformation approach, where the unknown coupling coefficients are extracted into a single vector so that the parameter estimation is done accordingly. While this is an effective way to estimate the MC coefficients, this approach is only considered for the antenna arrays with special geometries such as linear, circular, and rectangular. In this letter, we introduce a new data transformation approach that is applicable for a general array geometry, namely 3-D arrays. The proposed approach is based on the decomposition of the MC matrix in terms of the unknown distinct MC coefficients. The performance of the proposed technique is evaluated through simulations for different 3-D arrays, and it is compared to the theoretical performance bounds. It is shown that the proposed technique can be used effectively for 3-D array structures with high estimation accuracy.Öğe Robust Hybrid Beamforming with Quantized Deep Neural Networks(IEEE Computer Society, 2019) Elbir, Ahmet Musab; Mishra, Kumar VijayHybrid beamforming is integral to massive multiple-input multiple-output (MIMO) communications in reducing the training overhead and hardware cost associated with large antenna arrays. Prior works have employed optimization and greedy search to jointly estimate the precoder and combiner weights. High computational complexity of these methods apart, their performance strongly relies on accurate channel information. In this paper, we propose a computationally efficient, deep learning approach that also provides robust performance against the deviations in the channel characteristics. Further, we employ a convolutional neural network with quantized weights (Q-CNN) so that it is deployable in mobile devices that have less memory resources and low overhead requirements. We show that the proposed Q-CNN, saved in at least 6 bits, yields superior performance over conventional massive MIMO hybrid beamforming. © 2019 IEEE.Öğe Sensor array calibration with joint-block-sparsity in the presence of multiple separable observations(Springer London Ltd, 2019) Elbir, Ahmet MusabIn sparsity-based optimization problems, one of the major issue is computational complexity, especially when the unknown signal is represented in multi-dimensions such as in the problem of 2-D (azimuth and elevation) direction-of-arrival (DOA) estimation. In this paper, a low-complexity sparsity-based method is proposed for DOA estimation in the presence of array imperfections such as mutual coupling. In order to reduce the complexity of the optimization problem, this paper introduces a new sparsity structure that can be used to model the optimization problem in case of multiple data snapshots and multiple separable observations where the dictionary can be decomposed into two parts: azimuth and elevation dictionaries. The proposed sparsity structure is called joint-block-sparsity which exploits the sparsity in both multiple dimensions, namely azimuth and elevation, and data snapshots. In order to model the joint-block-sparsity in the optimization problem, triple mixed norms are used. In the simulations, the proposed method is compared with both sparsity-based techniques and subspace-based methods as well as the Cramer-Rao lower bound. It is shown that the proposed method effectively calibrates the sensor array with significantly low complexity and sufficient accuracy.Öğe Two-Dimensional DOA Estimation via Shifted Sparse Arrays with Higher Degrees of Freedom(Springer Birkhauser, 2019) Elbir, Ahmet MusabSparse antenna arrays provide larger virtual arrays to estimate the direction of arrivals (DOAs) of more sources than the number of physical antennas in the array. While the degrees of freedom (DOF) can be increased by the special structure of the antenna array, a shift in the antenna positions can generate new lags in the difference co-array; hence, more sources can be resolved. In this paper, we propose shifted sparse array structures composed of two overlapping arrays shifted by one lag. It is shown that the shifting property fills the holes in the co-array, which yields larger virtual arrays. We derive stationary and moving array models where overlapping sparse arrays can be realized. The proposed shifting property is applied to coprime, nested and sparse linear arrays, and we show that the proposed technique guaranteed to increase the DOF. Using the proposed sparse array structures, we also propose a 2-D DOA estimation algorithm by utilizing the cross-covariance matrix of an L-shaped sparse array. The performance of the proposed approach is evaluated through numerical simulations, and we show that it can resolve more sources than the conventional sparse arrays with the same number of physical antennas, providing less computational complexity.Öğe V-Shaped Sparse Arrays for 2-D DOA Estimation(Springer Birkhauser, 2019) Elbir, Ahmet MusabThis paper proposes a new sparse array geometry for 2-D (azimuth and elevation) direction-of-arrival (DOA) estimation. The proposed array geometry is V-shaped sparse array, and it is composed of two linear portions which are crossing each other. The degrees of freedom of the sparse array are enhanced by sparse sampling property. In this respect, V-shaped coprime (VCA) and V-shaped nested array (VNA) structures are developed. VCA can resolve both azimuth and elevation angles up to MN sources with 2M+N-1 sensors in each portion, and the total number of sensors is 4M+2N-3. VNA can resolve O(N2) sources with 2N sensors. Instead of 2-D grid search, the proposed method computes 1-D search for azimuth and elevation angle estimation in a computational efficient way. In order to solve the pairing problem in 2-D scenario, the cross-covariance matrix of two portion is utilized and 2-D paired DOA estimation is performed. The performance of the proposed method is evaluated with numerical simulations, and it is shown that the proposed array geometries VCA and VNA can provide much less sensors as compared to the conventional coprime planar arrays.
