The performance of traditional high-resolution direction-of-arrival(DOA)estimation methods is sensitive to the inaccurate knowledge on prior information,including the position of ar-ray elements,array gain and phase,a...The performance of traditional high-resolution direction-of-arrival(DOA)estimation methods is sensitive to the inaccurate knowledge on prior information,including the position of ar-ray elements,array gain and phase,and the mutual coupling between the array elements.Learning-based methods are data-driven and are expected to perform better than their model-based counter-parts,since they are insensitive to the array imperfections.This paper presents a learning-based method for DOA estimation of multiple wideband far-field sources.The processing procedure mainly includes two steps.First,a beamspace preprocessing structure which has the property of fre-quency invariant is applied to the array outputs to perform focusing over a wide bandwidth.In the second step,a hierarchical deep neural network is employed to achieve classification.Different from neural networks which are trained through a huge data set containing different angle combinations,our deep neural network can achieve DOA estimation of multiple sources with a small data set,since the classifiers can be trained in different small subregions.Simulation results demonstrate that the proposed method performs well both in generalization and imperfections adaptation.展开更多
Array imperfections will lead to serious performance degradation of the deep neural network(DNN)based direction of arrival(DOA)estimation in the low earth orbit(LEO)satellite communication by producing a mismatch betw...Array imperfections will lead to serious performance degradation of the deep neural network(DNN)based direction of arrival(DOA)estimation in the low earth orbit(LEO)satellite communication by producing a mismatch between inference data and training data.In this paper,we propose a lightweight deep learning-based algorithm for array imperfection correction and DOA estimation.By preprocessing the covariance matrix of the array antenna outputs to the image,the array imperfection correction and DOA estimation problems are correspondingly converted into the image-to-image transformation task and image recognition task.Furthermore,for the deployment of real-time DNN-based DOA estimation on the resource-constrained edge system,generative adversarial network(GAN)model compression is applied to obtain a lightweight student generator of Pix2Pix for array imperfection correction.The Mobilenet-V2 is then used to extract the DOA information from the covariance matrix image.Simulations results demonstrate that the DOA estimation performance is significantly improved through the array imperfection correction.The proposed algorithm also better satisfies the real-time demand with decreased inference time on the resource-constrained edge system.展开更多
基金the National Natural Sci-ence Foundation of China(No.62101340).
文摘The performance of traditional high-resolution direction-of-arrival(DOA)estimation methods is sensitive to the inaccurate knowledge on prior information,including the position of ar-ray elements,array gain and phase,and the mutual coupling between the array elements.Learning-based methods are data-driven and are expected to perform better than their model-based counter-parts,since they are insensitive to the array imperfections.This paper presents a learning-based method for DOA estimation of multiple wideband far-field sources.The processing procedure mainly includes two steps.First,a beamspace preprocessing structure which has the property of fre-quency invariant is applied to the array outputs to perform focusing over a wide bandwidth.In the second step,a hierarchical deep neural network is employed to achieve classification.Different from neural networks which are trained through a huge data set containing different angle combinations,our deep neural network can achieve DOA estimation of multiple sources with a small data set,since the classifiers can be trained in different small subregions.Simulation results demonstrate that the proposed method performs well both in generalization and imperfections adaptation.
基金supported in part by the National Natural Science Foundation of China(NSFC)under Grant 61971379in part by the Key Research and Development Program of Zhejiang Province under Grant 2020C03100+2 种基金in part by the Leading Innovative and Entrepreneur Team Introduction Program of Zhejiang under Grant 2018R01001in part by the Fundamental Research Funds for the Central Universities under Grant 226202200096in part by the Program of Innovation 2030 on Smart Ocean in Zhejiang University under Grant 129000*194232201.
文摘Array imperfections will lead to serious performance degradation of the deep neural network(DNN)based direction of arrival(DOA)estimation in the low earth orbit(LEO)satellite communication by producing a mismatch between inference data and training data.In this paper,we propose a lightweight deep learning-based algorithm for array imperfection correction and DOA estimation.By preprocessing the covariance matrix of the array antenna outputs to the image,the array imperfection correction and DOA estimation problems are correspondingly converted into the image-to-image transformation task and image recognition task.Furthermore,for the deployment of real-time DNN-based DOA estimation on the resource-constrained edge system,generative adversarial network(GAN)model compression is applied to obtain a lightweight student generator of Pix2Pix for array imperfection correction.The Mobilenet-V2 is then used to extract the DOA information from the covariance matrix image.Simulations results demonstrate that the DOA estimation performance is significantly improved through the array imperfection correction.The proposed algorithm also better satisfies the real-time demand with decreased inference time on the resource-constrained edge system.
