2025/02/26 by Youval Klioui, Klioui, Youval · 2 citations
Computer Science · Engineering · #Artificial neural network #Circulant matrix #Computational complexity theory #Direction-of-Arrival Estimation Techniques #Mean squared error #Memory footprint #Operator (biology) #Selection (genetic algorithm) #Sparse and Compressive Sensing Techniques #Speech and Audio Processing #Square (algebra) #eess.SP
paper · pdf · doi:10.48550/arxiv.2502.19076
published in arXiv (Cornell University) (Cornell University)
openalex publication_date 2025/02/26 · openalex created_date 2025/10/10 · openalex updated_date 2026/08/05
This paper introduces CADMM-Net and CHADMM-Net, two deep neural networks for direction of arrival estimation within the least-absolute shrinkage and selection operator (LASSO) framework. These two networks are based on a structured deep unfolding of the alternating direction method of multipliers (ADMM) algorithm through the use of circulant as well as Hermitian-circulant matrices. Along with a computational complexity of O(Nlog(N)) per layer for the inference, where N is the length of the dictionary A, they additionally exhibit a memory footprint of N and approximately half of N for CADMMNet and CHADMM-Net, respectively, compared with N2 for ADMM-Net. Furthermore, these structured networks exhibit a competitive performance against ADMM-Net, LISTA, TLISTA, and THLISTA with respect to the detection rate, the angular root-mean square error, and the normalized mean squared error.