T-ADD: Enhancing DOA Estimation Robustness Against Adversarial Attacks

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Zheng, Shilian, Wu, Xiaoxiang, Zhang, Luxin, Yue, Keqiang, Qi, Peihan, Zhao, Zhijin
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866911313156898816
author Zheng, Shilian
Wu, Xiaoxiang
Zhang, Luxin
Yue, Keqiang
Qi, Peihan
Zhao, Zhijin
author_facet Zheng, Shilian
Wu, Xiaoxiang
Zhang, Luxin
Yue, Keqiang
Qi, Peihan
Zhao, Zhijin
contents Deep learning has achieved remarkable success in direction-of-arrival (DOA) estimation. However, recent studies have shown that adversarial perturbations can severely compromise the performance of such models. To address this vulnerability, we propose Transformer-based Adversarial Defense for DOA estimation (T-ADD), a transformer-based defense method designed to counter adversarial attacks. To achieve a balance between robustness and estimation accuracy, we formulate the adversarial defense as a joint reconstruction task and introduce a tailored joint loss function. Experimental results demonstrate that, compared with three state-of-the-art adversarial defense methods, the proposed T-ADD significantly mitigates the adverse effects of widely used adversarial attacks, leading to notable improvements in the adversarial robustness of the DOA model.
format Preprint
id arxiv_https___arxiv_org_abs_2512_10496
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle T-ADD: Enhancing DOA Estimation Robustness Against Adversarial Attacks
Zheng, Shilian
Wu, Xiaoxiang
Zhang, Luxin
Yue, Keqiang
Qi, Peihan
Zhao, Zhijin
Signal Processing
Deep learning has achieved remarkable success in direction-of-arrival (DOA) estimation. However, recent studies have shown that adversarial perturbations can severely compromise the performance of such models. To address this vulnerability, we propose Transformer-based Adversarial Defense for DOA estimation (T-ADD), a transformer-based defense method designed to counter adversarial attacks. To achieve a balance between robustness and estimation accuracy, we formulate the adversarial defense as a joint reconstruction task and introduce a tailored joint loss function. Experimental results demonstrate that, compared with three state-of-the-art adversarial defense methods, the proposed T-ADD significantly mitigates the adverse effects of widely used adversarial attacks, leading to notable improvements in the adversarial robustness of the DOA model.
title T-ADD: Enhancing DOA Estimation Robustness Against Adversarial Attacks
topic Signal Processing
url https://arxiv.org/abs/2512.10496