Downlink Power Allocation in Massive MIMO via Deep Learning: Adversarial Attacks and Training

被引:8
|
作者
Manoj, B. R. [1 ,2 ]
Sadeghi, Meysam [3 ]
Larsson, Erik G. [4 ]
机构
[1] Linkoping Univ, Dept Elect Engn, Div Commun Syst, S-58183 Linkoping, Sweden
[2] Indian Inst Technol Guwahati, Dept Elect & Elect Engn, Gauhati 781039, India
[3] Linkoping Univ, Div Commun Syst, S-58183 Linkoping, Sweden
[4] Linkoping Univ, Dept Elect Engn ISY, S-58183 Linkoping, Sweden
关键词
Adversarial attacks; adversarial training; black-box attack; deep neural networks; massive MIMO; regression; white-box attack; CLASSIFICATION; NETWORKS;
D O I
10.1109/TCCN.2022.3147203
中图分类号
TN [电子技术、通信技术];
学科分类号
0809 ;
摘要
The successful emergence of deep learning (DL) in wireless system applications has raised concerns about new security-related challenges. One such security challenge is adversarial attacks. Although there has been much work demonstrating the susceptibility of DL-based classification tasks to adversarial attacks, regression-based problems in the context of a wireless system have not been studied so far from an attack perspective. The aim of this paper is twofold: (i) we consider a regression problem in a wireless setting and show that adversarial attacks can break the DL-based approach and (ii) we analyze the effectiveness of adversarial training as a defensive technique in adversarial settings and show that the robustness of DL-based wireless system against attacks improves significantly. Specifically, the wireless application considered in this paper is the DL-based power allocation in the downlink of a multicell massive multi-input-multi-output system, where the goal of the attack is to yield an infeasible solution by the DL model. We extend the gradient-based adversarial attacks: fast gradient sign method (FGSM), momentum iterative FGSM, and projected gradient descent method to analyze the susceptibility of the considered wireless application with and without adversarial training. We analyze the deep neural network (DNN) models performance against these attacks, where the adversarial perturbations are crafted using both the white-box and black-box attacks.
引用
收藏
页码:707 / 719
页数:13
相关论文
共 50 条
  • [41] Optimal Power Allocation and Training Duration for Uplink Multiuser Massive MIMO Systems With MMSE Receivers
    Muharar, Rusdha
    IEEE ACCESS, 2020, 8 : 23378 - 23390
  • [42] HyperRNN: Deep Learning-Aided Downlink CSI Acquisition via Partial Channel Reciprocity for FDD Massive MIMO
    Liu, Yusha
    Simeone, Osvaldo
    SPAWC 2021: 2021 IEEE 22ND INTERNATIONAL WORKSHOP ON SIGNAL PROCESSING ADVANCES IN WIRELESS COMMUNICATIONS (IEEE SPAWC 2021), 2020, : 31 - 35
  • [43] Backdoor Attacks Against Deep Learning-based Massive MIMO Localization
    Zhao, Tianya
    Wang, Xuyu
    Mao, Shiwen
    IEEE CONFERENCE ON GLOBAL COMMUNICATIONS, GLOBECOM, 2023, : 2796 - 2801
  • [44] Robustness and Security in Deep Learning: Adversarial Attacks and Countermeasures
    Kaur, Navjot
    Singh, Someet
    Deore, Shailesh Shivaji
    Vidhate, Deepak A.
    Haridas, Divya
    Kosuri, Gopala Varma
    Kolhe, Mohini Ravindra
    JOURNAL OF ELECTRICAL SYSTEMS, 2024, 20 (03) : 1250 - 1257
  • [45] Joint Flexible Duplexing and Power Allocation with Deep Reinforcement Learning in Cell-Free Massive MIMO System
    Deng, Danhao
    Wang, Chaowei
    Zhang, Zhi
    Li, Lihua
    Wang, Weidong
    CHINA COMMUNICATIONS, 2023, 20 (04) : 73 - 85
  • [46] Deep Reinforcement Learning Based Resource Allocation for Network Slicing With Massive MIMO
    Yan, Dandan
    Ng, Benjamin K.
    Ke, Wei
    Lam, Chan-Tong
    IEEE ACCESS, 2023, 11 : 75899 - 75911
  • [47] Melanoma detection using adversarial training and deep transfer learning
    Zunair, Hasib
    Ben Hamza, A.
    PHYSICS IN MEDICINE AND BIOLOGY, 2020, 65 (13)
  • [48] 1-bit Precoding Algorithm for Massive MIMO OFDM Downlink Systems with Deep Learning
    Zhou C.
    Wen L.
    Qian H.
    Kang K.
    Dianzi Yu Xinxi Xuebao/Journal of Electronics and Information Technology, 2024, 46 (03): : 886 - 894
  • [49] Closed-form ASE of Downlink Massive MIMO System with Uplink and Downlink Training
    Dong, Yitian
    Qiu, Ling
    2015 INTERNATIONAL CONFERENCE ON WIRELESS COMMUNICATIONS & SIGNAL PROCESSING (WCSP), 2015,
  • [50] Deep Transfer Learning-Based Downlink Channel Prediction for FDD Massive MIMO Systems
    Yang, Yuwen
    Gao, Feifei
    Zhong, Zhimeng
    Ai, Bo
    Alkhateeb, Ahmed
    IEEE TRANSACTIONS ON COMMUNICATIONS, 2020, 68 (12) : 7485 - 7497