Corpus ID: 54469562

Deep Neural Networks Meet CSI-Based Authentication

  title={Deep Neural Networks Meet CSI-Based Authentication},
  author={Amirhossein Yazdani Abyaneh and Ali Hosein Gharari Foumani and V. Pourahmadi},
The first step of a secure communication is authenticating legible users and detecting the malicious ones. [...] Key Method Our approach presents a way of extracting features from raw CSI measurements which are stable towards rotation. We extract these features by the means of a deep neural network. We also present a scenario in which users can be {efficiently} authenticated while they are at certain locations in an environment (even if they rotate); and, they will be rejected if they change their location…Expand
1 Citations
Physical-Layer Authentication Using Channel State Information and Machine Learning
  • K. S. Germain, F. Kragh
  • Computer Science, Engineering
  • 2020 14th International Conference on Signal Processing and Communication Systems (ICSPCS)
  • 2020
  • PDF


Authenticating Users Through Fine-Grained Channel Information
  • 26
  • Highly Influential
  • PDF
Broadcasting Into the Uncertainty: Authentication and Confidentiality by Physical-Layer Processing
  • 58
Detecting and Localizing Identity-Based Attacks in Wireless and Sensor Networks
  • 185
  • PDF
Physical layer based authentication without phase detection
  • 5
ProxiMate: proximity-based secure pairing using ambient wireless signals
  • 189
  • PDF
Wireless device identification with radiometric signatures
  • 631
  • PDF
Deep Learning for Massive MIMO CSI Feedback
  • 222
  • PDF
Batch Normalization: Accelerating Deep Network Training by Reducing Internal Covariate Shift
  • 21,987
  • Highly Influential
  • PDF
Deep Sparse Rectifier Neural Networks
  • 4,690
  • PDF
Incorporating Nesterov Momentum into Adam
  • 789