RF Fingerprinting of DSP Implementations using Machine Learning: An Over-the-Air Case Study

Key points

• This work instead investigates whether differences between digital signal processing implementations of a transmitter can be identified from their radiated signals.1
• As a practical case study, five different software-based DSP implementations for generating a 2-FSK RTTY communication waveform are considered.1
• In a monitoring receiver, a classifier can identify these DSP implementations using convolutional neural networks that are trained on augmented software-generated signals represented as raw IQ, power spectrum or instantaneous frequency.1
• The results demonstrate that DSP implementations leave distinguishable characteristics in the transmitted signals and suggest that DSP-dependent fingerprinting can provide complementary information for specific emitter identification systems.1

Sources (1)

• [1]RF Fingerprinting of DSP Implementations using Machine Learning: An Over-the-Air Case Study
  arXiv (AI, ML, NLP, CV, robotics, multi-agent) · Oct 11, 09:47 AM

  Radio frequency fingerprinting aims to identify transmitters by exploiting small differences in the emitted signal waveform.

  This work instead investigates whether differences between digital signal processing implementations of a transmitter can be identified from their radiated signals.

Extractive summary: sentences quoted from the sources.