An Energy-Efficient and Modeling Attack-Resilient SNN-Driven PUF Security for Internet of Medical Things

The healthcare industry is rapidly evolving with the digitalisation of medical devices. Despite the advantages of IoMT devices, they remain vulnerable to cyberattacks that threaten patient data security. Authentication typically relies on a static key, making it susceptible mainly to modelling attacks. Existing techniques fail to balance security with lightweight operation for resource-constrained IoMT. A Physically Unclonable Function (PUF)-based technique provides genuine authentication with on-the-fly key generation but lacks resilience against modelling attacks. The proposed work presents a fusion of facial biometrics and a Spiking Neural Network(SNN) with a PUF. This fusion generates stronger entropy than conventional methods that rely on fixed stored keys. This work strengthens authentication by integrating user-specific facial biometrics, device-specific PUF responses, and SNN-induced nonlinearity. It enables joint authentication of both the device and the legitimate user. The experimental evaluation is conducted on the Digilent Basys3 Artix 7 FPGA board and achieves higher randomness in the generated responses, with a Shannon entropy of 0.693 nats and a normalized Hamming distance of 0.50. The security evaluation demonstrates that machine learning and deep learning attacks remain random-guess predictions, indicating that the proposed lightweight Face-SNN-PUF resists effective modelling of the challenge-response. Tailored for non-life-critical IoMT devices, the proposed resilient framework provides practical and energy-efficient authentication.

@article{T2026,
  title     = {An Energy-Efficient and Modeling Attack-Resilient SNN-Driven PUF Security for Internet of Medical Things},
  journal   = {International Journal of Advanced Computer Science and Applications},
  volume    = {17},
  number    = {9},
  year      = {2026},
  publisher = {The Science and Information Organization},
  author    = {Agila Harshini T and Harini Sriraman},
  doi       = {10.14569/IJACSA.2026.0170938},
  url       = {https://doi.org/10.14569/IJACSA.2026.0170938}
}