Privacy‑Preserving Camera Surveillance for 6G Edge Networks Using FHE

Purpose and goal

The project develops an integrity-preserving camera solution for future 6G environments. Through fully homomorphic encryption, facial recognition can be performed entirely locally and in encrypted form, without exposing sensitive data. The result is a proof-of-concept that demonstrates how secure, data-minimizing edge AI can be used in public safety and critical infrastructure.

Expected effects and result

The project is expected to deliver a working proof-of-concept that demonstrates that camera surveillance can be done in fully encrypted form at the edge of the network. The results provide metrics on performance, energy efficiency and accuracy, as well as guidance on how privacy-preserving AI can be introduced in 6G environments. The effect will be increased data security and minimized exposure of personal data.

Planned approach and implementation

The project is implemented by developing and testing a prototype for encrypted camera surveillance at the network edge. The work includes pipeline design, implementation of encrypted feature matching with FHE and optimization for 6G-near environments. The system is evaluated through measurements of performance, accuracy, energy and bandwidth.