This paper presents a real-time ECG monitoring system that encrypts heart signals for secure cloud transmission in telemedicine. The key innovation is using each patient's own ECG signal characteristics to generate unique encryption keys on the fly — so the encryption key is biometric and patient-specific, refreshed continuously, and doesn't need traditional key exchange. The system encrypts data immediately after acquisition, sends it to the cloud, and doctors decrypt it remotely using matching keys.
Main takeaways:
- Uses a learnable key generator that derives encryption keys from the patient's ECG signal itself
- Keys control a chaotic logistic map that encrypts the data through permutation and XOR operations
- Achieves strong security metrics (Shannon entropy 7.678 bits, passes NIST randomness tests)
- Encryption adds minimal latency, preserving real-time performance
- Reconstruction is nearly lossless (PSNR > 52 dB), so diagnostic features are preserved