A Lightweight Cryptographic Framework for Medical IoT Using PUF-Enhanced ECDH and Proof-of-Trusted-Hash Blockchain

Internet of Medical Things (IoMT) systems handle sensitive clinical data under strict hardware limits, so secure key management remains difficult.Attacks often arise because stored keys expose devices to cloning and extraction during physical compromise.Several existing schemes also use multi-step cryptographic architectures and methods, that strain sensor capacity and reduce reliability in practical deployments.The aim of this study is to develop a compact security method that uses device variation for key generation and supports clear auditing in medical environments.The proposed method addresses these issues through a hybrid method that applies Static Random-Access Memory Physical Unclonable Function (SRAM-PUF) variation, Elliptic Curve Cryptography (ECC) key exchange, Advanced Encryption Standard (AES) key wrapping, and a Proof-of-Trusted-Hash ledger.A Physical Unclonable Function (PUF) module generates a raw response and supports helper data construction, so a stable root key forms without stored secrets.A PUF-driven Elliptic Curve Diffie-Hellman (ECDH) step creates a session key with one elliptic curve operation, so sensor load stays small.An AES key wrap method protects transport keys during movement between nodes.A confirmation token links session values with payload metadata and supports later checks.A blockchain layer records compact trust entries and supports review by peer nodes.The method solves the above problems by removing permanent keys, by grounding trust in hardware variation, and by providing a clear audit path for key activity.Prototype tests used an SRAM-PUF unit and a light ECC library on constrained hardware.Experiments show clear gains under three network settings.Throughput reaches 940 kbps in stable conditions and stays above 710 kbps in heavy traffic.Latency ranges from 42 ms to 91 ms and response time stays within 58 ms to 118 ms.Key generation stays between 2.9 ms and 3.4 ms and encryption stays between 6.2 ms and 7.5 ms.Decryption stays within 7.1 ms to 8.4 ms.These results confirm that the proposed method supports secure and predictable medical communication on constrained IoT hardware.