Authors: Mahesh S, Dr. Jayasree P
Abstract: The use of Internet of Medical Things (IoMT) in smart healthcare ecosystems has led to a large-scale attack surface which current perimeter-based security solutions cannot sufficiently cover. Resource limitations, diverse protocols, and life-critical telemetry require security designs which function across the device, edge, and cloud layers while avoiding significant computation costs. In this work, we design a cross-layer security architecture which involves selective elliptic curve cryptography for IoT device authentication, sidecar proxy zero trust enforcement at the edge, and protocol-aware intrusion detection for behavior monitoring. Our design utilizes only asymmetric cryptography at registration time while shifting online mutual authentication to hash-based message authentication code and timestamp validation in order to reduce computational cost of each session without losing authentication assurances. Experimental evaluation on emulated IoMT traffic shows that our architecture is able to provide sub-10 ms end-to-end authentication latency, preserve masquerade and replay attempt detection under the attack injection, and retain tenant isolation in multi-tenant cloud setups. Comparative evaluation with state-of-the-art frameworks proves that our solution provides a better trade-off between latency, energy consumption, and cross-layer visibility.
DOI: http://doi.org/
