AN EDGE-CENTRIC SECURITY RISK ASSESSMENT FRAMEWORK FOR INTERNET OF MEDICAL THINGS WITH QOS-AWARE SIMULATION

Abstract

Background The rapid adoption of the Internet of Medical Things (IoMT) has enabled continuous monitoring and transmission of sensitive healthcare information through interconnected medical sensors, wearable devices, edge nodes, and cloud services. However, the heterogeneous and resource-constrained nature of IoMT environments introduces significant security risks, while attacks on edge-connected medical devices can adversely affect both security and quality of service (QoS). Existing studies have primarily focused on individual security mechanisms, intrusion detection, authentication, or blockchain-based protection, with comparatively less emphasis on integrating security-risk assessment with edge-resource and QoS evaluation. Recent literature also demonstrates the need for systematic approaches to IoMT risk assessment and management. This paper proposes an Edge-centric IoMT Security Risk Assessment (EISRA) framework for evaluating security risks in edge-enabled IoMT environments. The proposed framework models heterogeneous medical devices, edge nodes, an edge gateway, and a healthcare cloud while considering multiple security threats, including Distributed Denial of-Service (DDoS), Man-in-the-Middle, replay, data injection, and device impersonation attacks. A composite risk score is formulated using likelihood, impact, device criticality, vulnerability, and edge exposure. A simulation environment is developed to evaluate the proposed framework under varying IoMT device densities and attack intensities. The performance is assessed using security and QoS parameters including risk score, latency, throughput, packet delivery ratio, CPU utilization, memory utilization, energy consumption, and network overhead. The simulation results are subsequently compared with representative existing approaches under common evaluation criteria. The framework provides an integrated basis for evaluating security risk and edge-level QoS performance in IoMT environments.