The dynamic risk assessment identifies, evaluates, and mitigates the vulnerabilities that compromise the reliability of the cyber-physical system (CPS). The slow degradation in network behaviour is known as the gradual and long-term shifts in traffic characteristics (variations in latency, packet flow distribution, and protocol level patterns) that accumulate over time. These subtle variations cause faults, stealthy intrusion, and network aging in CPS. However, the prevailing works overlooked the slow degradation that persists in the network behaviour. Thus, Quantum State-based Exponentially Weighted Moving Average (QS-EWMA)-based slow drift analysis is proposed. Initially, the medical devices are registered in the cloud server, followed by data sensing. During data transfer, the intrusion is detected by the Network Intrusion Detection System (NIDS). In NIDS, the data collection, pre-processing, and feature extraction are performed. Further, the QS-EWMA-based slow drift analysis is performed, followed by Sparse Random Walk Softplus S-shaped Rectified Gated Recurrent Unit (SRWSSR-GRU)-based intrusion detection. Next, the dependency-aware aggregation is performed using the Choquet k-Additive Function Integral Model (CkAFIM). Then, the uncertainty-based risk is assessed using Min–Max Normalization-based Neutrosophic Logic System (MMN-NLS). Here, the explainability is improved using SHapley Max–Abs Scaling Additive exPlanation (SMAS-HAP). Finally, the decision-making is performed using the Markov Linear Discrete-Time Propagation Decision Process (MLDTPDP). Hence, the proposed system effectively assessed the risk with an Indeterminacy Detection Rate (IDR) of 13.43%, showing superiority over prevailing works.
Kiruthika et al. (Thu,) studied this question.