E-PTES-S: Enhanced Trust Evaluation via Multidimensional Spatiotemporal Fusion and Variance-Based Stability Sequence...

Area of Science:

• Internet of Things (IoT) Security
• Trust Management in Wireless Sensor Networks
• Cybersecurity and Data Integrity

Background:

• Mobile data collectors (MDCs) are crucial for IoT sensing networks.
• Insider threats like on-off attacks and spatiotemporal fabrication challenge MDC trustworthiness.
• Existing trust evaluation methods lack sufficient evidence dimensions and behavioral stability quantification.

Purpose of the Study:

• To propose an enhanced proactive trust evaluation system (E-PTES-S) for IoT networks.
• To improve trust evaluation accuracy against insider threats by integrating multidimensional evidence and stability metrics.
• To enhance robustness against data scarcity and physical-layer uncertainties.

Main Methods:

• Developed E-PTES-S integrating five evidence factors, stability computation, and adaptive weights.
• Introduced regional consistency (TRC) and task timeliness (TTT) to address location and time deviations.
• Utilized a sliding window for integrated evidence sequences, including a novel continuous stability sequence (FCSS) with variance-based incentives.

Main Results:

• E-PTES-S achieved a 98.7% normal node detection rate under dynamic conditions, surpassing PTES and Trust-SIoT.
• Demonstrated improved cumulative data collection profit by 4.8%.
• Maintained an 84.4% detection rate even under severe physical-layer uncertainties and shadowing.

Conclusions:

• E-PTES-S significantly enhances trust evaluation accuracy and robustness in IoT networks.
• The system effectively mitigates insider threats by incorporating behavioral stability and spatiotemporal consistency.
• Proposed methods offer a more reliable and profitable solution for securing IoT sensing networks.