Blockchain-Enabled Uncertainty-Aware Passive Wi-Fi Localization for Secure Critical Infrastructure Sensor Networks.

Area of Science:

• Cybersecurity
• Wireless Networking
• Sensor Networks

Background:

• Existing passive Wi-Fi localization methods lack calibrated uncertainty and auditable data trails, hindering automated responses in security operations centers (SOCs).
• Uncalibrated confidence values in localization systems lead to asymmetric operational costs due to potential erroneous actions or missed escalations.

Purpose of the Study:

• To present a Blockchain-Enabled Uncertainty-Aware Passive Wi-Fi Localization framework for heterogeneous sensor networks.
• To address limitations in automated SOC response by providing calibrated uncertainty and evidentiary trust.

Main Methods:

• Developed a framework combining Bayesian estimation with robust loss functions and range-dependent error modeling.
• Implemented an uncertainty calibration component converting posterior confidence into SOC response modes (AUTO, VERIFY, OBSERVE) using empirical coverage metrics.
• Integrated a permissioned, hybrid on-chain/off-chain logging layer for tamper-evident auditability and scalability.

Main Results:

• The framework derives a posterior spatial distribution with calibrated uncertainty from passive Wi-Fi observations.
• Empirical coverage metrics and reliability diagrams were used to calibrate SOC response thresholds.
• Field validation demonstrated a practical operational validity boundary of approximately 40 meters for confident automated action.

Conclusions:

• The proposed framework integrates passive sensing, uncertainty-aware localization, and blockchain-based trust for secure critical infrastructure sensor networks.
• Key contributions include a posterior-distribution-based localization pipeline, empirical metrics for SOC response calibration, and a hybrid ledger architecture.
• The study establishes a 40m operational validity boundary for the developed passive Wi-Fi localization system.