Spatiotemporal trust evaluation using a neutrosophic model under sensor uncertainty

In dynamic and harsh environments, ensuring the accuracy and reliability of sensor data is essential for robust real-time decision-making. Conventional trust evaluation approaches often overlook the compounded uncertainties arising from sensor degradation, environmental fluctuations, and measurement noise, relying instead on binary or probabilistic models that inadequately capture partial truth and indeterminacy. This study introduces a neutrosophic-based framework for evaluating sensor trust under uncertainty and degradation. The proposed model characterizes truth, indeterminacy, and falsity through neutrosophic membership functions derived from sensor attributes such as coverage, battery level, noise, drift, and failure risk. Based on thresholded neutrosophic values, the sensing field is segmented into three zones: interior (trustworthy), boundary (ambiguous), and exterior (unreliable) to delineate sensor reliability. Trust scores are then computed by aggregating normalized truth values over effective coverage areas, while temporal degradation is captured through a spatiotemporal modeling process. Extensive Monte Carlo simulations with 50 runs are employed to assess resilience and variability, with performance evaluated using per-sensor metrics including mean trust, standard deviation, and top-3 occurrence frequency. Results show that the framework robustly tracks sensor trust dynamics, captures spatial and temporal degradation patterns, and enables informed, real-time sensor selection in uncertain environments.