Multi-tier heterogeneous Model for real-time hazard detection in Air Particulate

Abstract

ABSTRACT In recent years, several nations have built an environmental gathering data system. Using multi-tier heterogeneous networks to evaluate concentrations for an infinite number of contaminants, a novel mathematical framework was suggested in this study. The efficiency of the suggested method was examined using actual measurements of municipal particle pollution concentrations. The method can detect local inhomogeneities that expose people to dangerous air pollution in actual time, such as a lack of toxin dispersion and concentration. Moreover, real-time inhomogeneity identification in a contaminated environment is now possible thanks to technology. In credible emergency scenarios with contaminated air, the value of local heterogeneity was calculated at 0.015 level or 8 counts. As a result, the level of heterogeneity was decreased to 0.0025 units or 2 counts. Long Short-Term Memory (LSTM) provided precise data on the accessible bandwidth and co-channel disruption, making the suggested technique easy to apply and adaptable to any heterogeneous network design. The labeling & judging elements would then be built once the expected outcome of the suggested approach had been evaluated using precision, recall, F1 scoring, and dependability measures. The overall results demonstrate that our performance using LSTM was superior to the Hybrid Ant Colony Optimizer (HACO) method.