A Novel Adaptive AI-Based Framework for Node Scheduling Algorithm Selection in Safety-Critical Wireless Sensor Networks

Wireless Sensor Networks (WSNs) play a critical role in diverse applications, from environmental monitoring to mission-critical operations. In this type on applications, selecting the most suitable node scheduling algorithm for a given scenario remains a challenge, as no single approach consistently outperforms others under all conditions. Consequently, this study presents an AI-driven selection framework that evaluates scenario-specific requirements—such as coverage, connectivity, and network lifetime—to identify the optimal scheduling algorithm from a pool of algorithms including Hidden Markov Models (HMM), BAT, Bird Flocking, Self-Organizing Feature Maps (SOFM), and Long Short-Term Memory (LSTM). This framework employs a trained neural network, informed by a simulated dataset, to match algorithms to five real-world scenarios: healthcare monitoring, military operations, industrial IoT monitoring, forest fire detection, and disaster recovery. Experimental results show that LSTM frequently achieves near-optimal performance across scenarios, improving dependability e.g., excelling in adaptability, latency reduction, and fault tolerance, while other algorithms demonstrate valuable strengths in specific metrics such as connectivity (HMM) and lifetime (Bird Flocking). These findings demonstrate the effectiveness of the proposed framework in selecting scenario-aware, high-performance WSN scheduling solutions. They also highlight the complementary strengths of different algorithms in meeting diverse operational requirements across scenarios.