AI-based routing algorithms improve energy efficiency, latency, and data reliability in wireless sensor networks.

Area of Science:

• Computer Science
• Electrical Engineering
• Artificial Intelligence

Background:

• Wireless Sensor Networks (WSNs) face challenges in dynamic environments with limited resources.
• Traditional routing protocols often lack adaptability to real-time network changes.
• Static or standalone algorithms struggle to optimize WSN performance effectively.

Purpose of the Study:

• To propose a novel modular AI-based routing framework for WSNs.
• To integrate multiple AI techniques for adaptive and optimized routing.
• To enhance WSN performance metrics like packet delivery, latency, and energy efficiency.

Main Methods:

• Developed a modular framework integrating Reinforcement Learning (RL), Supervised Learning, Genetic Algorithms (GA), and Particle Swarm Optimization (PSO).
• Implemented a structured decision-making pipeline for dynamic adaptation to network topology, traffic, and energy.
• Utilized MATLAB R2021b for simulations to evaluate the framework's performance.

Main Results:

• The proposed AI framework demonstrated significant improvements in packet delivery ratio and energy efficiency.
• End-to-end latency was reduced compared to traditional routing protocols.
• The hybrid AI approach showed enhanced adaptability in dynamic WSN environments.

Conclusions:

• Hybrid AI strategies offer a promising approach for optimizing WSN routing.
• The modular framework provides a foundation for future real-world WSN implementations.
• Addressing challenges like scalability, resource usage, and security is crucial for deployment.