Load Balancing in IoT: Leveraging Learning Automata for RPL Optimization

The Internet of Things (IoT) is a technologically groundbreaking concept that offers considerable obstacles because of the restricted hardware and communication capabilities of its gadgets, but also enhances a wide range of applications through device interconnectivity. To help navigate these challenges, the Internet Engineering Task Force (IETF) created the Routing Protocol for Low-Power and Lossy Networks (RPL) to answer the unique needs of IoT environments. However, load allocation and traffic congestion are issues with RPL that negatively impact network performance and reliability. This study suggests a new enhancement to RPL by incorporating learning automata designed to maximise network traffic allocation. By dynamically adjusting routing decisions based on real-time network conditions, this improved protocol—the Learning Automata-based Load-Aware RPL (LALARPL)—achieves more effective load balancing and dramatically reduces network congestion. Extensive statistical analysis across 30 independent simulation runs demonstrates that LALARPL significantly outperforms current techniques ( p < 0.001 for most metrics), achieving 6.45–8.43% improvements in packet delivery ratio, 3.26–7.38% increase in throughput, and 7.52–10.06% decrease in end-to-end latency with large to very large effect sizes (Cohen’s d ranging from 1.2 to 5.3). The results demonstrate how our strategy may improve IoT network performance and prolong the life of network elements. The ability of learning automata to improve routing within RPL provides insightful information that could propel future developments in IoT networking toward more resilient, effective, and long-lasting network architectures. The effect sizes confirm the practical significance of these improvements for energy-constrained IoT deployments. Even modest efficiency gains translate to substantial operational benefits in battery-powered networks.