BATFE: design of a hybrid bioinspired model for adaptive traffic flow control in edge devices

Effective traffic flow management entails monitoring edge devices to ensure traffic is evenly dispersed across networks. However, existing flow control systems, sometimes incorporating machine learning, struggle with complicated configurations and inefficiencies, particularly in large-scale device networks. This study provides a hybrid bioinspired system to optimize traffic flow control in edge device networks and address these issues. The proposed methodology utilizes request-response time data to forecast traffic flows across multiple device sets. Using this predictive capacity, edge resources are dynamically assigned, considerably enhancing Quality of Service (QoS) in large-scale systems. The model analyzes this data using a hybrid Elephant Herding Particle Swarm Optimizer (EHPSO), which assigns temporal weights to IP groups to estimate future demands, permitting effective resource allocation depending on system capacity. A performance-based fitness function further modifies edge configurations to respond to incoming traffic. By using EHPSO, the suggested model achieves an 8.3% improvement in resource allocation efficiency, a 4.5% reduction in calculation time, and a 6.4% decrease in computational burden for processing huge numbers of requests, making it very useful for large-scale applications.