AN INTELLIGENT FUZZY REINFORCEMENT LEARNING-BASED ROUTING ALGORITHM FOR LATENCY-ENERGY TRADE-OFF IN IOT-ENABLED SDN

Abstract

In this paper, we present an Intelligent Fuzzy Re-inforcement Learning Based Routing Technique Algorithm (IFRLTRA) for Router Optimization in IoT and SDN Environment. The approach integrates fuzzy reasoning, reinforcement learning (RL), and dynamic resource allocation in the latency-driven feedback loop to achieve low-latency, high-bandwidth-utilization, load-balanced network. In addition, the utilization of these approaches simultaneously provides IFRLTRA with efficient network operation to maintain real time applications like video conferencing and other IoT services under their desired QoS parameters. The proposed IFRLTRA was evaluated against state-of-the-art routing algorithms through extensive simulations, showing that it outperforms them in terms of the rate of admission, path length, e2el latency, traffic utilization and load. Due to its dynamic traffic rerouting based on current network conditions in real-time, this algorithm allocates resources efficiently, and it helps the system to reduce congestion and latency as well as energy consumption. The paper presents potential future enhancements: purchase of edge computing to reduce latency even more and improved resource management as well as security mechanisms for applications with high data protection requirements. Above all, the IFRLTRA is considered as a more robust and practical scheme for routing in SDN-based IoT environment and thereby can be an appropriate candidate for improving the quality of real-time application on the modern IoT network which benefits from ensuring reliability and enhanced performance of network traffic..