NetAI-Gym: Customized Environment for Network to Evaluate Agent Algorithm using Reinforcement Learning in Open-AI Gym Platform

The growing size of the network imposes computational overhead during route establishment using conventional approaches routing protocol. alternate approach in contrast to table updating mechanism is rule-based method, but this also provides a limited scope dynamic networks. Therefore, reinforcement learning promises better way finding route, it requires an evaluation platform build model synchronization between and agent. Unfortunately, de-facto for agent evaluation, namely Open-AI Gym, does not provide suitable networking environment. paper aims propose environment as novel contribution by designing customized synchronically with Gym. successful deployment proposed environment: NetAI-Gym functional practical result that can be used further develop mechanisms based on Q-learning. validation carried out different nodes regarding Episodes Vs. Reward. experimental outcome justifies validity solving network-related problems.