An expert-system-based structure for active queue management
Jin Zhuo Wu, Karim Djemame · 2004
In this paper, the concepts of intelligent control are used in the design of AQM algorithm. Intelligent control is commonly used as alternative way to solve those uncertain problems in human or creature like approaches. We believe that the network congestion control is a problem that with large scale and complicity, where no accurate and reliable model can be proposed. It is believed that artificial intelligent methods have the potential to apply on it and solve uncertainties. Although there already exist some sorts of controllers in the communication networks designed either from intuition or probabilistic and conventional control theories that do include intelligence for some extent, they are not constructed in the way as intelligent control. As a result, those share the benefits from the research contributions of artificial intelligent and intelligent control. From intelligent control approaches, an expert-system-based structure is proposed for network congestion control. Follows this structure, a novel AQM algorithm is introduced. Simulation experiments are designed to show that the algorithm that generates from expert-system-based structure has a better performance than conventional approaches.