FPGA based implementation of Fuzzy controllers for internet traffic

Abstract

Recent research results propose and show the usefulness of a fuzzy control based approach to the development of intelligent systems for congestion control in Internet routers. However, adoption of this new technology is handicapped because of operational requirements, mostly in terms of inference speed, a hard constrain on the practical implementation of key traffic controlling systems. We report on the implementation of intelligent fuzzy controllers for Internet traffic using an FPGA based prototyping platform. A development methodology and a tool chain, a flexible and open prototyping platform, a set of fuzzy controllers and implementation results for a number of traffic controllers are presented. Our prototypes are shown to satisfy the requirements of high performance routing hardware deployed in the current Internet.