Intelligent transport protocol design using genetic algorithm

Abstract

With the growth of Information Technology in all aspects of management, business and entertainment, there is an increasing demand for networking for very varied purposes: fast file transfer, video streaming and voice over Internet Protocol. Configurability, adaptability, intelligence and QoS orientation are the qualities of a successful future transport protocol. A constant process of analyzing networking through fitness levels allows feedback on QoS achievement and subsequent protocol engineering. We present an intelligent protocol, genetic algorithm transport protocol (GATP), that can evolve and adapt to the network environment and achieve an evolving user-configurable QoS. Using a constant process of negotiation, implementation and feedback, a QoS-focused networking between peers can be achieved. A review of experimental results will provide an interesting insight into genetic algorithm behavior in an actual network environment. © 2002 IEEE.