Performance evaluation of fuzzy ant based routing method for connectionless networks

This paper introduces a novel algorithm, called FuzzyAntNet inspired by swarm intelligence and optimized by fuzzy systems. FuzzyAntNet is a new routing algorithm which is constructed by the communication model observed in ant colonies, and enhanced with fuzzy systems. Two special characteristics of this method are scalability to network changes and capability to recognize the best route from source to destination with low delay, traffic and high bandwidth. Using Ants (or agents) in this method cause to avoid congestion in data packet transmission because Ants walk through paths and intermediate routers and gather information about their delay, congestion and in return, they update the current delay of each visited link and consequently they bring up to date the routing probabilities table for every traversed router. We compare FuzzyAntNet with other routing algorithms such as AntNet and Destination Sequenced Distance Vector (DSDV).