Dynamic Decentralized Packet Clustering in Networks

In this paper we study a dynamic version of the Decentralized Packet Clustering (DPC) Problem. For a network consisting of routers and application servers the DPC problem is to find a good clustering for packets that are sent between the servers through the network. The clustering is done according to a data vector in the packets. In the dynamic version of DPC the packets data vector can change. The proposed algorithms to solve the dynamic DPC are inspired by the odor recognition system of ants. We analyze the new algorithms for situations with different strengths of dynamic change and for different number of routers in the network.

[1]  Salima Hassas,et al.  Self-Organisation: Paradigms and Applications , 2003, Engineering Self-Organising Systems.

[2]  Nicolas Monmarché,et al.  Visual Clustering with Artificial Ants Colonies , 2003, KES.

[3]  Nicolas Monmarché,et al.  A new clustering algorithm based on the chemical recognition system of ants , 2002 .

[4]  Daniel Merkle,et al.  Decentralized packet clustering in router-based networks , 2005, Int. J. Found. Comput. Sci..

[5]  Julian F. Miller,et al.  Genetic and Evolutionary Computation — GECCO 2003 , 2003, Lecture Notes in Computer Science.

[6]  Marco Dorigo,et al.  On the Performance of Ant-based Clustering , 2003, HIS.

[7]  Nicolas Monmarché,et al.  AntClust: Ant Clustering and Web Usage Mining , 2003, GECCO.

[8]  Lakhmi C. Jain,et al.  Knowledge-Based Intelligent Information and Engineering Systems , 2004, Lecture Notes in Computer Science.

[9]  Jiong Yang.  Dynamic clustering of evolving streams with a single pass , 2003, Proceedings 19th International Conference on Data Engineering (Cat. No.03CH37405).

[10]  Marco Dorigo,et al.  Strategies for the Increased Robustness of Ant-Based Clustering , 2003, Engineering Self-Organising Systems.

[11]  Peter J. Rousseeuw,et al.  Finding Groups in Data: An Introduction to Cluster Analysis , 1991 .