Peer-to-Peer Overlay based on Swarm Intelligence

Context

Distributed Computing, intelligent algorithms

Task Description

The task is to create an overlay network for the lookup of data /containers in the Internet. The idea is to search by using some heuristic algorithm like algorithms based on Ant Colony Optimization. In this scenario it is assumed that the search space is the set (subset of the set) of names, the environment is static (concerning the number of nodes, we assume the finite number of nodes), but dynamic (concerning the connection between nodes), the role of a "neighbour" in the Ant Algorithm has the next lookup container and a "distance" between nodes can be interpreted either as some constant C or the assigned value according to the fact how much the node is bussy. The proposed algorithms that can be applied are MIN-MAX AntSystem and AntNet. These algorithms are going to be applied in a hybrid forms i.e. combined with Local Search algorithm (in order to achieve the best possible solution that is time consuming as less as possible). The work includes implementation of the Peer-to-Peer Overlay structure and benchmarking.

Requirements

Java

Student

open topic (please apply using ID = "P2P-Swarm")

Supervisor

A.o. Univ.-Prof. Dr. Dipl.-Ing. eva Kühn

Related Projects in the Network

• Vesna Sesum-Cavik and eva Kühn, A Swarm Intelligence Appliance to the Construction of an Intelligent Peer-to-Peer Overlay Network, 1st Workshop on Coordination in Complex Software Intensive Systems (COCOSS-2010), Co-located with CISIS 2010: 4th International Conference on Complex, Intelligent and Software Intensive Systems [link]
• Vesna Sesum-Cavic, eva Kühn, Peer-to-Peer Overlay Network based on Swarm Intelligence, The 10th Annual International Workshop "Engineering Societies in the Agents' World" (ESAW 2009), Utrecht University, The Netherlands, Nov. 18-20, 2009 [link]