|
Progress - Overlay Networks - Search
Decentralized peer-to-peer (P2P) networks like Gnutella are increasingly
becoming attractive for certain applications because they require no centralized
directories and no precise control over network topology or data placement.
The greatest advantage is the robustness provided by them.
However, the efficient management of such distributed P2P networks, specially
developing an efficient search algorithm,
poses a fundamental challenge to researchers.
In this respect, our ongoing research work derives inspiration
from natural immune systems to improve various aspects of P2P search algorithm.
In order to improve the performance, we have a two-fold strategy (a) devising a
fast propagation scheme of the query message packets, and (b) arranging the peers to form clusters of peers which
host similar data so that the search operation can be consequently speeded up.
Our aim is to built up a semi-structured P2P network, with the same level of robustness provided by decentralized
P2P network; however offering better quality of search service.
To design faster propagation of message packets, we use concepts from the proliferation
and mutation mechanism operative in natural immune systems.
The intelligent rearrangement of peers helps us to build memory within the P2P system,
whereby, the search operation
becomes faster when it is repeated over several times. The idea is directly derived from the memory mechanism
displayed by adaptive immune systems.
Detailed experimental results illustrating the effectivity of the designed immune-inspired algorithms
which entirely depend upon local knowledge can be found in
[GCD04b,GCD04a,GD04a,GD04b].
However, we believe, mere experimental results confirming superiority of biologically
inspired algorithms over others is not enough. In order to `really' establish the superiority
of such algorithms, explaining their dynamics is absolutely necessary.
We need to have analytical explanations about why such local interactions lead to the desirable global
emergent behavior.
An initial theoretical explanation behind the superiority of immune system-based algorithms
can be found in [GBD05].
We are in the process of further refining and developing the search algorithm.
We will continue to update this page with future interesting results.
- GCD04b
-
Niloy Ganguly, Geoff Canright, and Andreas Deutsch.
Design of a robust search algorithm for p2p networks.
In Proceedings of the 11th International Conference on High
Performance Computing, Bangalore, India, December 2004.
[PDF],
[Bibtex].
- GCD04a
-
Niloy Ganguly, Geoff Canright, and Andreas Deutsch.
Design of an efficient search algorithm for p2p networks using
concepts from natural immune systems.
In Proceedings of the 8th International Conference on Parallel
Problem Solving from Nature (PPSN VIII), Birmingham, UK, September 2004.
[PDF],
[Bibtex].
- GD04a
-
Niloy Ganguly and Andreas Deutsch.
A cellular automata model for immune based search algorithm.
In Proceedings of the Sixth International Conference on Cellular
Automata for Research and Industry (ACRI 2004), Amsterdam, Netherlands,
October 2004.
[PDF],
[Bibtex].
- GD04b
-
Niloy Ganguly and Andreas Deutsch.
Developing efficient search algorithms for p2p networks using
proliferation and mutation.
In Proceedings of the International Conference on Artificial
Immune Systems, Catania, Italy, September 2004.
[PDF],
[Bibtex].
- GBD05
-
Niloy Ganguly, Lutz Brusch, and Andreas Deutsch.
Design and analysis of a bio-inspired search algorithm for
peer-to-peer networks.
In Ozalp Babaoglu, Márk Jelasity, Alberto Montresor, Christof
Fetzer, Stefano Leonardi, Aad van Moorsel, and Maarten van Steen, editors,
Self-Star Properties in Complex Information Systems, volume 3460 of
Lecture Notes in Computer Science. Springer-Verlag, 2005.
[PDF],
[Bibtex].
|