DocumentCode :
3144871
Title :
SPUN: A P2P Probabilistic Search Algorithm Based on Successful Paths in Unstructured Networks
Author :
Himali, D. M Rasanjalee ; Prasad, Sushil K.
Author_Institution :
Dept. of Comput. Sci., Georgia State Univ., Atlanta, GA, USA
fYear :
2011
fDate :
16-20 May 2011
Firstpage :
1610
Lastpage :
1617
Abstract :
Efficient searching for information is an important goal in peer-to-peer (P2P) networks. Searching in an unstructured P2P network is particularly challenging due to the random nature of the P2P overlay links. In this paper, we propose a novel probabilistic search mechanism called SPUN, that increases the success ratio of queries while keeping the bandwidth consumption considerably low. SPUN is an informed search mechanism that improves upon state-of-art probabilistic mechanism, namely, the Adaptive Probabilistic Search (APS). The core principle of our algorithm is to exploit the successful query paths that develop during the lifetime of P2P network and converge toward the target objects. Our work introduces a new neighbor selection criterion which allows a peer to evaluate its neighbors based on the strength of successful paths the neighbor leads to. We also introduce a peer profile exchange mechanism that supports the reduction in uncertainty in peer selection decision. Our extensive simulation results confirm that our path-based algorithm performs 25% better than APS and several of its variants and is capable of achieving higher success ratios with fewer walkers each with an average message size of 71 bytes.
Keywords :
peer-to-peer computing; probability; search problems; P2P overlay links; P2P probabilistic search algorithm; SPUN; adaptive probabilistic search; neighbor selection criterion; peer profile exchange mechanism; peer selection decision; peer-to-peer networks; unstructured P2P network; Heuristic algorithms; History; Indexes; Peer to peer computing; Probabilistic logic; Search problems;
fLanguage :
English
Publisher :
ieee
Conference_Titel :
Parallel and Distributed Processing Workshops and Phd Forum (IPDPSW), 2011 IEEE International Symposium on
Conference_Location :
Shanghai
ISSN :
1530-2075
Print_ISBN :
978-1-61284-425-1
Electronic_ISBN :
1530-2075
Type :
conf
DOI :
10.1109/IPDPS.2011.316
Filename :
6008960
Link To Document :
بازگشت