Use this URL to cite or link to this record in EThOS: http://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.698949
Title: Enhanced resource discovery mechanisms for unstructured peer-to-peer network environments
Author: Jamal, Azrul Amri bin
ISNI:       0000 0004 5993 5396
Awarding Body: Prifysgol Bangor University
Current Institution: Bangor University
Date of Award: 2016
Availability of Full Text:
Access through EThOS:
Access through Institution:
Abstract:
This study explores novel methods for resource discovery in unstructured peerto-peer (P2P) networks. The objective of this study is to develop a lightweight resource discovery mechanism suitable to be used in unstructured P2P networks. Resource discovery techniques are examined and implemented in a simulator with high scalability in order to imitate real-life P2P environments. Simulated topology generator models are reviewed and compared, the most suitable topology generator model is then chosen to test the novel resource discovery techniques. Resource discovery techniques in unstructured P2P networks usually rely on forwarding as many query messages as possible onto the network. Even though this approach was able to return many resources, the flooding of the network with query messages have an adverse effect on the network. Flooding the network has undesirable consequences such as degenerative performance of the network, waste of network resources, and network downtime. This study has developed alpha multipliers, a method of controlling query message forwarding to deal with the flooding effect of most resource discovery techniques in unstructured P2P networks. The combination of alpha multipliers and breadth-first search (BFS), ↵-BFS, was able to avoid the flooding effect that usually occurs with BFS. The ↵-BFS technique also increases the combined query efficiency compared to the original BFS. Aside from improving a uninformed search technique such as the BFS, this study also examines the network communication cost of several informed resource discovery techniques. Several issues that arise in informed resource discovery techniques, such as false positive errors, and high network communication costs for queries to update search results are discussed. This detailed analysis forms the basis of a lightweight resource discovery mechanism (LBRDM) that reduces the network communication cost by reducing the number of backward updates inside the network when utilising the blackboard resource discovery mechanism (BRDM). Simulations of BRDM and LBRDM show that the lightweight version can also return an almost identical combined query efficiency than the BRDM. The solution to control query message forwarding in ↵-BFS, and the removal of unnecessary exchange of information in LBRDM open a new perspective on simplifying resource discovery techniques. These approaches can be implemented on other techniques to improve the performance of resource discovery.
Supervisor: Not available Sponsor: Not available
Qualification Name: Thesis (Ph.D.) Qualification Level: Doctoral
EThOS ID: uk.bl.ethos.698949  DOI: Not available
Share: