Abstract
Peer-to-peer networks have drawn their strength from their ability to operate functionally without the use of a central agent. In recent years the development of the structured peer-to-peer network has further increased the distributed nature of p2p systems. These networks take advantage of an underlying distributed data structure, a common one is the distributed hash table (DHT). These peers use this structure to act as equals in a network, sharing the same responsibilities of maintaining and contributing. But herein lays the problem, not all peers are equal in terms of resources and power. And with no central agent to monitor and balance load , the heterogeneous nature of peers can cause many distribution or bottleneck issues on the network and peer levels. This is due to the way in which addresses are allocated in these DHTs. Often this function is carried out by a consistent hashing function. These functions although powerful in their simplicity and effectiveness are the stem of a crucial flaw. This flaw causes the random nature in which addresses are assigned both when considering peer identification and allocating resource ownership. This work proposes a solution to mitigate the random nature of address assignment in DHTs, leveraging two methodologies called hierarchical DHTs and content based addressing. Combining these methods would enable peers to work in cooperative groups of like interested peers in order to dynamically share the load between group members. Group formation and utilization relies on the actual resources a peer willingly shares and is able to contribute rather than a function of the random hash employed by traditional DHT p2p structures.
Library of Congress Subject Headings
Hashing (Computer science); Data structures (Computer science); Peer-to-peer architecture (Computer networks)
Publication Date
2009
Document Type
Thesis
Department, Program, or Center
Computer Science (GCCIS)
Advisor
Kwon, Minseok
Recommended Citation
Srebro, Joseph, "Content-based addressing in hierarchical distributed hash tables" (2009). Thesis. Rochester Institute of Technology. Accessed from
https://repository.rit.edu/theses/466
Campus
RIT – Main Campus
Comments
Note: imported from RIT’s Digital Media Library running on DSpace to RIT Scholar Works. Physical copy available through RIT's The Wallace Library at: QA76.9.H36 S74 2009