Search in complex network
Decentralized searching finds targeted nodes in network without asking help from server, but using local information only. (1) By examining and evaluating several degree‐based decentralized algorithms and related factors, this final year report shows how does the searching efficiency affected by...
Saved in:
| Main Author: | |
|---|---|
| Other Authors: | |
| Format: | Final Year Project |
| Language: | English |
| Published: |
2009
|
| Subjects: | |
| Online Access: | http://hdl.handle.net/10356/16702 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | Decentralized searching finds targeted nodes in network without asking
help from server, but using local information only. (1) By examining and
evaluating several degree‐based decentralized algorithms and related factors, this
final year report shows how does the searching efficiency affected by the nodal
degree, memory, searching mechanism and network topology. It is also found
that the benefit of using more memory can be achieved by using better searching
mechanism. With the careful design of the searching algorithm, it is possible to
improve the searching efficiency in terms of success rate and delivery time.
In this final year report, the effects of four chosen factors to the degreebased
decentralized algorithms had been examined in different chapters. The 4
factors are the effect of average nodal degree, effect of memory, effect of
backtracking and effect of clusters. The simulation results show that larger
average nodal degree, larger memory size and more backtracking steps and
smaller modularity in a network imply better searching efficiency. The mechanism
of those effects were revealed the in this report and an improvement for the
degree‐based decentralized algorithm was proposed. |
|---|