Information Technology Reference
In-Depth Information
1
P2P
P4P
Improved−P4P
0.95
0.9
0.85
0.8
0.75
0.7
0.65
400
450
500
550
600
650
700
750
800
Swarm Size
Fig. 4. Average Completion Time
time, and they all reduce the average completion time by approximately 20%
compared with the native BitTorrent.
In other words, in intradomain settings, Improved-P4P can reduce the bot-
tleneck link utilization and therefore relieve the pressure on networks with bot-
tlenecks without degrading the performance of the P2P applications.
5Con lu on
We analyzed the P-distance in a P4P cooperative model theoretically and showed
the relationship between the P-distance and the link trac. We concluded that
for a network topology with bottlenecks, the application of MLU as the optimiza-
tion objective of the P4P cooperative model does not achieve good performance.
Therefore, we proposed an improved model, Improved-P4P , which introduced a
link utility function as its objective. We analyzed the relationship between the
link price and the trac in this model, and demonstrated that Improved-P4P
could make the trac in a network more homogeneous. We carried out a consid-
erable number of simulations on P2P, P4P, and Improved-P4P in order to verify
that the Improved-P4P could implement the inter-active control of ISP and P2P
with respect to the network trac and that it could benefit ISPs with a reduc-
tion of the bottleneck trac without compromising the performance of the P2P
applications. Further, Improved-P4P performed stably for various swarm sizes,
proving its feasibility, scalability, and effectiveness. In conclusion, Improved-P4P
could solve the cooperation problem of P2P and ISP eciently.
Acknowledgement. This research is supported by the National Natural Sci-
ence Foundation of China (No.60873253 ), and National Key Basic Research
Program (”973” Program) of China (No.2009CB320500 and 2009CB320501).
 
Search WWH ::




Custom Search