what-when-how
In Depth Tutorials and Information
10
3
39 BT Torrents with
more than 5000 Peers
10
2
10
1
141 BT Torrents
with less than 300 Peers
10
0
10
0
10
1
10
2
10
3
AS Rank
Figure6.28
DistributionofASclustersize.
studied both the real traces and simulation results. hey evaluate the beneit of
several architectures and present the concept of locality in a particular solution.
Another work from H. Xie et al. [34] enables cooperation between peer-to-peer
applications and ISPs by a brand new locality architecture—P4P. According to the
large scale tests, P4P can reduce both the external traffic and the average down-
loading time. On the other hand, D. R. Choffnes et al. [12] proposed Ono—a
BitTorrent extension that leverages on a CDN infrastructure. his approach is
aimed to find the location of peers and group peers that are close to each other.
R. Bindal et al. [9] also examined a novel approach to enhance BitTorrent traffic
locality: biased neighbor selection. Using this method, a peer chooses the major-
ity, but not all, of its neighbors from peers within the same ISP. he simulation
results show that this modification can greatly reduce the inter-ISP traffic of BT
networks. S. L. Blond et al. [10] performed an extensive experiment-based study on
a controlled environment. he authors show that high locality values (deined by
[20]) enable up to two order of magnitude saving on inter-ISP links without any
significant impact on peers' download completion time.
However, most of these pervious studies are exclusively focused on a global infra-
structure. he content and the peer diversities are seldom discussed. Consequently,
most BT based locality approaches are processed upon every single peer in the
