• 沒有找到結果。

Using strategies plus bandwidth limiting

在文檔中 中 華 大 學 (頁 49-56)

CHAPTER 5 Performance Evaluation

5.2. Performance Analysis

5.2.3. Using strategies plus bandwidth limiting

because we have DPA to improve the system performance. The file download time of PCS is increased only slightly (about 5%), and it has a fewer effect upon the users.

But BNS is increased by about 13%.

0 1000 2000 3000 4000 5000 6000 7000 8000 9000 10000

No limiting 2.5Mbps 1.5Mbps 500Kbps ISP bandwidth limiting

Time (seconds)

BNS BNS + UPS BNS + UPS + DQA BNS + UPS + DPA (PCS)

Fig. 5-6. Download time of combining bandwidth limiting with 50 extra peers

CHAPTER 6 Conclusions and Future Work

We have presented Peer Collaboration Strategy (PCS), which integrates the advantage of BNS (decrease redundancy) and DQA (decrease the file download time) into our strategy, and combines with UPS to let redundancy down. PCS can ensure redundancy is controlled, besides the status of ISP lost pieces. Performance results show that our PCS can reduce cross-ISP traffic without sacrificing system performance, and we have to modify original framework of BitTorrent, and it does not need additional equipment and backup mechanism. In terms of the user perspective, they only care about the file download time and if the seed does not exist. Therefore, a lot of previous approaches only focus on the file download time and most of the ideas are based on personal demand, and they do not consider cross-ISP traffic. However, the reality of truth, a large number of cross-ISP traffic will let ISPs adopt bandwidth limiting, and it brings the low download rate and redundancy out of control. In this thesis, we consider the whole performance of BitTorrent in terms of the ISPs perspective, and we focus on the reduce cross-ISP traffic and the next is the whole performance of BitTorrent. Improving the BitTorrent peers into clusters and we conform to the principle of economic. Performance results show that we adopt PCS is based on BNS and DQA to solve cross-ISP traffic, not only restrain the redundancy but users can obtain a great performance, and decrease the file download time significantly. We have concluded from this thesis that reforming the performance of BitTorrent should be considering the whole ISP first. Instead of only improving the download rate, we also need to consider the cross-ISP traffic. Thus, ISPs can reduce cross-ISP traffic without sacrificing system performance, and let ISPs and users create

a win-win situation. We expected to accommodate the more complex network to our simulator in the future, and make the more complete analysis and consider the factors of influence, which causes the cross-ISP traffic.

Actually, we cannot only rely on decrease redundancy to reduce the cost of equipment in the reality. Because of the lowest redundancy indicates reduce redundancy of only one file. However, a large number of files were shared by P2P systems. We cannot avoid the P2P traffic has been increasing gradually. Thus, ISPs still need to economize on cost of extended equipment by bandwidth limiting. But, the cross-ISP traffic may get out of control if there is no strategy, which reforms the P2P system (like PCS). We cannot only use PCS or bandwidth limiting to reduce cross-ISP traffic. We thought an approach combines PCS with bandwidth limiting can reduce cross-ISP traffic and conform to the principle of economic.

Now, there are many proposals intend to integrate CDN into P2P system or developing the system such as P4P, even letting P2P system merge with monitor system. They promote QoS through this equipment to manage the files and users, and monitor the traffic. Most importantly, they also the advantages of P2P are preserved, such as scalability, expansibility and arbitrariness. It depends on the consideration of ISPs between the cost and the service. On the other hand, desire of collaboration between ISPs and users, and the incentive of users adopt the new P2P system should be considering. Problems of copyright, privacy and price also should be considering when using P4P or CDN.

References

1. http://www.ipoque.com/resources/internet-studies/internet-study-2008_2009.

2. BitTorrent, http://www.bittorrent.com/.

3. CacheLogic, http://www.cachelogic.com/.

4. eMule, http://www.emule.org/.

5. Gnutella, http://www.gnutelliums.com/.

6. Napster, http://www.napster.com/.

7. V. Aggarwal, A. Feldmann and C. Scheideler, Can ISPs and P2P Users Cooperate for Improved Performance, In Proceedings of ACM SIGCOMM Computer Communication Review, Volume 37, Issue 3, Pages 29-40, July 2007.

8. V. Aggarwal, O. Akonjang and A. Feldmann, Improving User and ISP Experience through ISP-aided P2P Locality, In Proceedings of IEEE Conference on Computer Communications Workshops, INFOCOM 2008, Pages 1-6, 13-18 April 2008.

9. A. R. Bharambe, C. Herley and V. N. Padmanabhan, Analyzing and improving a BitTorrent network’s performance mechanisms, In Proceedings of the 25th IEEE International Conference on Computer Communication, INFOCOM 2006, Pages 1-12, April 2006.

10. R. Bindal, P. Cao, W. Chan, J. Medved, G. Suwala, T. Bates and A. Zhang, Improving traffic locality in BitTorrent via biased neighbor selection, In Proceedings of the 26th IEEE International Conference on Distributed Computing Systems (ICDCS), Pages 66-66, July 2006.

11. M. D. Burger and T. Kielmann, MOB: zero-configuration high-throughputmulticasting for grid applications, In Proceedings of the 16th International Symposium on High performance distributed computing, Pages 159-168, Monterey, California, USA, June 25-29, 2007.

12. D. R. Choffnes and F. E. Bustamante, Taming the torrent: a practical approach to reducing cross-isp traffic in peer-to-peer systems, In Proceedings of the ACM SIGCOMM 2008 conference on Data communication, Pages 363-374, Seattle, WA, USA, 2008.

13. B. Cohen, Incentives Build Robustness in BitTorrent, In Proceedings of First Workshop on Economics of Peer-to-Peer Systems, 2003.

14. M. Feldman, C. Papadimitriou, J. Chuang and I. Stoica, Free-riding and whitewashing in Peer-to-Peer systems, In Proceedings of the ACM SIGCOMM workshop on Practice and theory of incentives in networked systems, Pages 228-236, Portland, Oregon, USA, 2004.

15. C. Gkantsidis and P. R. Rodriguez, Network coding for large scale content distribution, In Proceedings of the 24th Annual Joint Conference of the IEEE Computer and Communications Societies, INFOCOM 2005, Volume 4, Pages 2235-2245, 13-17 March 2005.

16. L. Guo, S. Chen, Z. Xiao, E. Tan, X. Ding, and X. Zhang, Measurements, Analysis, and Modeling of BitTorrent-like Systems, In Proceedings of the 5th ACM SIGCOMM conference on Internet Measurement, Pages 4-4, Berkeley, CA, 2005.

17. K. Huang, L. Wang, D. Zhang and Y. Liu, Optimizing the BitTorrent performance using an adaptive peer selection strategy, Future Generation Computer Systems,

18. M. Izal, G. Urvoy-Keller, E.W. Biersack, P. Felber, A. Hamra, and L.

Garces-Erice, Dissecting bittorrent: Five months in a Torrent’s lifetime, In Proceedings of the 5th Passive and Active Measurements Workshop, France, Pages 1-11, 2004.

19. R. Izmailov, S. Ganguly, and N. Tu, Fast Parallel File Replication in Data Grid, In Proceedings of Future of Grid Data Environments workshop (GGF-10), Berlin, Germany, March 2004.

20. T. Karagiannis, P. Rodriguez, and K. Papagiannaki, Should internet service providers fear peer-assisted content distribution, In Proceedings of the 5th ACM SIGCOMM conference on Internet Measurement, Pages 6-6, Berkeley, CA, 2005.

21. A. Legout, G. Urvoy-Keller and P. Michiardi, Rarest first and choke algorithms are enough, In Proceedings of the 6th ACM SIGCOMM conference on Internet Measurement, Pages 203-216, Brazil, 2006.

22. Minglu Li, Jiadi Yu, and Jie Wu, Free-Riding on BitTorrent-Like Peer-to-Peer File Sharing System: Modeling Analysis and Improvement, IEEE Transaction on Parallel and Distributed System, Volume 19, Issue 7, Pages 954-966, July 2008.

23. A. Su, D. R. Choffnes, A. Kuzmanovic, and F. E. Bustamante, Drafting Behind Akamai: Travelocity-based detouring, In Proceedings of the ACM SIGCOMM 2006 conference on Application, technologies, architectures, and protocols for computer communications, Pages 435-446, Pisa, Italy, 2006.

24. Y. Tian, D. Wu and K. W. Ng, Modeling, analysis and improvement for BitTorrent-like file sharing networks, In Proceedings of the 25th IEEE International Conference on Computer Communications, INFOCOM 2006, Pages 1-11, Barcelona, Spain, April 2006.

在文檔中 中 華 大 學 (頁 49-56)

相關文件