Abstract
A Peer-to-Peer (P2P) network can boost its performance if peers are provided with underlying network-layer routing topology. The task of inferring the network-layer routing topology and link performance from an end host to a set of other hosts is termed as network tomography, and it normally requires host computers to send probing messages. We design a passive network tomography method that does not require any probing messages and takes a free ride over data flows in P2P networks. It infers routing topology based on end-to-end delay correlation estimation (DCE) without requiring any synchronization or cooperation from the intermediate routers. We implement and test our method in the real world Internet environment and achieved the accuracy of 92 % in topology recovery. We also perform extensive simulation in OMNeT++ to evaluate its performance over large scale networks, showing that its topology recovery accuracy is about 95 % for large networks.
Similar content being viewed by others
Notes
Rooted from a single source node, the network-layer routing structure to multiple destination nodes is always a tree structure at any given time instance, or otherwise routing loops exist.
We argue that passive network tomography using RTT measurement is problematic in P2P networks for at least two reasons: first, message delivery between peers may not use TCP, and thus we cannot use TCP-SYN and TCP-ACK for passive RTT estimation as in [17]; second, even if passive RTT estimation is possible, the correlation estimation of the return paths is hard due to the variable processing delay at the destination nodes.
This implies that the sender have data to both receivers. This assumption is reasonable due to the diverse data forwarding requirements in P2P networks. Otherwise, passive network tomography will not have enough information to accurately infer routing topology.
For simplicity we usually choose the root node as the base router. However, in fact any intermediate node can be selected as the base router since the dynamic algorithm can go both directions in different cases within a tree.
A slice is a set of allocated resources distributed across PlanetLab. Slice is implemented using a technique called distributed virtualization. After nodes have been assigned to a slice, virtual servers for that slice are created on each of the assigned nodes.
Waxman refers to a generation model for a random topology using Waxman’s probability model for interconnecting the nodes.
References
Androutsellis-Theotokis S, Spinellis D (2004) A survey of peer-to-peer content distribution technologies. ACM Comput Surv (CSUR) 36:335–371
Qiu D, Srikant R (2004) Modeling and performance analysis of bittorrent-like peer-to-peer networks. ACM SIGCOMM Comput Commun Rev 34:337–378
Chandler H, Shen H, Zhao L, Stokes J, Li J (2012) Toward p2p-based multimedia sharing in user generated contents. IEEE Trans Parallel Distrib Syst 23(5):966–975
Andersen D, Balakrishnan H, Kaashoek F, Morris R (2001) Resilient overlay networks. In: Proceedings of SOSP, Banff, AB, Canada, pp 131–145
Yao B, Viswanathan R, Chang F, Waddington D (2003) Topology inference in the presence of anonymous routers. In: Proceedings of the IEEE INFOCOM, San Francisco, CA, pp 353–363
Vardi Y (1996) Network tomography: estimating source-destination traffic intensities from link data. J Am Stat Assoc 91(433):365–377
Gui J, Shah-Mansouri V, Wong V (2011) Accurate and efficient network tomography through network coding. IEEE Trans Veh Technol 60(6):2701–2713
Rabbat M, Nowak R, Coates M (2004) Multiple source, multiple destination network tomography. In: Proceedings of the IEEE INFOCOM, Piscataway, NJ, USA, pp 1628–1639
Caceres R, Duffield NG, Horowitz J, Towsley DF (1999) Multicast-based inference of network internal loss characteristics. IEEE Trans Inf Theory 45(7):2462–2480
Presti FL, Duffield NG, Horowitz J, Towsley D (2002) Multicast-based inference of network-internal delay distributions. IEEE/ACM Trans Netw 10(6):761–775
Nguyen H, Zheng R (2013) A binary independent component analysis approach to tree topology inference. IEEE Trans Signal Process 61(12):3071–3080
Cao J, Davis D, Wiel SV, Yu B, Vander S, Yu WB (2000) Time-varying network tomography: router link data. J Am Stat Assoc 95:1063–1075
Padmanabhan VN, Qiu L, Wang HJ (2002) Passive network tomography using bayesian inference. Microsoft Research:1–2
Ricciato F, Vacirca F, Fleischer W, Motz J, Rupp M (2006) Passive tomography of a 3g network: Challenges and opportunities. In: Proceedings of the IEEE INFOCOM
Cao J, Chen A, Lee P (2011) Modeling time correlation in passive network loss tomography. In: 41st International Conference on Dependable Systems Networks (DSN), 2011 IEEE/IFIP, pp 550–561
Yao H, Jaggi S, Chen M (2012) Passive network tomography for erroneous networks: A network coding approach. IEEE Trans Inf Theory 58(9):5922–5940
Tsang Y, Barford P, Nowak R (2004) Network radar: Tomography from round trip time measurements. In: Proceedings of the 4th ACM SIGCOMM conference on Internet measurement(IMC 04), pp 175–180
Tsang Y, Nowak RD (2003) Network delay tomography. IEEE Trans Signal Process 51(8):2125–2136
Eriksson BD, Barford P, Nowak R (2010) Toward the practical use of network tomography for internet topology discovery. In: Proceedings of IEEE INFOCOM
Ni J, Xie H, Yang T (2010) Efficient and dynamic routing topology inference from end-to-end measurements. IEEE/ACM Trans Netw:123–135
Source code of bittorrent, http://bittorrent.cvs.sourceforge.net/viewvc/bittorrent/, the homepage of BitTorrent source code
Planetlab. http://www.planet-lab.org/, the homepage of PlanetLab
Park K, Pai V (2005) Comon: A monitoring infrastructure for planetlab. http://comon.cs.princeton.edu/
Soltesz S, Pötzl H, Fiuczynski ME, Bavier A, Peterson L (2007) Container-based operating system virtualization: a scalable, high-performance alternative to hypervisors. In: Proceedings of EuroSYS
Bavier A, Bowman M, Chun B, Culler D, Karlin S, Muir S, Peterson L, Roscoe T, Spalink T, Wawrzoniak M (2004) Operating system support for planetary-scale network services. In: USENIX Symposium on Networked Systems Design and Implementation
Sommers J, Barford P (2007) An active measurement system for shared environments. In: Proceedings of the ACM IMC, pp 303–314
Omnet++. http://www.omnetpp.org/, the homepage of OMNet++
Medina A, Lakhina A, Matta I, Byers J (2001) Brite: An approach to universal topology generation. In: Proceedings of the MASCOTS
Castro R, Coates M, Liang G, Nowak R, Yu B (2004) Network tomography:recent developments. J Stat Sci 19(3):499–517
Duffield NG, Presti FL (2004) Network tomography from measured end-to-end delay covariance. IEEE/ACM Trans Netw 12(6):978–992
Duffield NG, Horowitz J, Presti FL (2001) Adaptive multicast topology inference. In: Proceedings of IEEE INFOCOM, pp 1636–1645
Acknowledgments
This work was supported by the National Key Technology Research and Development Program of the Ministry of Science and Technology of China under Grant no. 2012BAH93F01, the Innovation Research Fund of Huazhong University of Science and Technology, no. 2014TS095, the National Science Foundation of China under Grant no. 60803005, the National Science Foundation of China under Grant no. 91338201 and the Grand Foundational Research Program of China (863 Program) under Grand no.2015AA015701.
Author information
Authors and Affiliations
Corresponding author
Rights and permissions
About this article
Cite this article
Qin, P., Dai, B., Xu, G. et al. Taking a free ride for routing topology inference in peer-to-peer networks. Peer-to-Peer Netw. Appl. 9, 1047–1059 (2016). https://doi.org/10.1007/s12083-015-0380-9
Received:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s12083-015-0380-9