Gossip protocol is one of the most scalable and reliable protocol for message dissemination in a large network where nodes joins and leaves the network frequently. In this paper we have implemented gossip protocol for message/ information dissemination using two style of epidemic information dissemination i.e. rumor mongering and antientropy. The membership is managed by using CYCLON membership protocol. Here we analyze the change in behavior of four performance metrics of gossip protocol which are: Average Path Length, Average Clustering Coefficient; and Indegree and average message exchange with respect to different number of nodes.

 

Published In : IJCSN Journal Volume 3, Issue 3

Date of Publication : 01 June 2014

Pages : 50 - 56

Figures : 05

Tables : 04

Publication Link : Gossip Based Information Dissemination in a Distributed System

 

 

 

Mr. Navaraj Chettri : is a student. He has achieved his Master of Computer Application (MCA) from Sikkim Manipal Institute of Technology, Sikkim India and currently pursuing M.Tech degree in CSE from Sikkim Manipal Institute of Technology, Sikkim India. He has recently published one paper“, in the proceeding of IRF International Conference, ISBN: 978-93-82702-58-0, pp 51-56. and his area of interest is networking and his present interest is on wireless Sensor Network routing protocol i.e. “gossip protocol”

Mr. Dhruba Ningombam : is a research scholar. He has completed his M.Tech from Sikkim Manipal Institute of Technology, Sikkim India, and B.Tech from Government College of Technology, Coimbatore, India. He has published several papers in international conferences and journals. He is also a member of IAENG International Association of Engineers. His area of interest includes Artificial Intelligence, Cloud Robotics Communication and platform.

 

 

 

 

 

 

 

Membership

Gossip Protocol

Average cluster coefficient

Average path length

Rumor Mongering

Anti Entropy

In this paper we have implemented gossip protocol for message/ information dissemination using two style of epidemic information dissemination i.e. rumor mongering and anti-entropy. The membership is managed by using CYCLON membership protocol. The results of our implementation mentioned above, states that the performance of the protocol increases as the number of network size increases as expected. The results of the performance metrics are relatively low as mentioned in literature for better performance of information dissemination. Hence, we can conclude from our simulation that gossip protocol is one of the best tools for information dissemination in large and dynamic distributed system. The future direction for this work is that the delay factor in the message exchange could be reduced by the use of other epidemics information dissemination strategies (like rumor feedback, counter). The membership maintenances strategies could be changed like instead of using CYCLON Membership protocol HyPar View Membership Protocol could be used to check the performance of the protocol. The continue pursuit of this paper will include this entire factor.

 

 

 

 

 

 

 

 

 

[1] A. Demers et al., “Epidemic algorithms for replicated database maintenance,” in Proc. of the 6th ACM Symp. On Principles of Distributed Computing (PODC’87). Vancouver, BC, Canada: ACM Press, Aug. 1987, pp. 1– 12

[2] Birman, K. P., Hayden, M., Ozkasap, O., Xiao, Z., Budiu, M., and Minsky, Y Bimodal multicast. ACM Trans. Comp. Syst., 17(2):41–88,1999

[3] Kermarrec, A.-M., Massouli´e, L., and Ganesh, A. J. (2003). Probabilistic Reliable Dissemination in Large- Scale Systems. IEEE Trans. Par. Distr. Syst., 14(2):248– 258.

[4] Eugster, P. T., Guerraoui, R., Kermarrec, A.-M., and Massouli´e, L. (2004). Epidemic information dissemination in distributed systems. IEEE Computer, 37(5):60–67.

[5] Spyros Voulgaris, Daniela Gavidia, and Maarten van Steen, “CYCLON: Inexpensive Membership Management for Unstructured P2P Overlays”, June 2005, Journal of Network and Systems Management, Vol. 1

[6] Ming Cao, Daniel A. Spielman, Edmund M. Yeh. “Accelerated gossip algorithms for distributed computation”. In Proc. 44th Annu. Allerton Conf. Commun. Control Comput., Monticello, IL, Sep. 2006.

[7] Stephen Boyd, Arpita Ghosh, Balaji Prabhakar, Devavrat Shah. “Analysis and optimization of randomized gossip algorithms”. Decision and Control, 2004. CDC. 43rd IEEE Conference on, volume 5, pages 5310 - 5315.

[8] Stephen Boyd, Arpita Ghosh, Balaji Prabhakar, Devavrat Shah. “Gossip algorithms: design, analysis and applications”. INFOCOM 2005. 24th Annual Joint Conference of the IEEE Computer and Communications Societies, Proceedings IEEE. Volume 3, pages 1653 - 1664.

[9] Paolo Frasca, Fabio Fagnani. “Broadcast gossip averaging algorithms: interference and asymptotical error in large networks”. arXiv preprint arXiv:1005.1292, 2010.

[10] Damon Mosk-Aoyama, Devavrat Shah. Fast distributed algorithms for computing separable functions. Information Theory, IEEE Transactions on, july 2008. Volume 54, issue 7, pages 2997 - 3007.

[11] Robbert van Renesse, Yawn Minsky, Mark Hayden. “ A gossip-style failure detection service”. Middleware'98, 1998, pages 55 - 70.

[12] Zygmunt J. Haas, Joseph Y. Halpern, Li Li, “ Gossip- Based ad hoc routing”. INFOCOM 2002. Twenty-First Annual Joint Conferences of the Bibliography 85 IEEE Computer and Communications Societies. Proceedings. IEEE. Volume 3, pages 1707 - 1716.

[13] Ali Ghodsi, Sameh El-Ansary, Supriya Krishnamurthy, Seif Haridi, “ A self-stabilizing network size estimation gossip algorithm for peer-to-peer systems”. Technical Report T2005:16, SICS (2005).

[14] Mauro Franceschelli, Alessandro Giua, Carla Seatzu, “ Load balancing on networks with gossip-based distributed algorithms”. Decision and Control, 2007 46th IEEE Conference on, pages 500 - 505.

[15] Mauro Franceschelli, Alessandro Giua, Carla Seatzu. Load balancing over heterogeneous networks with gossip-based algorithms. American Control Conference, 2009. ACC '09, pages 1987 - 1993.

[16] Eytan Modiano, Devavrat Shah, Gil Zussman, “ Maximizing throughput in wireless networks via gossiping”. SIGMETRICS '06/Performance '06 Proceedings of the joint international conference on Measurement and modeling of computer systems, pages 27 - 38.

[17] Alexandros G. Dimakis, Soummya Kar, Jos_e M. F. Moura, Michael G. Rabbat, Anna Scaglione. “Gossip algorithms for distributed signal processing”. Proceedings of the IEEE, november 2010, volume 98, issue 11, pages 1847 - 1864.

[18] Konrad Iwanicki, Maarten van Steen. “The PL-Gossip algorithm”. Technical Report IR-CS-034.

[19] http://peersim.sourceforge.net