David Kempe
Jon Kleinberg
Abstract
The dynamic behavior of a network in which information is changing continuously over time requires robust and efficient mechanisms for keeping nodes updated about new information. Gossip protocols are mechanisms for this task in which nodes communicate with one another according to some underlying deterministic or randomized algorithm, exchanging information in each communication step. In a variety of contexts, the use of randomization to propagate information has been found to provide better reliability and scalability than more regimented deterministic approaches.In many settings, such as a cluster of distributed computing hosts, new information is generated at individual nodes, and is most "interesting" to nodes that are nearby. Thus, we propose distance-based propagation bounds as a performance measure for gossip mechanisms: a node at distance d from the origin of a new piece of information should be able to learn about this information with a delay that grows slowly with d, and is independent of the size of the network.For nodes arranged with uniform density in Euclidean space, we present natural gossip mechanisms, called spatial gossip, that satisfy such a guarantee: new information is spread to nodes at distance d, with high probability, in O(log1 + ε d) time steps. Such a bound combines the desirable qualitative features of uniform gossip, in which information is spread with a delay that is logarithmic in the full network size, and deterministic flooding, in which information is spread with a delay that is linear in the distance and independent of the network size. Our mechanisms and their analysis resolve a conjecture of Demers et al. [1987].We further show an application of our gossip mechanisms to a basic resource location problem, in which nodes seek to rapidly learn the location of the nearest copy of a resource in a network. This problem, which is of considerable practical importance, can be solved by a very simple protocol using Spatial Gossip, whereas we can show that no protocol built on top of uniform gossip can inform nodes of their approximately nearest resource within poly-logarithmic time. The analysis relies on an additional useful property of spatial gossip, namely that information travels from its source to sinks along short paths not visiting points of the network far from the two nodes.
- Agrawal, D., Abbadi, A. E., and Steinke, R. 1997. Epidemic algorithms in replicated databases. In Proceedings of the 16th ACM Symposium on Principles of Database Systems. ACM, New York, pp. 161--172. Google Scholar
- Birman, K., Hayden, M., Ozkasap, O., Xiao, Z., Budiu, M., and Minsky, Y. 1999. Bimodal multicast. ACM Trans. Comput. Syst. 17, 41--88. Google Scholar
- Demers, A., Greene, D., Hauser, C., Irish, W., Larson, J., Shenker, S., Sturgis, H., Swinehart, D., and Terry, D. 1987. Epidemic algorithms for replicated database maintenance. In Proceedings of the 7th ACM Symposium on Operating Systems Principles. ACM, New York, 1--12. Google Scholar
- Estrin, D., Govindan, R., Heidemann, J., and Kumar, S. 1999. Next century challenges: Scalable coordination in sensor networks. In Proceedings of the 5th International Conference on Mobile Computing and Networking. 263--270. Google Scholar
- Göbel, F., Orestes Cerdeira, J., and Veldman, H. 1991. Label-connected graphs and the gossip problem. Disc. Math. 87, 29--40. Google Scholar
- Gupta, I., van Renesse, R., and Birman, K. 2001. Scalable fault-tolerant aggregation in large process groups. In Proceedings of the Conference on Dependable Systems and Networks. 433--442. Google Scholar
- Hedetniemi, S., Hedetniemi, S., and Liestman, A. 1988. A survey of gossiping and broadcasting in communication networks. Networks 18, 319--349.Google Scholar
- Heinzelman, W., Kulik, J., and Balakrishnan, H. 1999. Adaptive protocols for information dissemination in wireless sensor networks. In Proceedings of the 5th International Conference on Mobile Computing and Networking. Google Scholar
- Kahn, J., Katz, R., and Pister, K. 1999. Next century challenges: Mobile networking for 'smart dust'. In Proceedings of the 5th International Conference on Mobile Computing and Networking. 271--278. Google Scholar
- Karp, R., Schindelhauer, C., Shenker, S., and Vöcking, B. 2000. Randomized rumor spreading. In Proceedings of the 41st IEEE Symposium on Foundations of Computer Science. IEEE Computer Science Press, Los Alamitos, Calif., 565--574. Google Scholar
- Kempe, D., and Kleinberg, J. 2002. Protocols and impossibility results for gossip-based communication mechanisms. In Proceedings of the 43rd IEEE Symposium on Foundations of Computer Science. IEEE Computer Science Press, Los Alamitos, Calif., 471--480. Google Scholar
- Kempe, D., Kleinberg, J., and Kumar, A. 2002. Connectivity and inference problems for temporal networks. J. Comput. Syst. Sci. 64, 820--842. Google Scholar
- Lin, M., Marzullo, K., and Masini, S. 1999. Gossip versus deterministic flooding: Low message overhead and high reliability for broadcasting on small networks. Tech. Rep. CS99-0637, UC San Diego. Google Scholar
- Pittel, B. 1987. On spreading a rumor. SIAM J. Applied Math. 47, 213--223. Google Scholar
- van Renesse, R. 2000. Scalable and secure resource location. In Proceedings of the 33rd Hawaii International Conference on System Sciences. Google Scholar
- van Renesse, R., Minsky, Y., and Hayden, M. 1998. A gossip-style failure-detection service. In Proceedings of the IFIP International Conference on Distributed Systems Platforms and Open Distributed Processing. 55--70. Google Scholar
Index Terms
- Spatial gossip and resource location protocols
Recommendations
Spatial gossip and resource location protocols
STOC '01: Proceedings of the thirty-third annual ACM symposium on Theory of computingThe dynamic behavior of a network in which information is changing continuously over time requires robust and efficient mechanisms for keeping nodes updated about new information. Gossip protocols are mechanisms for this task in which nodes communicate ...
Sub-linear universal spatial gossip protocols
SIROCCO'09: Proceedings of the 16th international conference on Structural Information and Communication ComplexityGossip protocols are communication protocols in which, periodically, every node of a network exchanges information with some other node chosen according to some (randomized) strategy. These protocols have recently found various types of applications for ...
On the complexity of asynchronous gossip
PODC '08: Proceedings of the twenty-seventh ACM symposium on Principles of distributed computingIn this paper, we study the complexity of gossip in an asynchronous, message-passing fault-prone distributed system. In short, we show that an adaptive adversary can significantly hamper the spreading of a rumor, while an oblivious adversary cannot. ...
Reviews
Anthony Joseph Duben
Gossip and resource location protocols are closely related. Gossip protocols are mechanisms by which nodes are updated, and by which they share new information about their environment as it changes. Locating resources can be considered a special case, in which nodes can locate a resource, such as a server. These protocols apply in a dynamic environment. As information changes (namely, sensor readings) or as resources go online or off-line, the changes need to be updated across the entire network. This paper discusses the preference for randomized approaches over deterministic methods of propagation, and the use of distance-based bounds as a performance measure. It also presents a demonstration of the finding that new information is spread to nodes at distance d with a speed that is logarithmic to the distance. This last point is the major goal of the paper, and it resolves a conjecture by Demers et al. [1]. The paper is divided into three major portions: an introduction, presenting an overview of the problem, and elaborating on the definitions and behavior of the protocols; a section on spatial gossip protocols; and a discussion of monotone and nonmonotone resource location protocols. Monotone protocols never lose a resource. Nonmonotone protocols recognize the possibility that a resource may be lost, and that this information needs to be made known to the network. The algorithms and protocols are described verbally, without portions of code. The paper is almost entirely formal, consisting of theorems and proofs leading up the resolution of the conjecture. Online Computing Reviews Service
Access critical reviews of Computing literature here
Become a reviewer for Computing Reviews.
Comments