Distributed Consensus and Mitigating Risk Propagation in Evolutionary Optimized Networks
The consensus model has been extensively applied to solve coordination problems of distributed systems. A consensus algorithm specifies the information exchange rule between a system (node) and all of its neighbors on the network. The design issue of sensor networks capable of reaching a consensus quickly has received considerable attentions. The conditions for achieving a consensus depend on the properties of the graph of the nodes. The network is also subjected to external shocks both with respect to the size of the shock and the spatial impact of the shock. There is empirical evidence that as the connectivity of a network increases, we observe an increase in the performance, but at the same time, an increase in the chance of risk propagation which is extremely large. In this paper, we address the issue of designing desirable networks for any fixed number of nodes and links by employing an evolutionary optimization procedure. The resulting networks are optimal not only for fast consensus but also for minimizing risk propagation.
KeywordsNetwork Topology Adjacency Matrix Random Network Consensus Problem Consensus Algorithm
Unable to display preview. Download preview PDF.
- 1.Albert-László, B.: Linked: The New Science of Networks. Perseus Books Group, New York (2002)Google Scholar
- 3.Durrett, R.: Random Graph Dynamics. Cambridge University Press, Cambridge U.K. (2006)Google Scholar
- 7.Donetti, L., Neri, F., Muñoz, M.A.: Optimal network topologies: expanders, cages, ramanujan graphs, entangled networks and all that. J. Stat. Mech. 26, 134–144 (2006)Google Scholar
- 10.Boyd, S., Diaconis, F., Xiao, L.: Fastest mixing markov chain on graph. SIAM Review, 121–155 (2004)Google Scholar
- 12.Kar, S., Aldosari, S., Moura, J.M.F.: Consensus based detection in sensor networks. In: International Workshop on Information Theory in Sensor Networks (June 2007)Google Scholar
- 17.Chakrabarti, D., Wang, Y., Wang, C., Leskovec, J., Faloutsos, C.: Epidemic threshold in real networks. ACM Trans. on Information and System Security 10(4), 131–156 (2006)Google Scholar
- 18.Sato, H., Isao, O., Shigenobu, K.: A new generation alternation model of genetic algorithms and its assessment. Journal of Japanese Society for Artificial Intellligence 12(5), 734–744 (1997)Google Scholar
- 20.Komatsu, T., Namatame, A.: An evolutionary optimal network design to mitigate risk contagion. In: 2011 Seventh International Conference on Natunal Computation (ICNC), pp. 1954–1959. IEEE (2011)Google Scholar
- 21.Komatsu, T., Namatame, A.: Optimal diffusion in evolutionary designed networks. Proceedings of Econophysics Colloquium 2010 (November 2010)Google Scholar