Design of an Information Brokerage System for Wireless Sensor Networks Cristina Muñoz, Pierre Leone Department of Computer Science, University of Geneva, Switzerland Motivation Nowadays, sensor networks are used for a variety of purposes. Actual sensor networks include from simple Research field Key research findings Distributed event-based systems Wireless Sensor Networks Network protocols Permutations on the top of well-connected networks are applications as data collection to smart sensing that used to simulate intersections for the Information Brokerage System. allows to react to certain events using actuators. Distributed sensing and control in unstructured networks Figure 7. Permutations arises a common issue due to the easy deployment of the sensing nodes. Most of these networks use information brokerage to match the producers of certain events with their consumers. Figure 3. Wireless Sensor Devices Objective To design an Information Brokerage System for Design of Directional Random Walks unstructured and free geo-coordinates wireless Directional Random Walks Information Brokerage System ↓ time to intersection compared to Pure RW Balances the load Synchronicity → ↓time to intersection Our design is as good as a traditional Rumor Routing algorithm with an infinite memory The use of N2(v) is not efficient The nº of nodes in the path until intersection is smaller for DRWs than for Pure RWs. ↑ densities of nodes → shorter paths sensor networks. 1 branch is as efficient as 2 branches Information Brokerage System A greedy approach of a geographic algorithm uses less nodes in the path but does not balance the load Figure 8. Key research findings. Future work • To design a content-based system wich uses DRWs. Figure 4. Directional Random Walks Figure 1. Information Brokerage System The cardinality of intersections is taken into account in An Information Brokerage System matches producers and consumers taking into account that they are completely decoupled. Our strategy assumes the principle that two lines in a plane are likely to intersect. • To deploy a real sensing testbed. order to add nodes to a branch: C(v) = α | N(v) ∩ N(DRW) | + β | N(v) ∩ N2(DRW) | where: Strategy • To define the filters to be implemented at broker nodes. • v is a candidate node to be added to the DRW. • DRW denote all the nodes already added to the DRW. • α and β are considered as weights. • N(v) is the set of neighbor nodes of v. • N2(v) is the set of neighbor nodes of N(v) . References • LEONE, P., AND MUÑOZ, C. 2013. Content based routing with directional random walk for failure tolerance and detection in cooperative large scale wireless networks. In Proc. 2nd Intl. Workshop ASCoMS’13, SAFECOMP’13. • LIN, C.-H., KUO, J.-J., LIU, B.-H., AND TSAI, M.-J. 2012. Gps-free, boundary-recognition-free, and reliable double-ruling based information brokerage scheme in wireless sensor networks. IEEE Transactions on Computers 61, 6, 885–898. • MUÑOZ, C., AND LEONE, P. 2014. Design of an unstructured and free geo-coordinates information brokerage system for sensor networks using directional random walks. In SENSORNETS14. Acknowledgment Figure 5. Adding nodes to a branch. Figure 2. Strategy proposed: a) Two lines cross; b) Two lines cross using Directional Random Walks. Contact Information Cristina Muñoz, Pierre Leone Directional Random Walk Theoretical Computer Science & Sensor Lab A DRW is a probabilistic method that uses a forwarding loop-free technique to reach distant areas in the network. Computer Science Dept. University of Geneva Carouge, Switzerland The path constructed uses collaborative nodes of a mesh network which form a list of relaying nodes for data propagation. Figure 6. DRWs of one branch intersecting Poster template by ResearchPosters.co.za This work has been developed as part of the POPWiN project that is financially supported by the Swiss Hasler Foundation in its SmartWorld program. Email: [email protected] [email protected]