Reminder seminar invite: The School of Electrical, Computer, and Telecommunications Engineering (SECTE) cordially invites you to our research seminar. Title: Network Aware Nodes: A Localization-Free Self-Organization Approach for Internet of Things Presenter: Professor Anura Jayasumana (Dept. of Electrical & Computer Engineering, Colorado State University, Ft. Collins, CO 80523 USA, engr.colostate.edu/~anura) Date/time: Thursday 27 November 2014, 13:30-15:00 Venue: 6-210 Abstract: The cost and size trends of motes, RFIDs and nano-devices point to future massive-scale networks deployed in 2D and complex 3D spaces to form an Internet of Things. Most techniques for self-organization, routing and tracking in such networks rely on distances and localization in the physical domain. However, obtaining geographic coordinates depend on analog measurements such as signal strength and time-delay that are unreliable, error prone and expensive. Techniques based on geographical coordinates do not scale well to 3D either. We present a novel localization-free coordinate system, the Topology Coordinates (TC), for these networks. TCs are derived from the hop distances from a node to a small subset of anchor nodes. We demonstrate techniques for obtaining Topology-Preserving Maps (TPMs) of 2-D and 3-D networks, in which even the geographic voids and shapes are preserved. This novel topological domain combines the advantages of geographical and virtual coordinate domains. TC based Geo-Logical Routing (GLR), for example, significantly outperforms existing logical and physical coordinate based routing schemes in both 2D and 3D networks. A Directional Virtual Coordinate (DVC) transformation is proposed to restore the directionality lost in VCS and to introduce the concept of angles between virtual directions. Ability to specify virtual cardinal directions and angles is a radical change from the traditional approaches. A novel self-learning algorithm is presented to provide network awareness to individual nodes, a step toward large-scale evolving sensor networks
Posted on: Wed, 26 Nov 2014 05:25:04 +0000