Journal of Computer Science and Technology/JCST
- Scalable localized routing in wireless sensor networks
- Simulation in ad hoc and sensor networks (criticism of existing practices)
Professor, Ivan Stojmenovic
EECE, University of Birmingham, UK
Scalable localized routing in wireless sensor networks
Sensors need their position information in order to provide useful monitoring information. In a localized routing algorithm, each node forwards the message solely based on the location of itself, its neighbors and destination. Such path based routing provides better fault tolerance than tree maintenance approaches. In this talk we describe a cost to progress ratio framework for designing routing algorithms, and show that a number of existing schemes are special cases of the design. The cost depends on metric selected, such as hop count, power, remaining energy, delay, expected hop count (which considers realistic physical layer), etc. Hop count based methods are divided into greedy and recovery schemes. In greedy schemes, each node forwards the message to a neighbor based on the direction, progress or distance criterion. A memoryless (stateless) GFG (greedy-face-greedy) routing algorithm that guarantees delivery (if destination location is accurate) in unit graphs (where nodes can directly communicate iff they are within fixed transmission radius) is presented. We also give design guidelines for network layer protocols in sensor networks. Joint research with several students and colleagues, and published in several papers.
Simulation in ad hoc and sensor networks (criticism of existing practices)
Simulation is certainly useful for evaluating protocol performance. Several recent studies underlined the lack of rigor in its application that threatens the credibility of the published claims. While agreeing with some of the criticism, this article argues against some others and advocates for a different overall view. The primary goal of simulation is to provide sufficient support for new concepts and protocols, for an overall combined contribution. ‘Proof of concept’ is basic (not thorough or testbed based) simulation using assumptions in designed protocol, including comparison with truly competing existing solutions. We advocate for thorough literature review, solving one problem at a time and selecting independent variables that can assist best in explaining the performance under wide range of scenarios. While all models are unrealistic, some of them are useful. We advice for parallel advance of useful modeling and protocol design, where the simplicity is emphasized to preserve tractability.
Ivan Stojmenovic received Ph.D. degree in mathematics He published over 200 different papers and over 30 book chapters, and edited four books on wireless, ad hoc and sensor networks and applied algorithms with Wiley/IEEE. He is currently editor of 15 journals, and founder and editor-in-chief of three journals. Stojmenovic is in the top 0.56% most cited authors in Computer Science (Citeseer 2006). One of his articles was recognized as the Fast Breaking Paper, for October 2003 (as the only one for all of computer science), by Thomson ISI Essential Science Indicators. He founded three ongoing workshop series, chaired 25 program committees, is steering committee member of three conferences, chaired 20 events, and served in 150 program committees. Professor Ivan Stojmenovic received a Royal Society Wolfson Research Merit Award to join EECE at University of Birmingham, UK.