Output details
11 - Computer Science and Informatics
University of Edinburgh
Distributed and compact routing using spatial distributions in wireless sensor networks
<06> Originality: Provides a more general perspective of routing in graph metrics than available previously.
Significance: Shows that nearly shortest paths can be found in networks where an approximate distance oracle is available, such as a geographic location system. Uses a probability distribution generalised from small-world models. Operates in a distributed fashion where each sensor computes only a few (~log(n)) shortest paths from itself in preprocessing. Published in ACM Transactions on Sensor Networks, the most important journal for sensor networks.
Rigour: The properties of short paths and low overhead are proven analytically and verified in simulations.
