Provided by: University of Washington School of Public Health & Community Medicine
Date Added: Dec 2010
Many sensor network protocols utilize the existence of disjoint paths (e.g., perfectly secure message transmission or multi-path key establishment), but do not address how a node actually determines these paths in the presence of an adversary. In this paper the authors investigate what assumptions are necessary to gather information about the local network topology when adversarial nodes are present and capable of lying about their identity or neighbors in the network. These assumptions are practical, and realizable through existing tools such as combinatorial key pre-distribution, fingerprinting, and localization.