@MISC{Kowshik11informationaggregation, author = {Hemant Jagadish Kowshik}, title = { INFORMATION AGGREGATION IN SENSOR NETWORKS }, year = {2011} }

Share

OpenURL

Abstract

In many sensor network applications, one is interested only in computing some relevant function of the sensor measurements. In this thesis, we study optimal strategies for in-network computation and communication in such wireless sensor networks. We begin by considering a directed graph G = (V,E) on the sensor nodes, with a designated collector node, where the goal is to characterize the rate region in R |E|, i.e., the set of vector rates for which there exist feasible encoders and decoders which achieve zero-error computation for large enough block length. For directed tree graphs, we determine a necessary and sufficient condition for each edge that yields the optimal alphabet, from which we then calculate the minimum worst case and average case complexity. For general directed acyclic graphs, we provide an outer bound on the rate region by finding the disambiguation requirements for each cut, and describe examples where this outer bound is tight. Next, we consider undirected tree networks, where each node has an integer measurement, and all nodes want to compute a symmetric Boolean function. For a class of functions called sumthreshold functions, we derive an optimal strategy which minimizes the worst-case number of bits exchanged on each edge. In the case of general graphs, we present a cut-set lower bound, and an