Power efficiency through tuple ranking in wireless sensor network monitoring

Abstract

In this paper, we present an innovative framework for efficiently monitoring Wireless Sensor Networks (WSNs). Our framework, coined KSpot, utilizes a novel top-k query processing algorithm we developed, in conjunction with the concept of in-network views, in order to minimize the cost of query execution. For ease of exposition, consider a set of sensors acquiring data from their environment at a given time instance. The generated information can conceptually be thought as a horizontally fragmented base relation R. Furthermore, the results to a user-defined query Q, registered at some sink point,
can conceptually be thought as a view V . Maintaining consistency between V and R is very expensive in terms of communication and energy. Thus, KSpot focuses on a subset V′ (⊆ V ) that unveils only the k highest-ranked answers
at the sink, for some user defined parameter k. To illustrate the efficiency of our framework, we have implemented a real
system in nesC, which combines the traditional advantages of declarative acquisition frameworks, like TinyDB, with the ideas presented in this work. Extensive real-world testing and experimentation with traces from University of California-Berkeley, the University of Washington and Intel Research Berkeley, show that KSpot provides an up to 66% of energy savings compared to TinyDB, minimizes both the size and number of packets transmitted over the network (up to 77%), and prolongs the longevity of a WSN deployment to new scales.