Approximate hybrid query processing in wireless sensor networks
Proceedings of the 20th International Conference on Advances in Geographic Information Systems
Energy efficient in-network data indexing for mobile wireless sensor networks
SSTD'13 Proceedings of the 13th international conference on Advances in Spatial and Temporal Databases
| Hi-index | 0.00 |
Data applications in wireless ad hoc sensor networks require scalable localization methods and dynamic data space partitioning among sensors to support efficient query processing. These challenges are exacerbated by sensors' inherent constraints in storage capability, battery power, processing speed and communication bandwidth. To meet the imperative of energy preservation in data manipulation operations without performance degradation, we propose a scalable and self-configuring scheme, which synergistically integrates localization and data space partitioning, thus minimizing the otherwise prohibitive cost of separately performing localization and/or indexing. We first introduce a basic scheme, called simple PRDS, which enables sensors to self-configure into a consistent coordinate system and to dynamically partition the data space in a network with a small core (at least 15% to 20%) of randomly scattered location-aware sensors within the geographical area spanning the network. We then propose an adaptive version of PRDS, which adds a refinement phase driven by a forcebased relaxation method called mass-spring optimization, to the basic scheme to handle cases where the fraction of location-aware sensors is significantly reduced. Seamless transition is achieved between the basic scheme and refinement phase by assessing the current requirement of the network, while concomitantly, without user intervention, performing efficient local re-adjustment of the data space partitioning among neighborhood sensors. Our experiments, under a varying number of location-aware sensors, show that the "adaptive PRDS" achieves better results than the "simple PRDS" at the cost of a small increase in latency, and is significantly superior to the combination of the best previously proposed techniques for localization and indexing.