Adaptive Stream Filters for Range Queries with Fraction-based Tolerance

Speaker: Alan KWAN Kang Lun

Abstract

This talk is based on the paper in VLDB05 and our recent work on its extensions. The problem of fraction-based tolerance protocol by applying adaptive filters for approximate range query processing in a distributed stream environment will be discussed. The proposed filter bound assignment protocols are with the objectives of reducing communication cost and energy consumption for sensors. We conduct extensive experiments based on both real and synthetic data to evaluate the protocols' performance. Our results show that the protocols achieve significant savings in both communication overhead and computation, and the extended works further improve the performance significantly.