Portal de Periódicos da CAPES

An Extensibility Approach for Spatio-temporal Stream Processing Using Microsoft StreamInsight

2011; Springer Science+Business Media; Linguagem: Inglês

10.1007/978-3-642-22922-0_38

1611-3349

Jeremiah Miller, Miles Raymond, Josh Archer, Seid Adem, Leo Hansel, Sushma Konda, Malik Luti, Yao Zhao, Ankur Teredesai, Mohamed Ali,

Peer-to-Peer Network Technologies

Integrating spatial operators in commercial data streaming engines has gained tremendous interest in recent years. Whether to support such operators natively or to enable the operator through an extensibility framework is a challenging and interesting debate. In this paper we leverage the Microsoft StreamInsightTM extensibility framework to support spatial operators enabling developers to integrate their domain expertise within the query execution pipeline. We first justify our choice of adopting an extensibility approach over a native support approach. Then, we present an example set of spatiotemporal operations, e.g., KNN search, and range search; implemented as user defined operators using the extensibility framework within Microsoft StreamInsight. More interestingly, the demo showcases the how embedded devices and smartphones are shaping the future of streaming spatiotemporal applications. The demo scenario specifically features a smartphone based input adapter that provides a continuous stream of moving object locations as well a continuous stream of moving queries. To demonstrate the scalability of the implemented extensibility framework, the demo includes a simulator that generates a larger set of stationary/moving queries and streams of stationary/moving objects.