Towards a Scalable Dynamic Spatial Database System by Joaquín Keller, Raluca Diaconu, Mathieu Valero.
Abstract:
With the rise of GPS-enabled smartphones and other similar mobile devices, massive amounts of location data are available. However, no scalable solutions for soft real-time spatial queries on large sets of moving objects have yet emerged. In this paper we explore and measure the limits of actual algorithms and implementations regarding different application scenarios. And finally we propose a novel distributed architecture to solve the scalability issues.
At least in this version, you will find two copies of the same paper, the second copy sans the footnotes. So read the first twenty (20) pages and ignore the second eighteen (18) pages.
I thought the limitation of location to two dimensions understandable, for the use cases given, but am less convinced that treating a third dimension as an extra attribute is always going to be suitable.
Still, the results here are impressive as compared to current solutions so an additional dimension can be a future improvement.
The use case that I see missing is an ad hoc network of users feeding geo-based information back to a collection point.
While the watchers are certainly watching us, technology may be on the cusp of answering the question: “Who watches the watchers?” (The answer may be us.)
I first saw this in a tweet by Stefano Bertolo.