Title :
Dynamically Defined Processes for Spatial Computers
Author_Institution :
BBN Technol., Cambridge, MA, USA
Abstract :
A program executing on a spatial computer must be able to react to changes in its environment. For example, a sensor network tracking flocks of birds needs to be able to create a spatially-extended tracking process for each flock that it detects, and these processes should not interfere with one another. When dynamically defined processes are identified using fixed equivalence classes (e.g. UIDs), however, the independent creation dilemma means that no algorithm can safely create processes in less than O(diameter/c) time, where c is the speed of information propagating through the network. This dilemma can be evaded by defining the extent of a process with a comparator that does not form an equivalence class. I an example of the dilemma and its resolution using the example of a sensor network tracking flocks of birds, as well as proposing an extension of the Proto spatial computing language to handle dynamically defined processes.
Keywords :
distributed processing; equivalence classes; Proto spatial computing language; bird flock tracking; dynamically defined process; fixed equivalence class; independent creation dilemma; sensor network; spatial computer; spatially-extended tracking process; Approximation methods; Birds; Computers; Heuristic algorithms; Manifolds; Robot sensing systems; Tracking;
Conference_Titel :
Self-Adaptive and Self-Organizing Systems Workshop (SASOW), 2010 Fourth IEEE International Conference on
Conference_Location :
Budapest
Print_ISBN :
978-1-4244-8684-7
Electronic_ISBN :
978-0-7695-4229-4
DOI :
10.1109/SASOW.2010.74