Title :
EcoCast: Interactive, object-oriented macroprogramming for networks of ultra-compact wireless sensor nodes
Author :
Tu, Yi-Hsuan ; Li, Yen-Chiu ; Chien, Ting-Chou ; Chou, Pai H.
Author_Institution :
Department of Computer Science, National Tsing Hua University, Taiwan
Abstract :
EcoCast is an execution framework for macroprogramming of wireless sensor networks. Users access sensor nodes as dynamic objects in Python by invoking methods on them without being concerned with network protocols, and type marshalling and demarshalling ensure proper data access. EcoCast extends Python´s functional programming primitives map(), reduce(), and filter() to macroprogramming with several synchrony semantics and job-control options. EcoCast can compile Python lambda expressions and functions to run on the nodes at native speed without requiring most users to write code in C or assembly, and it patches the firmware transparently without rebooting. The use of Python also facilitates host-side application development by enabling developers to take full advantage of the rich code libraries and data structures in Python. Experimental results show the reprogramming and execution latencies of EcoCast scale well over the size of the network while occupying a small memory footprint.
Keywords :
dynamic loading; macroprogramming; wireless sensor network;
Conference_Titel :
Information Processing in Sensor Networks (IPSN), 2011 10th International Conference on
Print_ISBN :
978-1-61284-854-9
Electronic_ISBN :
978-1-4503-0512-9