Affiliation:
1. MIT CSAIL, Cambridge, MA
2. University of California, Los Angeles, CA
3. Microsoft Research, Redmond, WA
Abstract
Recent work in wireless embedded networked systems has followed
heterogeneous
designs, incorporating a mixture of elements from extremely constrained 8- or 16-bit “Motes” to less resource-constrained 32-bit embedded “Microservers.”
Emstar is a software environment for developing and deploying complex applications on such heterogeneous networks. Emstar is designed to leverage the additional resources of Microservers by trading off some performance for system robustness in sensor network applications. It enables fault isolation, fault tolerance, system visiblity, in-field debugging, and resource sharing across multiple applications.
In order to accomplish these objectives, Emstar is designed to run as a multiprocess system and consists of
libraries
that implement message-passing IPC primitives,
services
that support networking, sensing, and time synchronization, and
tools
that support simulation, emulation, and visualization of live systems, both real and simulated. We evaluate this work by discussing the Acoustic ENSBox, a platform for distributed acoustic sensing that we built using Emstar. We show that by leveraging existing Emstar services, we are able to significantly reduce development time while achieving a high degree of robustness. We also show that a sample application was developed much more quickly on this platform than it would have been otherwise.
Publisher
Association for Computing Machinery (ACM)
Subject
Computer Networks and Communications
Cited by
41 articles.
订阅此论文施引文献
订阅此论文施引文献，注册后可以免费订阅5篇论文的施引文献，订阅后可以查看论文全部施引文献