Absynth: Making Wireless Sensor Networks Easy to Design for Applications Experts via Languages and Synthesis Technologies


Summary People Contact

Summary

Wireless sensor networks have the potential to revolutionize research areas and industries that require the distributed collection and aggregation of data, e.g., civil engineering, biology, and geology. However, to date their impact has fallen short of this potential. This is not surprising; wireless sensor networks are difficult to design and program. Experts in application domains such as biology and civil engineering have neither the training in embedded system programming and design required to develop adequate wireless sensor networks, nor the time or inclination to become embedded systems designers. For wireless sensor networks to live up to their potential, they must be easy for application experts to design and program instead of requiring embedded systems design expertise.

This project seeks to put wireless sensor network design and deployment within the reach of applications experts. This project will identify a small set of application archetypes, described in terms meaningful to application developers, that capture the most common application structures. Each archetype will be backed by a compilable specification written in a simple high-level archetype-specific language. After customization, a specification will be passed to a synthesis algorithm to produce a working hardware-software system. Synthesis of efficient wireless sensor networks would be intractable without an appropriate hardware-software platform. Therefore, this project will also determine the particular types of hardware and software components necessary to enable support for wide range of archetypes. A configurable hardware-software platform will be developed based on these findings.


People

Page maintained by Lan Bai.