Hardware Independent Networked Active Sensor Middleware

The pioneering efforts within the DARPA SENSIT and NEST programs, proved the potential of wireless sensor network (WSN) technology for compelling DoD applications, including surveillance and target localization. These projects'' legacies include hardware platforms, used extensively in academic research today and the TinyOS software suite. At the same time, these programs revealed the limitations of passive sensors and the challenges of developing WSN applications. Not only application development is hard and error-prone, but the resulting code is tightly coupled with the underlying hardware and sensor suite. This tight integration prevents code reuse as the hardware and sending platforms evolve. In this project we will address both limitations through a hardware-independent middleware for multi-static WSNs. Doing so requires designing novel distributed target detection and localization algorithms for active WSNs. These algorithms maximize coverage while minimizing energy use. Furthermore, the middleware abstracts application logic from hardware intricacies. To do so, we introduce the programming abstractions and system services that evolve TinyOS to a system that provides a clear but flexible delineation between applications and an evolving kernel. The key innovation is the ability to provide this interface and the system services that define it at no expense of system efficiency.