NeTS - NOSS: Reality-Aware Sensor Network Protocols

University Of Virginia Main Campus, Charlottesville VA

Investigators

Abstract

Abstract: Wireless sensor networks (WSN) have two key advantages over more traditional sensor networks: they can be quickly deployed, and can provide fine-grained sensing. However, many of the proposed solutions use simplifying assumptions about wireless communication and the environment. While many of these current solutions work well in simulation, it is either unknown how the solutions work in the real world or they can be shown to work poorly in practice. Our objectives are to: (i) Assess how the theoretical properties of wireless communication are exhibited in today's and tomorrow's devices, (ii) Establish models of communication realities to feed back into improved simulation tools, (iii) Invent new network protocols that work in real world environments, and (iv) Synthesize novel solutions into a complete system-wide protocol stack for real applications. Our work contributes to an understanding of the limitations of current and future devices and creates viable, deployable systems. New fundamental models are also derived from our work. The expected results of our work are that realistic wireless sensor networks can be deployed in agriculture, environmental science applications, and for emergency response systems. These systems can provide true benefits to society, e.g., more efficient production of food, environmental protection of watershed areas, and saving lives in response to a natural disaster. Dissemination of WSN software through SourceForge enables other groups to build applications on top of our work. Dissemination of our simulation models provides an amplification effect of our research. New courses make novel WSN technologies accessible to graduate and undergraduate students.