The overall goal of this project is to demonstrate the effectiveness of the emerging technology of Wireless Sensor Networks in solving the increasingly important problem of species interactions between domestic animals and wildlife. Specific objectives include the development of a reliable wireless sensor network that permits recognition of individual animals and that also provides a distance metric between them in real time. This network will be adapted to include free-ranging white-tailed deer and to collect cow and deer interaction data. These data will be collected and analyzed to quantify the potential for these interactions to facilitate disease transmission between domestic animals and wildlife.

Applications for sensor networks are extremely diverse covering almost every walk of human life. The most unique features of sensor networks are non-traditional: extremely limited battery power, scalability, and durability to extremely treacherous environments. Many new proposals have been made to handle these non-traditional issues. The common theme of these proposals is energy-saving while delivering high performance quality. Although this theme is necessary and may result in high impact, this work offers only a few design choices (of high optimized networks)for application designers. If we plot these offered design choices in the design space formed by tradeoffs between energy consumption, and performance quality such as response time, throughput, resilience to failure, etc., the existing work tends to produce only a few data point in the design space (if not only one) -their performance study focuses on finding the data points in this space that are not possible by earlier work. Although this shows the very existential and motivational reasons for their "new approaches", it is not very helpful to application designers. Because of diverse applications, designers demand (sensor) network protocols that can provide a wide spectrum of design choices, especially for very low energy budget applications. In this proposal, we propose an approach, called Route-aware Media Access Protocols (RASMAC), that can take existing MAC layer protocols and greatly diversify their design choices expressed in the tradeoffs between energy and performance qualities. RASMAC offers many choices, especially over low energy budget areas, by utilizing its unique energy saving feature: opportunistically turning off the radio when a node is not on an active routing path. This approach meshes well with existing MAC schemes such as SMAC and TDMA and also with on-demand routing approaches such as directed diffusion and SPIN. The proposed work uses RASMAC to explore many different choices of MAC for sensor networks which can serve diverse applications, and develops building blocks for a design tool that designers can use to find the right set of "MAC-aware" routing and "route-aware" MAC protocols and their corresponding parameters to use to meet their performance and energy requirements. Our plan includes the development of new MAC and routing protocols that can fit well with RASMAC and their performance modeling.

Support is requested to enable research into the use of battery powered networks of wireless computer systems as elements of a new approach to solving wildlife tracking problems. Wildlife tracking data (e.g., GPS locations, time, acceleration) can be collected by means of sensor equipped collars placed on selected animals. If this data can be efficiently relayed through an inexpensive ad hoc wireless network to a long term data storage facility, wildlife biologists would be able to improve their data collection and thus extract more complete information regarding wildlife behavior and habitat. Research issues that must be explored to leverage this new technology include the definition and understanding of parameters necessary for power management throughout sensor and relay elements of the network. The project proposed here will complement a collaborative effort to improve wildlife tracking techniques that has been underway between NCSU Computer Science faculty and students, and the North Carolina Zoological Park and its affiliate agency, the North Carolina Zoological Society.