SENSORS: Biological Sensor Data Acquistion System

Cornell University, Ithaca NY

Investigators

Abstract

Tagging animals with miniature electronic instruments provides unique opportunities to observe physiology, movements, and social behavior in a free-ranging context. The addition of environmental sensors to animal tags provides the capacity to monitor ecological and oceanographic processes, and the animals can be viewed as self- sustaining autonomous vehicles that focus sampling effort on regions of biological interest. Although some scenarios permit recovery of the tags, a much broader domain of applications requires telemetry of data from the tags, usually using radio frequency signals. Radio tags confront conflicting demands: minimization of transmit power to enable extended operations with small batteries, and maximization of the area over which the tags can be tracked. In the current systems for marine animal (e.g., whales) data acquisition, the attached (implanted) radio tags measure data about the animals and their environments and store the data in memory as packets. When the animal surfaces in view of a satellite, this data is transmitted to the satellite where the biologist can retrieve it. Once the packet is transmitted to the satellite, it may be deleted from storage on the animal-implanted device. The devices on animals transmit at preset intervals to the satellite. However, the use of the satellite system accepts certain limitations: substantial transmission power requirements, high cost, and probable limits on connections, because satellites may not always be in view. The last problem may prove serious if the tags cannot transmit at any time. Many marine mammal species have dive durations exceeding 30 minutes, and may spend less than 15% of their time on the surface. Moreover, the cost and energy consumption are critical issues due to the low battery power of the implants. In this proposal, the proposer advocate a different approach: the infrastructure is extended to the animals (mobile nodes) by the mobile nodes themselves, by creating a sensor network infrastructure composed of the animal tags. The information that is created in a network node is allowed to be diffused (replicated) among the nodes, as the nodes come into contact due to their mobility. In other words, as a node comes into reception distance from another node, the information stored in the memory of the node is replicated and stored in the memory of the recipient node. An information collecting station is placed within the animal habitat, and when an animal comes within the reception range if a collecting station, all stored information in the memory of that animal tag is offloaded to the collecting station. To increase the probability that the information is recovered from the network, a number of collecting stations can be distributed throughout the habitat. Thus, only one of each information piece needs to reach only one collecting station to be successfully offloaded. The propoer term this networking model the Ad Hoc Infostation model. As part of the research, it propose to address three areas: research into the design and performance evaluation of the Ad Hoc Infostation model, development of a practical sensor tag that would operate in a network of Ad Hoc Infostations, and integration of the sensor tag into a network of Ad Hoc Infostations. In particular, examples of topics address under the first area include design of the network of Ad Hoc Infostations to reach the terrestrial network, networks that include a hierarchy of different-mobility nodes, such as those that include mobile Ad Hoc Infostations, and examination of the "capacity-delay" tradeoff offered by the Ad Hoc Infostation model. In their work, plan on using the vast knowledge in epidemic modeling to leverage from in studying the Ad Hoc Infostation model. Under the second area, a new transceiver tag will be developed, which will support all of the RF capabilities needed for the wireless Ad Hoc Infostation networking nodes. Under the third area, the tags and Ad Hoc Infostation receivers will be tested locally, and evaluated in two pilot projects to assess their biological utility. Initial functional tests will be conducted by setting up Ad Hoc Infostations in the local Tompkins Country, and driving tags around to simulate animal movements. The second project will utilize the networking capabilities to gather bioacoustic data from a matrix of dispersed fixed sensors. As the broader impact of the proposal, whales and other marine mammals are crucial to several marine environmental issues and improved understanding of the factors that enable prediction marine mammal distributions, and of the responses of marine mammals to anthropogenic noise, is a long0term research priority of naval research. Also, NSF has proposed a National Environmental Observation Network, a 10 year $120M program. Two of the proposed 12 sites have been earmarked for funding in the President's proposed budget. Our proposed system could implement a heterogeneous network of stationary instruments, relatively sedentary animals, and wide-ranging animals to efficiently transmit telemetry data from areas that are inaccessible or very difficult to access.