IAD: Rsch & Educ Infr for Enabling Autonomic Sensor Grid Systems in Multidisciplinary Applications
Project Proposed:
This project, acquiring infrastructure to enable autonomic sensor grid systems (ASGS), integrates two emerging technologies, wireless sensor networks and grid computing. The large-scale sensor network and grid computing platform service the following research projects: . High-Throughput Networking Protocols; . An Autonomic Sensor Grid System Framework; . Collaborative Dynamic Tracking and Information Processing; . Intrusion Detection; and . Enabling Multidisciplinary Applications. The first project investigates techniques to enhance dynamic sensor network throughput while contributing to meet the emerging data intensive and high data rate requirements of many applications. The second designs, implements, and evaluates the ASGC framework that integrates sensor networks and grid computing technologies with self-managing capabilities providing self-managing middleware and user and programming interfaces for query, visualization, analysis, and physical discovery phenomena, and further command and control of actuators/robots for synergistic tasks in coordinated manners. The third develops high confidence coordination algorithms and tools for hybrid (static and motion-enabled) sensor networks and explores power-aware collaborative information processing techniques. The fourth investigates attack modeling and analysis, large-scale intrusion detection and tracing, and deceiving attackers' techniques. The fifth includes Precision Agriculture and Ranching, Protective Clothing, Vehicle Navigation, Homeland Security, Environmental Monitoring, and Rapid Disaster Recovery.
Broader Impacts: The infrastructure contributes in the development of new courses, recruiting under-represented/minority students, outreach and minority participation, and technology transfer to industry. The new courses include Cooperative Robotic Networks, Sensor Networks, Embedded Systems, and Grid Computing. Plans include involving Native Americans and Langston University, an HBCU. Moreover, the facility contributes to broaden education and research in an EPCSoR state. Precision agriculture in crop and livestock production appears to be an interesting and useful application for the state.
We investigated issues in designing, implementing, and deploying cyber-physical systems. By real-world deployment in a farmland, we demonstrated the feasibility and efficiency using low-energy low duty cycle wireless sensor networks for enhancing precision agriculture. We deployed multiple wireless sensor nodes with a carrier frequency of 2.4GHz in an experimental wheat field to measure soil moisture, temperature, and electrical conductivity. A series of systematic field experiments were conducted during three complete winter wheat growth seasons from seeding to harvest. Models were developed to predict the path-loss of radio signals in the field during various growing stages. As one of the first real-world experiments, our experiences and experiments have opened significant opportunity in the state and beyond. To facilitate efficient data gathering in an irregular region (such as a large farmland), we designed an anchor-free SinkTrail protocol that minimizes data collection process in a realworld settings (e.g. a tractor following irregular trails in a land). The novel algorithm features naturally-formed virtual coordinates without relying on any landmarks or predefined locations. In related work, robots with sensors were deployed in a lab setting to measure the most effective sink-sensor combination. Leveraging support from NSF/GENI program, we further extend our sensor network testbed to the cloud and build the OKGems/GemsCloud (http://okgems.ece.ufl.edu/). OKGems provides a public cloud for research and experiments on the ceiling wireless sensornet grid and mobile robots in our lab. To support multi-site sensornets, we proposed MagicLink to seamlessly integrate multi-site sensornets with virtualized cross-site links. Our testbeds have been demonstrated to local high school students, used to host a high school student from Mexico, a visiting PhD student from China, and many academic and industry visitors. Our research has been well received in the community with over a dozen publications in international conference proceedings and journals. Two PhD students supported by the project won the best PhD scholarship in the department in two consecutive years. This CRI grant was also instrumental to develop a robotic cyber-physical system test bed at the Department of Electrical & Computer Engineering, University of New Mexico. The test bed which is part of the MARHES Lab (http://marhes.ece.unm.edu) consists of a fleet of all-terrain ground robots and rotorcraft UAVs (quadrotors) equipped with embedded computers, an environmental sensor suite, and wireless communication capabilities. The support by the CRI program allowed us to attract additional funding from the NSF, DOE and DOD. Several projects use the test bed developed with the partial support of this cyber infrastructure grant. In addition to train students and researchers at the Electrical and Computer Engineering Department at UNM, this test bed has been used by other departments within UNM and other institutions participating in collaborative projects (Duke University, University of Oklahoma-- REU site). Three PhD students and several MS students developed their dissertations and theses using the CRI test bed. Moreover, this project provided the instrument that enabled us to obtain important research results which have been published in top journals and international conferences. Some of these breakthroughs include: (i) the formulation of communication-aware motion planning for robotic networks (collaboration with Prof. Yasamin Mostofi, NSF IIS); coordination of heterogeneous sensor networks via Adaptive Dynamic Programming and Hybrid Systems (collaboration with Prof. Silvia Ferrari, NSF ECCS); and agile transportation using teams of UAVs. On the educational and outreach side, we taught at UNM two successful courses that cover different aspects of the proposed work such as Autonomous Mobile Robots and Control of Hybrid Systems. We have advised several students including Hispanic and female students to work on different aspects of the proposed work towards their degrees. We have also hosted many K-12 visitors in our labs, with a particular emphasis on Native American and Hispanic students and developed workshops and robotic games for some of the local Hispanic high school students on our campus.
