U NIVERSITY OF M ASSACHUSETTS, A MHERST Department of Computer Science From Cloud Computing to Sensor Networks: Distributed Computing Research at LASS Prashant Shenoy Lab Description Seminar 2008
U NIVERSITY OF M ASSACHUSETTS, A MHERST Department of Computer Science 2 Internet-enabled Applications Proliferation of web-enabled and networked applications Increased use in consumer and business worlds Brokerage/ banking online gameonline store Growing significance in personal, business affairs Focus: networked server applications
U NIVERSITY OF M ASSACHUSETTS, A MHERST Department of Computer Science 3 Data Centers Networked apps run on data centers Data Centers Large clusters of servers Networked storage devices Allocate resources to meet application needs Modern data centers are increasingly virtualized
U NIVERSITY OF M ASSACHUSETTS, A MHERST Department of Computer Science 4 Cloud Computing Hosted applications Rent server resources (on-demand) Run applications on rented servers SLA: contract between server and app provider Examples Amazon Elastic computing cloud (EC2) Google App Engine Advantages No need to maintain expensive infrastructure Scale on demand
U NIVERSITY OF M ASSACHUSETTS, A MHERST Department of Computer Science 5 Ongoing Research Dynamic capacity provisioning Adaptive monitoring in data centers Modeling data center applications and systems Byzantine fault-tolerant internet applications
U NIVERSITY OF M ASSACHUSETTS, A MHERST Department of Computer Science 6 Sensor Networks as a Distributed System Hierarchical wireless networks composed of low power sensors. Enables densely and closely monitoring of phenomena. Tracking Surveillance Structure/Machinery Monitoring
U NIVERSITY OF M ASSACHUSETTS, A MHERST Department of Computer Science 7 Emerging Rich Sensor Networks Richer energy Tethered power High data rate Many MB/second Diverse users/applications needs E.g. First responders, Commuters, Insurance, for traffic monitoring Radar Sensor Network Camera Sensor Network
U NIVERSITY OF M ASSACHUSETTS, A MHERST Department of Computer Science 8 Ongoing Research Sensor Virtualization “rich” sensors will need to serve multiple users Virtualize the network to enable concurrent use Remote Management of sensor networks Remote deployments require remote management capability Management plane that separates control from data Energy harvesting in Solar-powered Networks
U NIVERSITY OF M ASSACHUSETTS, A MHERST Department of Computer Science 9 River Sensor Network Application: Monitor river dynamics (e.g: seasonal, flood events), ecological status, water quantity and quality information. Sensors: Water quality sensors, underwater cameras, etc. Research: Design of wide-area, remote wireless sensor network infrastructure. Collaboration between faculty at Mt. Holyoke, NSM and Engineering at UMass, and Hampshire College. Initial Deployment site: 12 mile stretch of Fort River, Amherst
U NIVERSITY OF M ASSACHUSETTS, A MHERST Department of Computer Science 10 Research Methodology Typical experience for research student Approach to advising Mailing list: Lass Alumni Combination of academia and industry 5 in academia: Abhishek (Minnesota), Bhuvan (Penn State), Puru (IIT Bombay), Huan (Beihang U), Peter (Northeastern) 2 in Industry: Vijay (Microsoft), Xiaotao (EMC)