GatorCloud: SDN-enabled Campus Cloud
We design the GatorCloud cyberinfrastructure to dramatically boost the Campus Research Network (CRN) to over 100Gb/s backbone networks connecting multiple primary sites of data centers across the campus and further enhance it with the cutting-edge software-defined networking (SDN) capabilities and novel cloud services. GatorCloud is the first SDN-enabled campus cloud initiative, offering novel cloud services, such as HPC as a Service (HPCaaS), Network as a Service (NaaS), and Cloud-in-Cloud as a Service (CCaaS). It will have transformative impact on shaping the future of cyberinfrastructure. It provides a production-level environment for research in the forefront of computer science and engineering. (1) CloudBay: Computational EBay for Resource Trading. (2) MyCloud: Creating Personalized Clouds in Conventional HPC/Supercomputing on Demand. (3) GolfClub: SDN controller, hypervisor, app store, and app engine.
CognitiveEngine (Florida Brain)
CognitiveEngine (Florida Brain) provides an intelligent engine with realtime deep learning capabilities. It consists of large-scale realtime messaging, processing, storage, and interaction cloud services. It supports large-scale deep learning networks, multiple jobs, multiple tenants, and multiple categories of applications. It also features crowdsourcing and online fusion. It has a broad spectrum of applications in multimedia, video on demand, surveillance, mobile social networks, security and privacy, business, engineering, science, and precision medicine.
SMART: Scalable Adaptive Runtime Management Algorithms and Toolkit
Emerging general-purpose high-end computing and storage resources provide great opportunities to tackle grand challenge problems in science and engineering. However, most systems do not have adequate support to meet special requirements of a number of ultra-scale dynamic scientific applications due to lack of application-awareness. These computation and data intensive applications are aimed to model and investigate highly dynamic and sometimes drastically changing phenomena in science and engineering, such as interacting black holes, global and regional high-resolution weather forecasting. Furthermore, it is challenging, error-prone, and time-consuming for scientists/engineers to develop their large-scale parallel and distributed scientific applications from scratch. SMART fills the gap and aims to create an integrated framework of scalable adaptive runtime management algorithms, libraries, and toolkit with friendly programming models; (2) design a suite of application-aware adaptive algorithms to holistically address various issues; and (3) enable high-impact real-world large-scale scientific applications.
HiCloud: Taming Clouds in Clouds
HiCloud addresses the most challenging problems of very-high-resolution numerical weather prediction, obtaining the optimal state estimations for initializing ensembles of predictions by assimilating the highest volume of weather observations available, and addressing problem sizes and scales that are only attainable on petascale computing platforms. An efficient and easy-to-use programming toolkit for a large class of weather forecasting applications will be developed. The programmer-friendly runtime management system and programming model will also be enabled in heterogeneous cloud environments. Apala adaptive partitioning anad load balancing for state-transition applications.
Guoguo: Fine-Grained Indoor Localization
Smartphones are emerging as the pervasive computing platform of the future. Despite widely available GPS for outdoor localization, practical indoor localization has been a signficant challenge and a major hurdle for a wide spectrum of potential indoor applications. The Guoguo algorithm and system are proposed to solve this long-lasting problem. Leveraging the pervasive microphone in a smartphone, we design fine-grained indoor localization algorithms using audible-band acoustic signals. Guoguo means grasshopper in Chinese.
ToGathor: the Mobile Social Network with Missions
Coming ToGathor is a beginning; Keeping ToGathor is progress; Working ToGathor is success. ToGathor provides a context-aware mobile social network to accomplish our daily events and activities. It features opportunistic sensing, messaging, localization, and rich context. It enables smart life, smart circles, and collaborative social efforts. (1) FindingNemo: finding lost children via crowdsourcing; (2) TotalRecall: automatic statistics for education, conferences, parties, and entertainment; (3) InstantWhisper: instant messaging on the spot; (4) SocialEngine: deep understanding of realworld social pulses with floating and massive information underexplored. Spatial-temporal information, events, crowds, physical contexts, and their interplay are yet to be revealed.
GemsCloud: Cloud of Sensors, Actuators, Networks, and People
GemsCloud is an umbrela project that covers many ongoing subprojects. (1) OKGems: GENI-Federated Cyber-Physical Sysems. OKGems is aimed to design a remotely programmable cloud of physical sensors and actuators. It can be accessed and managed through a GENI control framework. MagicLink is further designed to enable multi-site seemless integration. (2) YouGo: Physical Search Engine via Cyber-Physical Social Networks. YouGo leverages both computing power and human intelligence for dificult physical search and Q&A problems. (3) CirclesScope: Walking Circles. CirclesScope attempts to build a self-configured dynamic social networks that connect real people in the physical world. No typing but walking; building your friendship by your feet not by your hands.
VitalCloud: Weaving a Life Cloud
There are about one billion visits nation-wide to see physicians in offices per year. Millions of these are likely unnecessary. VitalNet is designed to address the needs of healthcare as the average age of our society grows significantly. Automatic vital sign sensing, no-hassle deployment, and pervasive networks enabled by VitalNet substantially release the overwhealming pressure on the capacity of our public healthcare facility and professionals. In the meantime, it further improves responses and context-aware diagnosis from our healthcare providers.
Privacy protection is of critical concern to Location-Based Service (LBS) users in mobile computing and mobile social networks. It enables users to enjoy LBS without worry. Privacy-preserving is supported at clients, servers, service providers, and their interactions and information sharing. iShare: enabling private media sharing in the clouds. iZone: enabling mixed zones for dynamic anonymization. pZone: mobile privacy-preserving substrate and platform.
TrustNet: Trust and Reputation Management in Distributed Systems
TrustNet builds reputation systems for interacting people, computing entities/objects, agents, mobile devices/apps, and other nodes/peers in distributed systems and mobile networks. We proposed one of the first personalized group trust mechanisms and investigated truthful reporting via incentive mechanisms. TrustNet offers a decentralized open framework and middleware for distributed systems to establish, infer, aggregate, and query trust ratings.
S3PAS: Scalable Shoulder-Surfing Resistant Password Authentication System
The vulnerabilities of the textual password have been well known. Users tend to pick short passwords or passwords that are easy to remember, which makes the passwords vulnerable for attackers to break. Furthermore, textual password is vulnerable to shoulder-surfing, hidden-camera and spyware attacks. Graphical password schemes have been proposed as a possible alternative to text-based scheme. However, they are mostly vulnerable to shoulder-surfing. S3PAS seamlessly integrates both graphical and textual password schemes and provides nearly perfect resistant to shoulder-surfing, hidden-camera and spyware attacks. It can replace or coexist with conventional textual password systems without changing existing user password profiles. Moreover, it is immune to brute-force attacks through dynamic and volatile session passwords. S3PAS shows significant potential bridging the gap between conventional textual password and graphical password.
More Past Projects
- ASGrid: Autonomic Sensor Grid System (Funded by NSF and OSU)
- Project A*: Autonomic Grid Computing (Funded by NSF, DHS, and OSU)
- GrACE: Grid Adaptive Computational Engine (Funded by DOE, Rutgers)
- ARMaDA: Adaptive Runtime Management of Dynamic Applications (Funded by DOE and NSF, Rutgers)
- AutoMate: Autonomic Component Framework for Grid Applications (Funded by NSF, Rutgers)
- Gridulator: Grid System Simulator (Grid ROI Calculator) for Commercial Applications (IBM Extreme Blue Program)
- PIng: IPv6 and Mobile IPv6 (CWC/I2R, Singapore)
- DIPS: Distributed Image Processing System (Funded by NUS, Singapore)
- CACDES: Computer-Aided Conceptual Design Expert System (Funded by NSF of China and CAD & CG National Key Laboratory, ZJU, China)
Acknowledgement: This material is based upon work supported in part by the National Science Foundation, National Institutes of Health, and other sponsors. Any opinions, findings, and conclusions or recommendations expressed in this material are those of the author(s) and do not necessarily reflect the views of the National Science Foundation and other sponsors.