Current Projects

  • Self organized networks like mobile ad hoc networks and sensor networks enable pervasive computing applications to observe and monitor the physical world. Queries present an application-friendly abstraction to gather information from these emerging opportunistic networks. However, these networks are inherently dynamic because of changes induced by mobility and the unpredictable lifetime inherent to battery operated devices. Consequently, the state of the environment changes during the execution of a query.

  • The Open Workflow project is investigating the implications of the use of dynamic workflows as a mechanism for coordination and collaboration in a mobile environment. The quintessential feature of the open workflow paradigm is the ability to construct a custom context-specific workflow specification on the fly in response to unpredictable and evolving circumstances by exploiting the knowhow and services available within a given spatiotemporal context.

  • We envision a future where applications written by third-party developers are deployed on networks of common devices (e.g., TVs, thermostats, refrigerators). To accomplish this vision, modern "closed" devices must be "opened" both physically through communication hardware and logically through software interfaces. Developers also need a software infrastructure (i.e., middleware) that allows them focus on the problem to be solved an hides complex and mundane details like discovery, communication, etc.

  • Servilla introduces service-oriented computing into the realm of wireless sensor networks for the purpose of allowing applications to adapt to network heterogeneity. The decoupling provided by service-oriented computing enables a modular middleware framework that adapts to even severely resource-constrained devices.

  • Real-time patient monitoring enables timely prediction of clinical deterioration of non-ICU inpatients. To provide timely care of patients with deteriorating conditions, it is essential to update existing electronic medical records with real-time sensor data continuously collected from patients. To meet this challenge we have proposed a multidisciplinary project to develop and evaluate novel wireless sensor network technologies for real-time patient monitoring.

  • To facilitate advanced research in wireless sensor network technology, we are currently in the process of deploying a wireless sensor network testbed at Washington University in St. Louis.

  • Wireless sensor networks have traditionally been used as specialized single-application platforms. However, with the emergence of integrated sensing systems such as automatic building management, wireless sensor networks are evolving into shared, multi-application infrastructure. This paradigm shift raises the need for a framework for deploying multiple applications on a wireless sensor network such that application QoS requirements are met while also optimizing the use of the limited sensor network resources.

Recent Projects

  • CiAN is a workflow management system that has been designed with mobile devices and mobile ad hoc networks in mind. CiAN offers a modular, mobility-aware workflow specification, a library of distributed algorithms to manage task allocation in the presence of mobility, and a distributed choreographed execution engine for executing workflows that involve humans and software services across the physical world.

  • Sliver is a micro workflow management system that supports the execution of workflows written in the Business Process Execution Language (BPEL) on small mobile devices such as cellphones and PDAs. Sliver uses a set of novel strategies to shoehorn a BPEL workflow management system that typically executes on enterprise servers into a smaller footprint suitable for mobile devices.

  • Agilla is a sensor network middleware that allows the rapid development and deployment of sensor network applications using mobile agent technology. Agilla provides the means to design simple agents using an assembly-like language and subsequently inject these agents into a sensor network where through logical mobility and cloning, they rapidly disburse within the network and provide the desired functionality.

  • The communication delay associated with a multi-hop transmission across a sensor network is affected by the spatial distribution of both data flows and sensors. At the same time, sensor networks need to deal with the dynamics introduced by the mobility of users, physical objects, and sensor nodes. This project established a spatiotemporal communication framework for wireless sensor networks through a field of study that integrates real-time systems and mobile computing research.

  • SPAWN is a service-oriented computing middleware for groups of mobile devices that are connected by a mobile ad hoc network. SPAWN builds on the principles of Sun's Jini middleware and offers basic service computing primitives as well as advanced features such as automated code management, run-time upgrade of services, logic ally mobile services, and services that exploit knowledge to produce better matches in a mobile environment.

  • EgoSpaces facilitates rapid development of sophisticated context-aware applications by allowing applications to define contexts that reflect diverse concurrent and changing needs and encompass data from multiple sources. EgoSpaces manages this information for the application, relieving the developer from having to handle network connections and disconnections common in mobile environments.

  • The Limone coordination middleware is a lightweight derivative of LIME that is is designed to be better suited to accommodate transient interactions and limited resources typical of ad hoc networks. Limone eliminates the strong guarantees offered by LIME and uses a system of beaconing and acquaintance lists to build a more fluid and dynamic snapshot of its peers enabling more efficient communication.

  • The LIME (Linda In a Mobile Environment) middleware enables rapid development of dependable applications that exhibit physical mobility of hosts, logical mobility of agents, or both. LIME extends the Linda model of process coordination to allow tuple spaces to be distributed in time and space and transiently shared when connectivity is available. Movement, logical or physical, results in implicit changes to the contents of the tuple space accessible to the individual components.