[Home] |
||||||||
|
||||||||
System
Support for Ubiquitous Computing Workshop at the Fifth Annual Conference on Ubiquitous Computing (UbiComp 2003) |
||||||||
This workshop offers the opportunity to bring together researchers and practitioners involved in the development of systems support for general purpose ubiquitous computing environments. It aims at exploring most recent research and findings in this area, comparing results, exchanging experiences, and promoting collaboration and cooperation among researchers in the field. The workshop focuses on the common abstractions and patterns found in the existing systems, as well as the core low-level services that are needed to build general-purpose ubiquitous computing environments. The workshop targets different aspects of system and middleware research and the challenges involved when applying them to support ubiquitous computing. Ubiquitous computing environments are envisioned as being populated with large numbers of computing devices and sensors to the extent that the physical and computational infrastructures become fully integrated, creating a dynamic programmable environment. To realize this vision, several projects have developed prototype environments, typically focused on a particular ubiquitous computing scenario or application. System support is often pragmatic, problem oriented and difficult to generalize to other domains. To fully realize programmable ubiquitous computing environments it is essential to provide services that coordinate software entities and heterogeneous networked devices and provide the low-level functionality needed to enable ubiquitous computing in the general case. Systems software provides a homogeneous computing environment where applications are supported with resource management (i.e. resource and service discovery), and common abstractions that leverage the implicit heterogeneity in such environments. The work of independent researchers has revealed patterns of service usage, indicating that systems software for ubiquitous computing may converge to a set of necessary core services. If such a set of requirements could be identified, applications could then be more easily ported across different implementations and interoperability would be simplified. One of the key issues for debate is the underlying structure of ubiquitous computing middleware. Current prototypes are characterized by three different architectural organizations. In the first case, the environment provides an infrastructure that coordinates the resources present in a specific geographical location. Applications can discover and access such resources only via the infrastructure. Furthermore, all communication between devices is mediated by the infrastructure. Additional information, such as large amounts of application data which for instance cannot be stored on small devices, can be maintained in the infrastructure. Direct interaction between devices is not considered and the infrastructure typically provides services localized to a specific geographical area such as a room, or a building, covered by one or more network types. The second architectural organization relies on spontaneous interaction among the devices present in the environment as a federation of peers. There is no common infrastructure per se -- applications have to store and maintain application data cooperatively. Such an organization is typically based on an underlying ad-hoc configuration. The third model is a hybrid that relies on a centralized system support infrastructure, but relies on peer-to-peer communication among entities. So far, the three approaches seem to offer benefits in distinct application domains and it is likely that they will continue to co-exist and complement each other. These three approaches lead to a variety of research questions concerning interoperability, architecture, service organization, application models, and application support in general. This workshop will provide a unique opportunity to foster discussion and interaction among researchers working on this new area.
The workshop focuses on presenting state of the art and emerging research, as well as experience reports, in the following topics: · System support infrastructures and services · Middleware for ubiquitous computing · Architectural structure, design decisions and philosophies
In order to ensure a high quality technical session, paper submissions must be 5-8 pages long in LNCS format and have to cover one of the topics listed above. Furthermore, we will prioritize experience papers describing lessons learnt from built systems, including information about approaches that did and did not work, unexpected results, common abstractions, abstraction mapping among different systems, common building blocks present in different architectures, and metrics for evaluating ubiquitous computing infrastructures. It is highly recommended that submissions meet these guidelines:
Submissions must be blinded for
peer review (no author names and affiliations and no obvious references).
Blinded submissions, in PDF format, must be
emailed to
ubisys@cs.uiuc.edu no later
than Submissions will be reviewed blindly and selected based on their originality, merit, and relevance to the workshop. All accepted papers must be presented during the workshop. At the end of the workshop, we will have a discussion session, where we summarize the findings of the workshop and discuss future work and possible collaborations. It is planned to publish selected papers in revised version in a LNCS post-proceedings. Please email us if you have any questions.
|
||||||||
|