Telecommunication/Telecollaboration

In collaboration with Microsoft, we have begun the TeleEducation/TeleCollaboration and Streaming Media project, which includes network protocols (including floor-control), multicasting, support for caching, and streaming media. These require extensive technology development beyond the current Internet, to effectively and affordably support quality real-time streaming media, dynamic
multicasting, and advanced user interfaces providing additional function and improved ease of use. Microsoft's DISC represents an initial setting for the investigation. On the one hand, it provides many of the envisaged features for the distributed classroom. On the other hand, it assumes no floor passing between teacher and students, and relies on a centralized architecture for the maintenance of the workspace. Based on the above observations, we will address the following research problems:
(1) Impact analysis of the Microsoft's DISC as is, insofar as its ability to provide a seamless integration of group collaboration into the tools used by instructors and students in courses.
(2) Floor control and conference management mechanisms that synchronize user actions and permit concurrent sharing of shared workspaces in a non-intrusive manner, i.e., without requiring the users of the system to adopt a strict protocol of interaction. An integral part of our study will be determining effective ways to embed floor control mechanisms in DISC.
(3) Scalability of the distributed classroom, in which potentially hundreds of students could attend a remote lecture. This aspect of our research entails the investigation of system aspects of the distributed classroom (e.g., what is an efficient way to eliminate
centralized servers or databases that can become bottlenecks in the system) as well as algorithmic and protocol aspects of its components (e.g., how should applications be adapted to the handling of hundreds of participants who are geographically distributed; how should classroom telecasts be supported, given that IP multicast is not supported beyond relatively small testbeds and research networks in the Internet).
(4) Definition and use of multimedia feedback channels. In the case of a lecture, limited audio feedback can be enough; for distance learning, however, a feedback window could be used to provide information to the instructor from the students. The size of a lecture impacts the nature and effectiveness of feedback channels; therefore,
different types of feedback should be available to students.