Goals
web site under construction – check back later for more detail
Here are a few goals of this project:
- information mass transit – exploit similarities in user requirements
- alternative protocols – explore alternatives in niche areas
- scalability – explore scalability issues in mass content delivery
- beyond current assumptions – look at how trends in technology and internet use break the basis of earlier designs
Previous work has focused on information mass transit (IMT). By analogy with mass transit, common user requirements are grouped as much as possible. Theoretically, an personal point-to-point link is faster. In practice, sharing bandwidth (e.g., by joining an existing stream and picking up the part you missed later) reduces congestion at the cost of a small overhead on all transactions.
The approach already investigated is called virtual multicasting (VMC); more work on that area is possible.
Other areas on which future work is possible include improved models for video on demand, specialized networking infrastructure for high bandwidth requirements and alternative models for content delivery in general.
Published Work
Here are a few publications in this area:
[Philip Machanick and Brynn Andrew] Virtual Multicasting as an Example of Information Mass Transit, Proc. SAICSIT-2000, Cape Town, November 2000, pp 252-255 (special issue of South African Computer Journal, no. 26, November 2000) (abstract and full paper)
Streaming vs. Latency in Information Mass-Transit, Computer Architecture News, vol. 26 no. 5, December 1998, pp 4-6. (abstract and full paper)
Design of a Scalable Video on Demand Architecture, Proc. SAICSIT '98, Gordon's Bay, South Africa, November 1998, pp 211-217. (abstract and full paper)
For other areas, see my publications page.
Comments?
If you like any of these ideas or have any comment, please email me: philip.machanick-AT.NO.SPAM-gmail.com.