Methods and apparatus for scheduling, serving, receiving media-on-demand for clients, servers arranged according to constraints on resources

Abstract

A media object is scheduled for transmission between a server and a client. The media object is partitioned into segments of blocks, wherein each block is a unit of media for which a client will wait to receive an entire block before playing out the block, and wherein each segment includes an integer number of blocks. One or more channels on which to serve each segment are determined, and a rate at which to serve each segment is determined. Additionally, a schedule pair for each channel is determined. The schedule pair includes a time at which the client may start receiving on the channel and a time at which the client may stop receiving on the channel.

Description

CROSS REFERENCES TO RELATED APPLICATIONS [0001] The present invention claims is a continuation of U.S. application Ser. No. 09 / 768,843 filed Jan. 23, 2001 entitled “Method and Apparatus for Scheduling, Serving, Receiving Media-on-Demand for Clients, Servers Arranged According to Constraints on Resources,” which claims priority from U.S. Provisional Application No. 60 / 254,514 (Atty. Docket No. 019186-002900US), filed Dec. 8, 2000 and entitled “Method For Media-On-Demand For Clients And Servers With Constrained Resources,” the entire disclosure of which is herein incorporated by reference for all purposes.FIELD OF THE INVENTION [0002] The present invention relates generally to media-on-demand systems, wherein a large number of clients might make requests of the system for a particular piece of media content that is served in real-time, or near real-time, to the clients. In particular, the present invention relates to optimized media-on-demand systems including systems using chain reac...

AI Technical Summary

Benefits of technology

[0021] Advantages of the present invention include providing improved scheduling of media objects that are to be served to clients so that the clients may play out the media objects uninterrupted after a startup latency. Advantages also include improved serving of media objects to clients. Advantages further include improved receiving of media objects served by a server.

Problems solved by technology

A true media-on-demand system is much more difficult to build because the system may have to serve as many different streams as there are active clients, as each client demands different content (media objects) or the same content at different times. Potentially each client will require an independent stream of the media object it requested, where a stream is the flow of data required by that client in order to play out the content from the beginning without interruptions.

This may be acceptable for a small number of users, but may be infeasible if the number of users grows very large for very popular content.

The startup latency, Ts, may include delays such as processing the client's request at the server, propagation delay in the network, and decoding and encoding delays at the client and server respectively.

Another restriction of all the above media-on-demand schemes is that if any of the data is lost, then they either don't play back in full fidelity, or else they have to wait for a complete cycle of that segment before the missing data piece is re-sent.

Thus, it is unlikely that a media object can be decoded in parts when the encoding is applied to the entire media object as a whole.

Images