1 / 16

FCAST: Scalable Object Delivery on top of the ALC Protocol

< draft-roca-rmt-newfcast-00.txt > IETF 68 th – Prague meeting, March 2007 Vincent Roca (INRIA). FCAST: Scalable Object Delivery on top of the ALC Protocol. If you missed the San Diego meeting…. initiative motivated by Nokia’s patent on Flute patent granted in 2004…

sian
Télécharger la présentation

FCAST: Scalable Object Delivery on top of the ALC Protocol

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript


  1. <draft-roca-rmt-newfcast-00.txt>IETF 68th – Prague meeting, March 2007 Vincent Roca (INRIA) FCAST: Scalable Object Delivery on top of the ALC Protocol

  2. If you missed the San Diego meeting… • initiative motivated by Nokia’s patent on Flute • patent granted in 2004… • but we (Flute co-authors and RMT group) only discovered it at the end of 2006 • today’s goals: • have a deeper look at the patent • see if (hopefully non patented) alternatives are feasible • decide if it’s worth the pain

  3. I hereby declare that I haven’t tried to patent any technique that is described in <draft-roca-rmt-newfcast-00.txt> and that I am not aware of any patent or other IPR claim related to this I-D.

  4. Outline • a quick look at the patent • a little bit of history, back in 2000 • the proposal (see the I-D for more details) • discussion

  5. A quick look at the patent • Nokia’s IPR disclosure https://datatracker.ietf.org/public/ipr_detail_show.cgi?ipr_id=733 • licensing conditions (INRIA/Nokia exchange, November 2006) • "INRIA is free to do what they want for their software, but if Nokia patent is used for commercial purpose, then a proper license is required. Nokia is committed to provide the license on FRAND (fair, reasonable and non-discriminatory) principles."

  6. A quick look at the patent… (cont’) • administrative aspects • priority date: January 31, 2003 • i.e. a few days before Toni submitted the “File Delivery over ALC” I-D • granted on August 2004 • title: DATACAST FILE TRANSMISSION WITH META-DATA RETENTION

  7. A quick look at the patent… (cont’) • claim1 (claim7 is similar): 1. A device for receiving datacast information comprising: […] - means for arranging the object packets into at least a first object and a second object based on the object packet's transmission object identifiers; - means for extracting meta-data from the first object pertaining to the second object as identified by its transmission object identifier; and - means for storing the second object in non-volatile memory as a file having one or more properties identified by the extracted meta-data. It’s Flute’s basic principle: an FDT Instance object contains the meta-data associated to files that are sent in separate objects…

  8. A quick look at the patent… (cont’) • other claims build on this idea • what can we do then? • question the validity of the patent, showing prior art? • design something different that does not use claim1 (and 7)? • it’s the purpose of this I-D…

  9. A little bit of history, back in 2000 • the first ALC I-D “Asynchronous Layered Coding: A scalable reliable multicast protocol”, <draft-ietf-rmt-pi-alc-00.txt> • from: Luby, Gemmell, Vicisano, Rizzo, Crowcroft, Lueckenhoff • published on March, 2000: • of particular interest • A.3. File Transfer using ALC - "Fcast“ • principle: append meta-data to the file’s content

  10. A little bit of history, back in 2000… (cont’) • object format: • the trailer contains MIME fields: • Content-Location • Last-modified • Content-Length • Content-Encoding… +----------------------------------------------+ | Object | 4-byte checksum | Null padding| +----------------------------------------------+ / \ | ........... | \ | .......... / \ +--------------------------------------+ | File data | Trailer | Trailer Length | +--------------------------------------+

  11. A little bit of history, back in 2000… (cont’) • BTW, there’s nothing new in Claim1 of Nokia’s patent… • in <draft-ietf-rmt-pi-alc-00.txt>, A.3., it is said: • is it sufficient to say there’s prior art? • I’m not a lawyer, so… When the object being transmitted is a file, the receiver usually requires "metadata" in addition to the file itself, such as the file name, its creation date, etc. Metadata can be sent a number of ways. For example, it could be part of the object session description or sent as a separate ALC object with a well-known object transmission label. The method described here combines the file with its metadata in a single ALC object.

  12. The proposal • follow-up of Fcast <draft-ietf-rmt-pi-alc-00.txt> • the initial Fcast has a major limitation: • extracting meta-data requires to download and decode the whole object, which may be: • costly (power consumption) • impossible (not enough memory to decode the object) • especially if the client discovers he cannot process the object (e.g. non supported format)

  13. The proposal… (cont’) • two complementary but optional ways to carry meta-data in new Fcast: • append meta-data to the object (idem) • add a dedicated EXT_OMD (Object Meta-Data) header extension to the ALC/LCT packets • the EXT_OMD can be processed immediately, and the client may decide whether or not he needs the object

  14. The proposal… (cont’) • in practice: • put only a subset (e.g. Content-Length/Content-Encoding) in the EXT_OMD, the rest being appended • with small objects, no need to use EXT_OMD • send EXT_OMD in the first few packets sent for a new object, and then periodically • add an Fcast session description object • it describes the session’s content, not the objects • lists the objects (i.e. their TOI) that are part of the current carousel instance

  15. Discussion • what’s next? • Option 1: is the patent valid? • NB: I’m not a lawyer… • Option 2: we don’t care about patents and leave everything as it is… • NB: I don’t like it but you might disagree…

  16. Discussion… (cont’) • Option 3: we continue along the lines sketched here (or something else if somebody has better ideas…) • NB: this proposal can be extended to work on top of both ALC and NORM… We (INRIA) did it in the past… • NB: thanks’ to the experience gained with Flute, it should not take too long • NB: even if Flute is included (definitively) into several standards (3GPP, DVB), there are use-cases where having a brand new application is not an issue • e.g. file distribution within clusters, grids, servers, etc.

More Related