Re: IP over DVB-S2

[Lloyd Wood <l.wood@eim.surrey.ac.uk>]

On Tue, 21 Jun 2005, Jérôme Lacan wrote to the ipdvb mailing list:

> With our colleagues from Alcatel-Space, we are working on the
> mapping of IP over DVB-S2.
>
> Recall that DVB-S2 is the "second generation" DVB system for
> broadband satellite service. It will probably replace DVB-S in the
> next few years. You can find the specifications in ETSI document -
> EN 302 307.
>
> One of the main business targets of DVB-S2 is IP packet
> transmission. The mapping/encapsulation of IP over this system
> remains an open research issue. Actually, DVB-S2 proposes 2 modes to
> transport the data. The first one, Transport Stream-TS, is designed
> to transport MPEG2-TS packets whereas the second one, Generic
> Stream-GS, is designed to transport continuous streams (which can be
> compound of variable size packets). Regardless of the chosen mode,
> the resulting data are then mapped on Data Fields (of length varying
> from 16008 to 58192 bits) to be encoded by the FEC (LDPC + BCH).
>
> Several options can be considered for this mapping. We think that
> the ipdvb group is a good place to start a discussion on this
> subject.

One approach to transporting continuous streams is to carry a
continuous HDLC stream; once you come out of your DVB-S2-capable modem
the HDLC stream would be ready to go right into a serial port on any
router for the IP packets to be handled there -- and integrating
modem and router becomes easy.

This HDLC approach has been proposed in the space agency domain for
their CCSDS protocols, as easier than the various existing and arcane
IP-in-CCSDS encapsulations. A new IP-over-CCSDS space links workgroup:
http://public.ccsds.org/sites/cwe/sis-ipo/default.aspx
has been formed to discuss this, in a kind of parallel way to this
IP-over-DVB IETF workgroup.

This sapproach separates the layering quite cleanly. You'll find a
description of the engineering advantages of layering an HDLC stream
over your link in:

K. Hogie, E. Criscuolo and R. Parise, Using standard Internet
Protocols and applications in space, Computer Networks, special issue
on Interplanetary Internet, vol. 47 no. 5, pp. 603-650, April 2005.

...which describes the HDLC-into-modem case (with a number of
practical anecdotes) and proposes the HDLC-into-CCSDS-device case.

One source for a copy of that paper is:
http://www.uni-bamberg.de/wiai/ktr/html/lehre/2005ss/hauptsem/unterlagen/hogie.pdf

If you think of DVB-S2 as, like CCSDS, being a very complex link
layer, this approach makes a lot of sense.

L.

> If you are interested on this subject, could you please manifest your
> interest?
>
> best regards,
>
> Jerome Lacan
> ENSICA/TeSA/LAAS-CNRS
> Toulouse, France

<http://www.ee.surrey.ac.uk/Personal/L.Wood/><L.Wood@eim.surrey.ac.uk>