RE: ip-dvb - PROPOSED CHARTER

["Harald Skinnemoen" <harald.skinnemoen@nera.no>]

I agree with Marie-Jose that it would be beneficial to consider "something"
that would not be tied to specifically MPEG and DVB, but that could support
efficient IP services over satellite across different transmission
technologies.

However, at present, the other options are not generally published
(proprietary technologies), and details of other air interfaces / transport
streams are not generally known as well as the DVB and the DVB-RCS spec.
Therefore, when specifics of the air interface needs to be considered it
makes sense to consider DVB. Perhaps it?s possible to separate those issues
that are general for satellite (or MPEG or ATM), from those that are
specific for DVB-S or DVB-RCS?

I think however, that focusing on DVB simplifies the work, at least in the
short term. It may also make it more directly applicable.

There is of course work in progress to update the DVB-S spec. that need to
be taken into account, where issues like adaptive coding and modulation is
taken into account.

Adaptive coding and modulation is also a characteristic of other satellite
systems (it will probably be important for Ka-band systems where severe rain
fades can occur), and although the exact method may not be generally known
the consequence is that there will be changes in the total channel capacity.
This will in turn translate into either changes in the capacity that can be
given to a fixed number of users, or changes in the number of users (or a
combination of both).

There could be variations also in the requirements for e.g. coding as a
function of the required error rate (in whatever metric used.). Robust
source coding techniques (not implemented in MPEG-2, but implemented in more
recent technology) can tolerate several percent of errors. (However,
limiting factors may be the Internet Protocols, and as such heavy protection
may still need to be used...?)

Issues like changes in channel capacity (due to rain/fading/blocking/...)
may influence the QoS handling, and somehow it would be beneficial if it was
standardized how this could be communicated from the lower layers to the
upper layers independent of the specifics of the air interface. (We are
thinking along these ways in ETSI.)

The following is cut from the scope of the ETSI document bsm07d08, where
work is just starting (the SI-SAP spec. is not completed:

?The present document specifies the Satellite Independent Service Access
Point (SI-SAP) for the provision of standard Broadband Satellite Multimedia
(BSM) bearer services. This interface corresponds to the OSI Layer 2 service
access point between the satellite dependent part of the air interface and a
satellite independent part...

The present document is a common specification that is applicable to any air
interface that complies with the BSM bearer service requirements. The BSM
bearer services are accessed via the SI-SAP and they can be used to
transport a range of standard network services such as:

* IP-based network services. The SI-SAP services are compatible with both
IPv4 and IPv6 packets. The SI-SAP can support the mapping of IP classes of
service based on either diffserv or intserv in addition to the basic
?best-effort? class of service.

* ATM-based network services. The SI-SAP services are compatible with
standard ATM services, including both Permanent Virtual Circuits (PVCs) and
Switched Virtual Circuits (SVCs).

The SI-SAP is a logical interface and the service primitives are defined in
abstract functional terms. The present specification is only intended to
define the functionality of this interface and is not intended to constrain
the implementation of the interface. The SI-SAP may be implemented as a
physical interface (e.g. using message based exchanges); as a logical
interface (e.g. using API function calls); as an internal interface (e.g.
using embedded procedure calls) or in any other format that provides the
specified functionality.?

I also hope we can work together in the IP-over-satellite issues (IETF, DVB,
ETSI).

/Harald Skinnemoen
Chairman ETSI WG BSM


- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
Dr. Harald Skinnemoen
Chief Scientist
Nera SatCom
P.O.Box 91
N-1375 Billingstad
NORWAY

Email : Harald.Skinnemoen@nera.no
Office: (47) 67 24 40 00
Direct: (47) 67 24 36 35
Fax:    (47) 66 85 91 15
Web: www.nera.no

>-----Original Message-----
>From: owner-ip-dvb@erg.abdn.ac.uk [mailto:owner-ip-dvb@erg.abdn.ac.uk]On
>Behalf Of Montpetit Marie-Jose (NVO/Boston)
>Sent: 8. november 2001 19:06
>To: ip-dvb@erg.abdn.ac.uk
>Subject: RE: ip-dvb - PROPOSED CHARTER
>
>
>Beware of the QoS quagmire and be aware that there is a new working
>group (NDIS) part of transport that will look at novel signaling for QoS
>of IP traffic starting with wireless and if we are doing to right thing
>coming up with a framework for all IP traffic over whatever.
>
>I would really like to get out of MPEG2 or even DVB specific approaches
>since compression changes, encoding changes but QOS remains :-) Let's
>start with an architecture of what we the DVB people need in terms of
>QoS and try to coordinate with others.
>
>I also would like the group to make sure that the work done in the DVB
>standardisation group for IP does not overlap too much with what IETF is
>doing.
>
>I think one issue that has to be addressed and maybe a new
>"encapsulation" could have fields for that is the issue of multiple
>gateways. Yes you can get IP traffic from your satellite but also (at
>the same time) from your DSL line, you cell phone, you PDA what not.
>Where to you send what? This goes beyond the current scope and is a real
>problem right now. Probably reaches onto the level 8 (content-based)
>routing.
>
>Another point is that an awful lot of hardware exists right now that
>uses the current MPE stuff. Do we have an idea on how this would degrade
>"gracefully"?
>
>The rest of the charter looks ok.
>
>Marie-Jose
>
>-----Original Message-----
>From: ext Gorry Fairhurst [mailto:gorry@erg.abdn.ac.uk]
>Sent: Thursday, November 08, 2001 3:59 AM
>To: ip-dvb@erg.abdn.ac.uk
>Subject: Re: ip-dvb - PROPOSED CHARTER
>
>
>
>Dear list member:
>
>This list is now reaching an important time, after discussing the
>relationship with DVB, ETSI, some technical issues, etc. We now
>need to agree on a charter - or at least have a firm idea of
>the sort of things we ARE going to do.
>
>Please read:
>
>http://www.erg.abdn.ac.uk/users/gorry/ip-dvb/charter.html
>
>Please send comments as soon as possible, preferably TODAY
>or FRIDAY am.
>
>Is there anything missed out which we should also consider?
>
>Is there anything here which you feel is of no value?
>
>- PLEASE send all comments to this list.
>
>Gorry
>