Re: [EFM-P2P] RE: [EFM] T.V. broadcast / unicast
Ingvar,
I've sent a document with our requirements to the group some time ago.
They're located at:
http://www.ieee802.org/3/efm/public/email/msg00540.html
At this document, I've tried to outline several options to provide voice,
video and data services using an EFM approach. Here follow a short resume
of the main alternatives for voice and video:
Voice service:
a) Voice over TDM (using sideband modulation)
b) Voice over packet, with three options:
b.1) Voice over TDM emulation, transparent mode.
b.2) Voice over TDM emulation, advanced signaling.
b.3) Voice over IP (soft switch approach).
Video service:
a) Analog video overlay, using a WDM approach (separate wavelength for
video).
b) Digital video over packet. This is the preferred approach. That is why I
believe that single-copy-broadcast is *needed* for EFM.
For the video delivery, we devised a VLAN-based approach to implement
channel segregation. At that time, we felt that IGMP snooping would not be
a good alternative to solve the selective channel distribution problem,
based on the assumption that the complexity, and thus the price, would be
too high. By using VLANs, we expected to be able to selectively deliver
only some channels of video to every customer, and then - from the ONU to
the settop box, where bandwidth is at a premium - we would use the same
system to deliver only a single TV channel. Now we see that IGMP snooping
may scale well enough, but we still count VLAN (with some advanced
membership protocol, such as GARP) as a viable alternative.
There are also other comments, and we could discuss this extensively. I
believe that this is more on the focus of what we are calling 'Ethernet
Subscriber Access', or ESA for short - the integrated solution to bring
services to the market, using Ethernet based technology.
Carlos Ribeiro
CTBC Telecom