While I agree with you that SONET/SDH are here to stay for a long time, I do not
agree with the "encapsulation" idea. While POS uses an encapsulation scheme, it
does not mean that 802.3 needs to use one. Part of the high cost of POS
interfaces is the additional processing that is required for the "encapsulation"
within HDLC/PPP. Even using SDL adds additional processing to the L1
implementation. Directly mapping the 802.3 frames, along with the IPG into the
"payload" will not add any additional processing. I hope that it was the
concept of directly mapping 802.3 frames in the "payload" that you were writing
"Perkins, Drew" wrote:
> >At the same time WANs are architected to transport data traffic more
> >seamlessly. In particular the trend is to move to photonic DWDM networks
> >which are not SONET based. It is these new photonic DWDM networks where a
> >9.584640 G Ethernet would be most important. In the DWDM networks Ethernet
> >would be directly carried, not encapsuled in OC-192. The issue here is one
> >of transition. The current DWDM networks were built for SONET transport.
> >Because of this they will retain some of the SONET character for a number
> >of years before they become completely transparent.
> I must respectfully disagree with you on this. Completely. The trend
> is NOT away from SONET-based networks. The trend IS away from SONET ADMs.
> But the SONET protocol is well designed to meet the requirements in the WAN
> with relative efficiency. if 10 Gb/s Ethernet is used in this environment, I
> think it will be encapsulated in OC-192.
> CIENA Corporation Email: ddp@xxxxxxxxx
> Core Switching Division Tel: 408-865-6202
> 10201 Bubb Road Fax: 408-865-6291
> Cupertino, CA 95014 Cell/Pager: 408-829-8298