I think, aside from timing, that Ethernet is already there. The advances in OAM and the ITU taking the technology seriously has promoted it into prime-time. With that said…..timing (sync) is still a biiiiiiiig issue. We’re left with a choice on timing; settle for timing that isn’t as accurate as we’re used to or replace every bit of ethernet gear in the network which is more than 3 years old…..

I’ve seen some really good transport/service solutions recently that are hardware based and are surprisingly inexpensive. If you’re interested we can talk about those

good you visit such events and get an insight. Very interesting! Also good to see the industry slowly turns into something more mature. Not surprised it’s the mobile carriers driving this.

“will come up with a way to solve these problems elegantly without going way over budget” – I enjoy the ironic aspect of this: ethernet is the solution, because it is so inexpensive. And packet networks offer you so much more flexibility. Both is true. But getting more for paying less, shouldn’t we be a bit more suspicious?

IMHO the Ethernet product doesn’t really exist yet, at least not the Ethernet that can replace TDM or ATM. I don’t say it cannot be done and it will happen for sure. But I wonder how much remains in terms of costs savings. Well, that is probably obfuscated with the next upgrade of linecards, i.e. when new ASICs/NPs have to be rolled out anyway.

Do we really wonder when the carriers are reluctant to buy?

While I don’t solve the problem my bet is on the modern NPs that allow a fairly generic programming, allowing a “simple” implementation by code change and yet delivering the performance and accuracy. Normal software solutions are possible but I’ve seen in our code how easy it is to break real-time behaviour despite a proper QNX underneath.

BR

-Antti

http://managedethernetaccess.blogspot.com/2010/03/mobile-world-congress-wants-active.html