The marketization and politicizing of science can result in dire consequences, such as in environmental matters. While the ramifications may be less important in communications, such as the meaning of Ethernet being a moving target over time or with Luxtera being the only vendor to provide actual silicon photonics, albeit only for parallel singlemode, stretching the definition of basic scientific terminology, at a minimum, results in unnecessary confusion in the industry. In a Light Reading post in June, Orange appears to be discussing “all-IP” in a marketing context, not in a technical one.
It is almost as if the headline could have been, “Stop the Presses – Orange Discovers Massive Number of Synchronous Optical Elements in its Network.” Of course, even with past projections by service providers years ago about moving to “all-IP,” which turned out to be inaccurate, there were SDH and ATM equipment in the infrastructure. So, we are left with more questions for Orange, including how exactly does it define “all-IP,” than with answers.
For example, IP is not deterministic. How can services with strict SLAs work on such an exclusive network?
Second, if Orange were to decide “to ‘do nothing’…in emerging markets where SDH technology is still relatively new,” how does it get to “all-IP” by the 2020 timeframe?
Third, in switching to OTN, how is it IP? Like SDH and ATM, OTN is a TDM, circuit-based technology. Furthermore, it goes without saying that all of the E1s do not magically turn into IP by moving to OTN.
To be sure, certainly some liberties are being taken in that article. In general, the piece does not seem to be very logical. Is it possible that Orange meant to stress more of a “programmable” network approach? If so, why has not the operator made an attempt to correct the record or to add a greater amount of context including with “its Essentials2020 transformation initiative”?
Despite executives at carriers tending to be only concerned about the next six months, there could also be the realization at Orange that 2020 for an “all-IP” goal may ultimately look a little silly down the road. Therefore, the decision may have been to start using its SDH installed base as a potential scapegoat. It may also be a trial balloon to see if the market buys OTN as “all-IP.”
Interestingly, Verizon, which is moving in an OTN direction, has apparently never provided any prognostications for “all-IP.” Then again, Verizon’s top leadership in engineering tends to be more upfront about network evolution, such as with its negative stance on open line systems.
Obviously, a service provider opting to go with MPLS-TP rather than OTN, such as NTT, would have an easier time in making the case for having “all-IP” technology. Nevertheless, we are unaware of NTT doing so since its prediction several years ago turned out to be quite aggressive.
In conclusion, the industry may be left with a similar situation as it has with silicon photonics. It is more important for some companies to say they have achieved “all-IP,” for marketing purposes with its supposed cost advantages, etc., ostensibly increasing its market worth than to stick to a strict definition.
[written by Mark Lutkowitz]