Swisscom SONATE: Deploying 100G for a Future-ready Network

Swisscom Optical Network for All-IP Transport Services (SONATE)

Swisscom, the largest full-service operator in Switzerland, has 5.6 million mobile subscribers, 3.5 million fixed network subscribers, and 1.8 million broadband subscribers, accounting for over 60% of market share in Switzerland. As one of the top 50 operators in the world, Swisscom has been running in good operational condition over years, and with rapid but stable development.

Challenges Before SONATE

Bruno Sievi, Network Architect Transport, Swisscom

First, the trend towards Ethernet based services and towards much higher capacity. Secondly, the reduction to complexity in network management, although at the same time introducing a new network management system in our current OSS environment. Third, we have a lot of parts which are end-of-life in SDH and WDM, we have to migrate them away without traffic affect to the customers.

Expectations for SONATE

Martin Joerg, Resource Manager SONATE, Swisscom

First of all we expected a highly scalable, end to end solution; from 2Mbit until 100Gbit with a single, highly powerful network management system. A WDM portfolio, built from edge to the Metro and core components, this makes Swisscom network bandwidth have high expandability for the further 10 years. Second, we expected a future proof solution with a clear roadmap and all together with a competitive and transparent price level. Third, from the vendor side we expected a strong global player which belongs to the leaders of the optical transport market, with a powerful local support.

Partnering with Huawei

Stefan Muehlematter, Leader, SONATE Operation, Swisscom

Because Huawei Swizerland is located in Bern Liebefeld, the collaboration is very close. Since the beginning of the SONATE Project, the technical staff on the Huawei Switzerland side remains the same. This brings us a big advantage in communication between the two companies. Huawei's reaction time in case we need technical assistance is very fast. If we recognize complex problems, R&D is involved very early and problems are solved within agreed SLA Time. The cooperation with Huawei is very professional; high knowledge based technical support are also important. They have a very good support by the R&D department. The procedure to bring in service a 100G Wavelength is nearly the same as it is for a 10G wavelength. For this reason we save a lot of time during introduction of 100G.

Regarding documentation and training, we can make use of the operational knowledge. We have done with 10G in the past 3 years. Since October 2012 we have 100G trails in service, so far they are working without any problem.

According to the actual huge traffic demand, and the huge forecast for the coming years, we had to decide either to build new 100G greenfield links, or to extend our existing links with 100G in the brownfield, or also called Hybrid mode.

Our business case shows us some advantages for the hybrid mode; for example, less room space, less power consumption, and last but not least, less fiber costs for the new links. Actually we use a lot of Huawei 100G boards. The planning with the designer tool MDS 6600, together with the guard band tool is easy to handle, and also the operation effort for bringing into service is much faster than with single 10G connections.

From a planning point of view, we're investigating now for the next generation 100G boards with soft decision FEC, to reduce the amount of guard bands, and to reduce the power consumption, and therefore we plan to introduce V1R8 as the next milestone in our successful OTN transport solution called SONATE.

Evolution of SONATE

It's all about scaling, lifecycle and higher capacity. It's not so only that the number of connections will be increasing by a factor of 2 to 4; it's more of a concern that the transported bandwidth will be higher by a magnitude, we assume in addition the increase of the network performance has to be taken into account, also the network management system; we have to look at the performance of the management system itself. Today we've introduced 100G in our network; in 5 to 10 years we assume to have 400G in place.

Beyond 100G

There's a lot of new stuff and not so new stuff that will come up in the next years. SDN concepts, just for example, the separation of data and control plane, that will become an issue. It will be implemented somehow in our network, even if we don't know how.

The integration of OSI layer 1 and layer 2 in one system that will be a trend,

and for sure the integration of 400G that will be an important step. But at the end it always comes down to the cost per bit, so the transport costs should be come down by a factor of 3 to 10.