Networks Part 6: SDH, the great survivor
March 2007
When I first wrote about Synchronous Digital
Hierarchy (SDH) and SONET (SDH is the European version SONET) back in 1992, it was seen to be truly
transformational for the network service provider industry. It marked a clear boundary from just continually enhancing an old asynchronous technology belatedly called Plesiochronous Digital
Hierarchy (PDH) to a new approach that could better utilise and manage the ever increasing bandwidths then becoming available through the use of optical fibre. An up-to-date overview of SDH /
SONET technology can be found in Wikipedia.
SONET was initially developed in the USA and adapted to the rest of world a little later which called SDH. This was needed as the rest of the world used different data rates to those used in the USA - this later caused interesting inter-connect issues when connecting SONET to SDH networks. For the sake of this post, I will only use the term SDH from now on as, by installation base, SDH far outweighs SONET.
Probably even more amazing was that when it was launched, following many years of standardisation efforts, it was widely predicted that along with ATM it would become a major transmission technology. It has achieved just that. Although ATM hit the end stop pretty quickly and the dominance of IP was unforeseen at that time, SDH and SONET went on to be deployed by almost all carriers that offered traditional Public Switched Telephone Network (PSTN ) voice services.
The benefits that were used to justify rollout of synchronous networking at the time pretty much panned out in practice.
-
Clock rates tightly synchronised within a network through the use of atomic clocks
-
Synchronisation enabled easier network inter-connect between carriers
-
Considerably simplified and reduced costs of extracting low data rate channels from high-data rate backbone fibre optic cables
-
Considerable reduction in management costs and overheads compared to PDH systems.
In the late 1990s, as SDH came out of the telecommunications world rather than the IT world, it was often considered to be a legacy technology along with ATM. This was driven by the fact that SDH was a Time Division Multiplexed (TDM) based protocol with its roots deeply embedded in the voice world whereas the new IP driven data world was packet based.
In reality, carriers had by this time had made hefty commitments to SDH and they were not about to throw that money away as they had done with ATM. What carriers wanted was to have a network infrastructure that could deliver both tradition TDM based voice and data services along with the newer packet based services i.e. a true multi-service network. In many ways SDH has been a technology that has not only survived the IP onslaught but will be around for many years to come. It will certainly be very hard to displace.
From a layer perspective, IP packets are now generally delivered using an MPLS infrastructure that was put in place to replace ATM switching. MPLS sits on top of SDH, which in turn sits on top of Dense Wave Division Multiplexing (DWDM) optical fibre. DWDM will be the subject of a future post.
One interesting aspect of all this. is that quite a few carriers that started up in the late 1990s (many didn't survive the telecommunications implosion) looked to a future packet-based world and did not wish to provide traditional TDM-based voice and data services. To this breed of carrier, the deployment of SDH did not seem in any way sensible and they looked to remove this seemingly redundant layer from their architecture by building a network where MPLS sat straight on top of DWDM. This is a common architecture today for a green field network start-up looking to deliver legacy voice and data services purely using an IP network.
A number of 'improved' SDH alternatives sprang up in the late 1990s. The most visible one being Cisco's Dynamic Packet Transport (DPT) / resilient packet ring (RPR) technology. To quote Cisco at the time:
DPT is a Cisco-developed, IP+Optical innovation which combines the intelligence of IP with the bandwidth efficiencies of optical rings. By connecting IP directly to fiber, DPT eliminates unnecessary equipment layers thus enabling service providers to optimize their networks for IP traffic with maximum efficiencies.
DPT never really caught on with carriers for a variety of technical and political reasons.
Another European initiative came from a small start-up in Sweden at the time - net insight. This was called Dynamic Synchronous Transfer Mode (DTM). To quote net insight at the time:
DTM combines the advantages of guaranteed throughput, channel isolation, and inherent QoS found in SDH/SONET with the flexibility found in packet-based networks such as ATM and Gigabit Ethernet. DTM, first conceived in 1985 at Ericsson and developed by a team of network researchers including the three founders of Net Insight, uses innovative yet simple variable bandwidth channels.
Again, DTM failed to gain market traction.
SDH has a massive installed base in 2007 and continues to grow at an albeit steady pace. For those carriers that have already deployed SDH, it is pretty much of a no-brainer to carry on using it, while new carriers who focus on all services being delivered on a converged IP network, would never deploy SDH.
SDH has always managed to keep up with the exploding data rates available on DWDM fibre systems so it will maintain its position in carrier networks until incumbent carriers really decide to throw everything away and build a fully converged networks based on IP. There are a lot of eyes on BT at present!
SDH extensions
In recent years, there have been a number of extensions to basic SDH to help it migrate to a packet oriented world:
Generic Framing Procedure (GFP): To make SDH more packet-friendly, the ITU, ANSI, and IETF have specified standards for transporting various services such as IP, ATM and Ethernet over SONET/SDH networks. GFP is a protocol for encapsulating packets over SONET/SDH networks
Virtual Concatenation (VCAT): Packets in data traffic such as Packet over SONET (POS) are concatenated into larger SONET / SDH payloads to transport them more efficiently.
Link Capacity Adjustment Scheme (LCAS): When customers' needs for capacity change, they want the change to occur without any disruption in the service. LCAs a VCAT control mechanism, provides this capability.
These standards have helped SDH / SONET to adapt to a packet based world which was missing in the original protocol standards of the early 1990s.
A more detailed over view of these SDH extensions is provided by Cisco .
Round up
At the end of the day there seems to be four core transmission technologies that lie at the core of networks: IP, MPLS, DWDM and, if the carrier was a traditional telco, SDH. It will be interesting to see how this pans out in the next decade. Have we reached the end game now? Are there other approaches that will start to come to the fore? What is the role of Ethernet? These are some interesting questions I will attempt to tackle in future posts.
Although considerable advances were taking place at the physical optical layer-1 with the advent of DWDM systems, the existing SONET and SDH standards of the time were limited to working with a single wavelength per fibre, and were also limited to working with optical circuits in the physical layer. They could cope with astounding data rates on a single wavelengths they could not be effectively used with emerging DWDM equipment.
This deficiency led to further SDH standards development initiatives to bring it "up to date". These are known as the Optical Transport Network (OTN) working in an Optical Transport Hierachy (OTH) world. OTH is the same nomenclature as used in PDH and SDH networks. This is covered in Making SDH packet friendly.
The follow on to this post is: Making SDH, DWDM more packet friendly