Chris Trown CSCI 198c Advanced Networking Presentation Notes for SONET/SDH I) INTRO SONET(Synchronous Optical NETwork) or SDH(Synchronous Digital Hierarchy) as it's known in Europe, is a set of standards for interfacing Operating Telephone Company(OTC) optical networks. They are a set of global standards for interfacing equipment from different vendors(One of the few where telephony is concerned). SONET is the protocol for North America and Japan while SDH is the definition for Europe. The differences between SONET and SDH are slight. At the current time SONET/SDH are being implemented for long-haul traffic, but there is no reason it can't be used for shorter distances. Indeed, AT&T has just announced a version of SONET to be used over copper for runs no more than 100 meters in length. Synchronous networking is a concept where all the clocks driving a network are running at the same speed. The reality is that in a large geographic-ly distributed network, this is virtually an impossible task. The protocols do allow for a certain amount of drift. II) HISTORY We go back to the mid-80's where single-mode optical fiber is fast becoming the medium of choice for long-haul, high-speed digital transport. Proprietary equipment was starting to make an appearance. With that in mind, Bellcore proposed a set of standards in 1985 to standardize carrier-to-carrier interfaces. CCITT expressed interest in what the T1 committee was up to in Summer of 1986. After much negotiation and compromise an initial set of standards was approved by CCITT(Now ITU). The main points of contention arose because of the differences in the asynchronous digital hierarchies between North America and Europe. The North American(and Japanese) T1 runs at 1.544 Mb/s while the European E1 runs at 2.048 Mb/s. The SONET standard was approved in 1988 by the CCITT plenary and in late 1989 by the ANSI T1 committee. III) DIFFERENCES The differences between SONET and SDH are slight. There are some differences in the basic frame format, but SDH and SONET are identical beyond the STS-3 signal level. The base signal for SONET is STS-1 and the base signal for SONET is STM-1. STS-3 is equivilant to STM-1 and the lower tributaries can be mapped inter- changeably between the two format from that point on. IV) ADVANTAGES Aside from the fact that proprietary protocols for fiber based digital transport have essentially been banished(Proprietary protocols rot the brain -- me), there are a few others: o SONET is based on the principal of direct synchronous multiplexing. Essentially, separate, slower signals can be multiplexed directly onto higher speed SONET signals without intermediate stages of multiplexing. o Provides advanced network management and maintenance features. nearly 5% of the total bandwidth is devoted to this. o Both SONET and SDH can transport signals for all the networks in existence today and it has the flexibility to accommodate any networks defined in the future. The benefits to this are obvious. o Can be used in the three traditional telecommunications areas: long-haul networks, local networks and loop carriers. It can also be used to carry CATV video traffic. V) TERMS Before continuing, some terms must be defined. This will provide some clarity later on. Framing is a method of indicating where to begin counting channels so the demultiplex-er knows where to start counting channels. A certain pattern of bits are transmitted with each frame that makes a pattern that is hard for data to imitate. A tributary can best be described as a smaller pipe feeding into or out of a larger one. A T1 signal feeding into a T3 for example. Plesiochronous Digital Hierarchy(PDH) from the Greek "plesio" meaning almost. Before SONET, and in some current installations today, multiplexers have a difficult task. When a multiplexer is trying to multiplex multiple signals onto one data stream, it has to take into account the fact the the clocks on all incoming tributaries are not exactly the same. There is a certain amount of drift. Trying to get all the signals in sync is virtually impossible. What the PDH multiplexer does is read data from all the incoming data streams at the highest allowable speed. When there are no more bits in the data buffer because the data is arriving according to a slower clock, the multiplexer stuffs the data stream with bits to make up the difference. The multiplexer has a means of notifying the receiving end that stuffing has taken place so the extra bits can be thrown out. OAMP is an acronym that stands for Operations, Administration, Maintenance and Provisioning. STS is another acronym which stands for Synchronous Transfer Signal and is the electrical specification for the various levels of the SONET hierarchy. STM stands for Synchronous Transfer mode and is the SDH equivalent of the SONET STS hierarchy. VI) THE PROTOCOL The SONET protocol stack consists of 4 layers. Each layer can communicate with layers above and below it. To facilitate the flow of information, each layer can communicate directly with it's counterpart on the other end. While the stack looks like the OSI model, all it concerns itself with is the frame itself. The 4 layers are: Photonic layer - This layer concerns itself with converting electrical signals to optical ones. Section layer - Deals with the transport of STS-n frames across the physical medium. Functions include framing, scrambling, section-error monitoring and communicating and adding the section layer overhead. Line layer - Deals with the transport of the path layer payload and it's overhead across the medium. It provides synchronization and multiplex- ing for the path layer. A line is the medium required to transmit data from the originating equipment to the terminating equipment. It is possible for a payload to go through many of these devices on it's way from point A to point B. Path layer - Deals with the transport and mapping of services between path terminating equipment. These services include but are not limited to: DS1, DS3 and video. It's responsible for mapping these services into an STS frame. VII) THE BASIC FRAME There are 3 basic parts to each frame: The section overhead, the line overhead and the synchronous payload itself. Within the payload is the path overhead. The section overhead contains information required for section to section communication. That is, repeater to repeater communication. The section overhead contains data for framing, performance monitoring and a voice channel for maintenance personnel and a channel for OAMP. The line overhead contains information required for line termination equipment communication such as an Add/Drop terminal. It also contains the payload pointer and has data for OAMP, line performance monitoring and another voice channel for maintenance personnel. The synchronous payload contains the actual information being transmitted. Contained within it is the path overhead. It is carried within the payload because the only time it is created or looked at is when the payload enters or exits the SONET network. The path overhead provides end-to-end monitoring of the payload and it's performance during it's journey across the network, makes sure that the correct connection was made, identifies the payload type and provides a user channel for the service providers information. VIII) HOW DOES IT WORK? The problem with PDH is that inserting or removing a lower level data stream from a higher order one, removing a T1 from a T3 for example, requires that the multiplexer perform all the functions of the three multiplexer that created the T3. This operation is called Add/Drop, it's a complex task and the equipment to do it is expensive because of all it has to do. Another problem is that it is difficult to perform management, performance monitoring, etc.. on the network created by these multiplexers. SONET is a new way of multiplexing tributary signals onto a higher order one. It has mechanisms for dealing with tributaries that are not running at the same clock rate. SONET/SDH transmits 8000 frames per second. That's one frame every 125 microseconds. See Fig. for what is transmitted. Pictured is an STS-1 frame. It should be noted that the difference between the various STS frames is the width of the frame. They are still composed of 9 rows, but the width of the row changes depending on the STS specification. The bytes are transmitted one at a time, from left to right, starting with the top row. The secret to making SONET work is the payload pointer. The tributaries coming into a multiplexer may have been created with a clock running at a different speed. They are not necessarily aligned with each other or with the clock in the multiplexer. To resolve this problem, remember that this is a SYNCHRONOUS network, the SONET multiplexer finds the beginning of a frame for each tributary. It then calculates a pointer designating where in the STS-1 frame it has placed the beginning of the tributary frame. This way, it is not necessary for the multiplexer to either get the signals in sync(impossible) or stuff the frame with bits. If a tributary signal's clock slips over time with respect to the multiplexer's clock, the SONET multiplexer simply recalculates the pointer. This is a description of how a SONET multiplexer creates STS-1 frame. It differs from creating a STS-3 or above frame in that since there are no slower definitions for SONET, all the tributaries are usually PDH encoded and therefore not synchronous. Creating a STM-3 or faster frame generally involves multiplexing slower frame types(usually STM-1) into the larger frame. The technique for multiplexing STM frames onto a larger(faster) frame is different. Each byte from the incoming tributaries is interleaved byte-by-byte into the larger frame. Each byte of the tributary signal, and thus the tributary signal itself is visible in the frame. This makes it relatively easy to pull a tributary signal out of the main signal as compared to de-multiplexing a PDH signal. For example, an STS-12 signal is formed by byte-interleaving 12 STS-1 frames. Each byte is separately visible. The various STS-1's could be carrying different types of traffic(voice, data...) and can be heading to different destinations. IX) VIRTUAL TRIBUTARIES Mapping a DS3 onto a STS-1 signal is relatively easy because of the compatibility in speeds of the two protocols. In reality, most users do not have access to DS3 signals. For this reason, the protocol designers created something called Virtual Tributaries, or VTs. Four sizes have been defined, designed to match the speeds of T1, T2, E1 and DS-1C speeds. Each VT is assigned a chunk of the actual payload. The size depends on the size of the VT. There are extra bytes available in the actual payload to allow for synchronization. These VTs can operate in two modes: Floating mode - This mode is the most efficient for transporting VTs. A pointer is used to define where, within the VT container, the actual data is. This makes synchronization easier. This mode is more complex, however. Locked mode - In this mode the data within each VT container is fixed. That is, it's always in the same place, there are no pointers. This is a much simplex interface. The problem is that it is slower due to the time necessary to synchronize the signals. X) CONCLUSION SONET is here, now. It will continue to penetrate the market through upgrades and retirement of existing equipment. It is an international standard that is being widely adopted. One reason it is so important is that SONET has been selected as transmission technology for BISDN. It can transport all signals currently defined in the world today.