Synchronous digital hierarchy

from Wikipedia, the free encyclopedia

The synchronous digital hierarchy ( SDH ) is one of the multiplex technologies in the telecommunications sector that allows low-rate data streams to be combined into one high-rate data stream. The entire network is synchronized .


In 1985 the USA began to specify a new generation of optical digital transmission systems under the name SONET ( Synchronous Optical Network ) , which should have decisive advantages over the widespread PDH technology ( Plesiochronous Digital Hierarchy ). For reasons of compatibility, it should also be able to transport signals from PDH technology, but otherwise form a new hierarchy of bit rates. In the USA, 51 Mbit / s was chosen as the basic bit rate and called STS-1 (Synchronous Transport System, Step 1). With this bit rate, the plesiochronous bit rate of 45 Mbit / s could be transmitted. The next multiplex level works with a factor of 3 and delivers 155 Mbit / s (STS-3). It transports three individual STS-1s, so it has three structured information fields that carry the user data . But this is often unfavorable, which is why a variant was defined that has a contiguous field three times the size instead of three fields. This procedure is called STS-3c, where the c stands for “concatenated”. Another concatenation method is called virtual concatenation and serves the same goal: enlarging the coherent information field. It is identified by an appended -vc (virtual concatenation). The introduction of virtual concatenation made it possible to transfer bit rates in steps of n times 2 Mbit / s (e.g. 2M, 4M, 6M, 10M, 20M, 40M, 50M, 100M), as well as high bit rate data signals (such as with Gigabit Ethernet ) to multiplex and transmit without data rate losses.

The international standardization organization ITU-T (Recommendations, G.707) took up the concept of a new hierarchy for digital transmission systems and standardized it under the name SDH. However, in contrast to the North American SONET, the 155 Mbit / s level with the designation STM-1 ( Synchronous Transport Module , Step 1) was chosen as the basis.

The data is transmitted transparently in containers using "Link Connections" and "Trails" through the SDH network. If an SDH network node or an optical fiber fails, SDH network elements can automatically switch the data streams to an alternative route within a few milliseconds ( protection ).

Compared to the previous PDH networks, SDH is equipped with significantly expanded OAM functionalities, ie errors (defects and anomalies) can be recognized more clearly and reported in a more differentiated manner. The interfaces used may have a maximum bit error rate of 10 −10 and even individual bit errors in an SDH signal of any rate can be detected. Overall, SDH networks are designed for the highest quality of service and availability.

SDH networks are increasingly being displaced by DWDM technology, with which a more efficient utilization of the capacity (higher bandwidths, multiple connections by using different wavelengths) of fiber optic cables is possible.


SDH is standardized by the ITU-T (G.707, G.783, G.803). It is derived from SONET (Synchronous Optical Network), which has been developed by Bellcore and AT&T since 1985 . SONET was standardized by ANSI . Today the differences between SONET and SDH are small, the two concepts are interoperable. Since PDH can only be used to a limited extent for broadband ISDN with bit rates above 100 Mbit / s, SDH was primarily designed as a transmission system for B-ISDN. However, it is also suitable for the transparent transport of all relevant payloads ( ATM cells, multiplex signals of the PDH hierarchy, SAN signals, Ethernet aggregation, etc.).

SDH in the layer model
PSTN / ISDN / ATM / IP Application layer
VC-12 layer Low order path
VC-4 layer High Order Path
Multiplex Section
Regenerator Section
Physical interface
  • Physical interface : Usually fiber optic, microwave, or satellite connection
  • Regenerator : Refreshes the damped and distorted signals with regard to timing and amplitude
  • Multiplexer : combine plesiochronous and / or synchronous signals into high-bandwidth SDH bit streams or insert / decouple signals
  • VC (Virtual Container): transport containers with user data. VC-4 layer regulates the integration / removal (mapping) of 140 Mbit / s signals (E4), the VC-3 layer the mapping of 34/45 Mbit / s signals (E3 / DS3) and VC- 12-layer mapping of 2 Mbit / s signals (E1)


SDH is a synchronous time division multiplex process that, like PDH, contains a multiplex hierarchy. The aim is to make the best possible use of the transmission capacity offered by optical fibers. In contrast to PDH, the clocks of the individual transmission links are synchronous with very little deviation. The PDH technology works with deviations of a maximum of 50 ppm, the SDH technology is more than 10 times more accurate. The principle of SDH is simple: the byte streams from n sources at the rate R is its synchronous multiplex to a byte stream to the rate n · R summarized.

In contrast to the PDH, the synchronous operation of the SDH makes it possible to generate a multiplex signal of the order n + 1 directly from the signals of all hierarchical levels 1, ..., n below . Likewise, a lower order multiplex signal can be extracted directly from the frames of higher hierarchical levels. These functions are called add / drop . The synchronous multiplex method also enables bit streams such as ATM cells and PDH multiplex signals to be transported. This function is known as "cross-connect" .

SDH knows the hierarchy levels according to the table. The frames of level n are referred to as STM- n (Synchronous Transport Module-n). The levels STM-1, STM-4, STM-16, and STM-64 are commonly used. SDH reserves around 5% of the gross data rate for OAM tasks (Operations, Administration and Maintenance).

SDH hierarchy levels

SDH hierarchy level
Bit rate
Bit rate
payload / payload
Bit rate
Bit rate
Synchronous (electrical) transport
signal SONET
Optical carrier signal
STM-0 * 51.84 Mbit / s 50.112 Mbit / s 49.536 Mbit / s 1.728 Mbit / s STS-1 OC-1 
STM-1 * 155.52 Mbit / s 150.336 Mbit / s 148.608 Mbit / s 5.184 Mbit / s STS-3 ** OC-3 
STM-2 207.36 Mbit / s  
STM-3 466.56 Mbit / s 451.044 Mbit / s 445.824 Mbit / s STS-9 OC-9 
STM-4 * 622.08 Mbit / s 601.344 Mbit / s 594.824 Mbit / s 20.736 Mbit / s STS-12 ** OC-12 
STM-6 933.12 Mbit / s 902.088 Mbit / s 891.648 Mbit / s STS-18 OC-18 
STM-8 1,244.16 Mbit / s 1,202.784 Mbit / s 1,188.864 Mbit / s STS-24 OC-24 
STM-12 1,866.24 Mbit / s 1,804.176 Mbit / s 1,783.296 Mbit / s STS-36 OC-36 
STM-16 * 2,488.32 Mbit / s 2,405.376 Mbit / s 2,377.728 Mbit / s 82.944 Mbit / s STS-48 ** OC-48 
STM-32 4,976.64 Mbit / s STS-96 OC-96 
STM-64 * 9,953.28 Mbit / s 9,621.504 Mbit / s 9,510.912 Mbit / s 331.776 Mbit / s STS-192 OC-192 
13,271.040 Mbit / s STS-256 OC-256 
STM-128 19,906.560 Mbit / s STS-384 OC-384 
STM-256 * 39,813.120 Mbit / s 38,486 Mbit / s - 1,327.104 Mbit / s STS-768 OC-768 
STM-512 79,626.240 Mbit / s STS-1536 OC-1536 
STM-1024 159,252.480 Mbit / s 153,944 Mbit / s - STS-3072 OC-3072 

The levels marked with * are included in the standard. The hierarchy levels marked with ** in SONET are the most widespread. STM-1 can be designed with electrical or optical interfaces, STM-4 and higher only with optical interfaces.

SDH network elements

Basic types of network elements in SDH multiplex technology are defined as follows:

  • REG ( regenerators ) amplify optical signals. The difference between the purely optical amplifiers used in the SDH-based OTN is the conversion of the received optical signal into an electrical one. Only the electrical signal is amplified, synchronized in time and corrected in its form. Then it is converted back into an optical signal and sent. The purely optical amplifier does not need to be converted into an electrical signal.
  • TM ( Terminal Multiplexer ) usually have several plesiochronous subscriber interfaces and one or more interfaces to the SDH network. They combine tributary signals that come from hierarchically subordinate network elements or from terminal devices to form an aggregate signal of an SDH hierarchy level, for example STM-1, which is forwarded to the SDH network.
  • ADM ( add-drop multiplexers ) are an extension of the terminal multiplexer. They have two interfaces on the aggregate side for SDH signals of the same hierarchy level. An ADM can split the received signals from the two aggregate interfaces into the partial signals contained therein and feed some of them to the corresponding tributary interfaces (drop), but pass the signals through otherwise unchanged between the two aggregate interfaces. In the opposite direction, the ADM inserts signals that arrive at the tributary interfaces back into the aggregate-side signals instead of the extracted partial signals (add). SDH networks in ring topology require ADMs, TMs cannot be used in rings.
  • Cross-connect multiplexers or DCS (Digital Cross-Connect System) (also called DXC in Europe) are, in turn, extensions of the ADMs. They have at least 4 interfaces on the aggregate side and can interconnect the partial signals taken from these or the signals arriving at the tributary interfaces on the VC level as required.

Functional model of the SDH

SDH contains functions that can be assigned to OSI layer 1. The functional blocks and their layering are identified by the following terms:

  • Optical sections (photonic) relate to optical signals on glass fibers and conversions optical - electrical and vice versa.
  • Regenerator section refers to a fiber optic section which is arranged between regenerators (REG) or between a regenerator and another network element. The regenerator section is assigned to the RSOH.
  • Multiplexer section connects two multiplexers (also across several regenerators). The multiplex section connects two terminating ports of the same rate STM-N. The MSOH is assigned to the multiplex section.
  • HO path (high order path or trail) can be transmitted across multiple network elements (e.g. via ADM, DCS and regenerators) (without resynchronization). As a signal mapped to AU4, it contains a VC4 (or a concatenation of VC4 containers, for example for ATM data signals) with a payload data signal of rate E4 or serves as a transport layer for LO paths. The VC4-POH is assigned to the HO path. Furthermore, there are also HO paths of rate VC3 when these are mapped into an AU3.
  • LO path (Low Order path or trail) of the rates VC11, VC12, VC3 are packed in a VC4 and transport the actual user data signals with bit rates equivalent to DS1 to E3. The VC11 / 12/3-POH is assigned to the LO path.

These layers are characterized by their own OAM functions (e.g. transmission error monitoring, alarming, protection) that function independently of the higher-level transmission layer. For example, the bit error rate can be measured at HO level without having to resort to data from the multiplex sections. In the opposite direction, however, if the higher layer fails, the subordinate layer is assigned an error signal, ie if a multiplex section fails, all HO paths and LO paths contained in it are discarded.

Topology of SDH networks

Example of an SDH ring structure

In most countries, the transport networks have now been expanded using SDH technology and the old PDH technology has largely been replaced. Therefore, topologies of various types are implemented, they are based on the geographical requirements. An essential feature of SDH technology is the automatic switch to alternative routes in the event of a fault ( protection ). The double ring is often chosen as an example to explain how the protection works; in undisturbed operation, a ring is used, the so-called work path. The second ring serves as a cold reserve , as an alternative route. Byte streams are introduced into and removed from the work path by the ADM ( add-drop multiplexer ). If the route to work is disturbed, the APS (Automatic Protection System) switches from the route to work to the alternative route. This topology is standardized under the designation 4-fiber MS-SPRing (Multi-Section-Shared-Protection-Ring) from STM-16 upwards.

A simplified version of the ring protection is known as 2-fiber MS-SPRing, where half of the available bandwidth is kept free for alternate route switching or filled with low-priority traffic. In the event of a fault, this bandwidth is loaded with the traffic of the failed ring link and the lower priority traffic is discarded.

MS-SPRing mechanisms are only suitable for ring structures and are therefore particularly applicable in backbone structures. The MSP (Multiplex Section Protection) protocol was developed for linear structures, where a replacement connection usually protects exactly one fiber connection (1 + 1). Further developments occupy the substitute circuit with low priority traffic (1: 1) or protect several multiplex sections with a substitute path (1: N). These protocols work at the multiplex section level, ie the equivalent circuit is used for the entire optical fiber.

Path-based sub-network connection protection , which offers 1 + 1 protection at the VC level , is ideal for heavily meshed structures .

What all these protective mechanisms have in common is that, according to the standard, the backup switching measures must be completed automatically within 50 milliseconds after a fault is detected. In modern SDH devices, however, the switching times actually achieved are significantly lower (depending on the cable length / propagation delay, around 1 millisecond per 200 kilometers).

As a rule, all connections 1 + 1 are protected in today's SDH transmission networks.

Frame structure and multiplex structure

Structure of the STM-1 frame
Lines 9 columns (1 to 9) 261 columns (10 to 270)
Section Overhead (RSOH)
261 * 9 bytes per frame (150.336 Mbit / s)
4th AU Pointer
(Administrative Unit)
Section Overhead (MSOH)
1 byte each 1 byte each

SDH transmits useful and control data in a sequence of frames, which are sent serially. Each frame consists of overhead (control data) and payload (user data and other data). The STM-1 frame consists of the payload, RSOH ( Regenerator Section Overhead ) and MSOH ( Multiplex Section Overhead ) and AU pointer areas . The frame is transmitted line by line from left to right and from top to bottom. The AU pointers (administrative unit) point to the position of the useful information in the payload area.

The terms for frame structure are defined as follows:

  • Container (Ci) Areas in the frame that correspond to a specific payload. The size of the container was adapted to the data rates defined in the PDH . The insertion of plesiochronous data streams requires stuffing processes (bit or byte synchronous). The POH (Path Overhead), which describes the user data, is added to each container.
  • Virtual containers (VC-i) are divided into lower order VC (VC11 to VC12, VC2 and VC3) and higher order VC (VC-4). Some lower-order VCs can be combined to form higher-order VCs, but do not have to.
  • Tributary Units (TU-i) are required because the VC coming from outside the SDH can have different phase positions with respect to the multiplex frame. Therefore the VC are embedded in the slightly larger TU. The start of a VC within a TU is indicated by pointers.
  • Tributary Unit Group (TUG) summarize TU-i according to the diagram.
  • Administrative units (AU-i) have the same function in relation to higher-order VC as the tributary unit group in relation to lower-order VC.
  • Administrative Unit Groups (AUG) are formed from AU-3 and AU-4 in the same way as the Tributary Unit Groups. The associated pointers are the AU pointers in line 4 bytes 1–9.
  • Synchronous Transport Module (STM-n): Higher-order frames ( ) are formed by multiplexing a corresponding number of frames from the next lower hierarchy level.

The introduction of the pointer allows (in contrast to the PDH) the direct addressing of a user data signal in a high bit rate signal without having to demultiplex the complete signal. Furthermore, small clock differences between the network elements can be compensated for using pointers.

SDH multiplex structure according to ITU-T G.707
STM-n ← AUG ← AU-4
pointer treatment
← VC-4
POH Path Overhead
← C4
149.760 Mbit / s (ATM signal)
139.264 Mbit / s (E4 signal)
← TUG-3 ← TU-3
pointer handling
← VC-3
← C3
48.384 Mbit / s (Ethernet signal)
44.736 Mbit / s (DS3 signal)
34.368 Mbit / s (E3 signal)
← TUG-2 ← TU-2
pointer handling
← VC-2
← C2
6.312 Mbit / s
3 * 2.048 Mbit / s
← TU-12
pointer handling
← VC-12
← C12
2.048 Mbit / s
← VC-11
← C11
1.544 Mbit / s

Note: ATM signals can be mapped directly into a C4 with a transmission rate of around 150 Mbit / s (with DSLAM).

Adaptation of the AU pointer

The administrative unit pointer can be adjusted at any time. The following situations are responsible for this:

  • Virtual containers are not bound by frame boundaries.
  • Under certain circumstances "wandering" virtual containers (VC)
  • In every fourth frame, pointers can be adjusted after notification.
  • The pointer structure may be chained (transport groups contain containers, etc.).

Structure of an STM-N signal

Structure of an STM-N signal

Standards according to ITU-T

  • G.707 / Y.1322 "Network Node interface for the synchronous digital hierarchy (SDH)", defines the bit rates, the basics of the multiplex structure and the signal structure for SDH at the network node interface (NNI)
  • G.780 “Vocabulary of terms for synchronous digital hierarchy (SDH) networks and equipment”, a glossary
  • G.783 "Characteristics of synchronous digital hierarchy (SDH) equipment functional blocks", defines the function blocks belonging to SDH in the form of information models
  • G.784 "Synchronous digital hierarchy (SDH) management" describes the operating technology belonging to SDH
  • G.803 "Architecture of transport networks based on the synchronous digital hierarchy (SDH)"


The LAPS (Link Access Procedure SDH) transport protocol was developed to map packet-oriented IP data directly into an SDH container. ITU-T X-85 defines IP over SDH and ITU-T X-86 defines Ethernet over SDH using LAPS.

Next generation SDH

The SONET / SDH was created to optically transmit voice and data traffic at higher transmission rates. The user data of the container is therefore defined to be downwardly compatible with the data transmission rates of the PDH hierarchy. The original idea was that the data traffic from IT equipment would also initially be transmitted electrically using a common PDH bit rate such as 2 Mbit / s ( E1 ), and that this would then be combined with other PDH signals in an SDH multiplexer optical SDH aggregate signal is multiplexed. This method is still common today, but at higher data rates the unused part of the transmission capacity is high: for example, an STM-1 signal with 155 Mbit / s is required for the data transmission rate of Ethernet traffic of 100 Mbit / s.

In order to transmit voice and data efficiently via a common platform, the GFP protocol, virtual chaining (VCAT) and granular adding or removing capacity (LCAS) have been defined at the ITU. These extensions of the conventional SDH are called Next Generation SDH.

General Framing Procedure

With the GFP protocol (ITU-T G.7041), Ethernet frames and frames from other common network technologies (Fiber Channel, ESCON, FICON, GbE, digital video) are mapped into the SDH container using GFP mapping. Two modes are defined: Transparent GFP (GFP-T) and Frame-mapped GFP (GFP-F).

Virtual concatenation (VCAT)

However, because the defined SDH container sizes for the transmission of data packets were not optimal, the “virtual concatenation” ( ITU-T G.707 ) of several containers (VC12, VC3 or VC4) was also introduced. This results in a correspondingly larger payload. For Fast Ethernet, only two VC3 are required instead of one VC4. The advantage of virtual concatenation: the individual containers are transported separately through the network, the hardware only needs to be adapted to the new functionality at the network edges - in contrast to "contiguous concatenation".

Link Capacity Adjustment Scheme (LCAS)

Using the LCAS protocol (ITU-T G.7042), individual virtual containers can be switched on or off during operation, so that a quasi-dynamic change in the transport capacity in the network with a relatively short response time and without operator intervention (for example in the event of faults in the Network) is possible. This means that, for example, connections (Ethernet over SDH, ...) can be split over two paths (50/50), so that if one path fails, the connection continues to function, albeit with a reduced / half bandwidth. A protective function using LCAS has the advantage over other methods such as SNCP that no additional transmission capacity is required (with SNCP, double the bandwidth is required - main and alternative paths each with the full target bit rate).

Future of the NG-SDH and NG-SONET

GFP and LCAS enable the SDH to transmit packet data cost-effectively without loss of bandwidth. However, 50% of the transmission capacity is required to protect the SDH links, which is unfavorable in terms of price. Restoration using GMPLS allows the SDH to use the high-speed lines (STM16 or STM64) more effectively. With restoration (shared mesh), an alternative route is calculated dynamically in advance; several routes share a replacement route. The NG-SDH is competitive with the IP / MPLS and Ethernet networks in wide area networks.

In the optical testbed VIOLA in Germany, the latest optical network technologies such as Ason-GMPLS and the Next Generation SDH are tested.

Multiservice platforms

IP-enabled NG-SDH network nodes that use SDH or WDM as a transport network are referred to as MSPP (Multi Service Provisioning Platform) or MSTP (Multi Service Transport Platform).


In October 2005, the first multifunctional platform was presented at the Broadband World Forum in Madrid, which combines a 100% mix of Ethernet / MPLS, SONET / SDH, and WDM / OTN in a single device. The Alcatel-Lucent 1850 Transport Service Switch no longer differentiates between packet-oriented (IP) and circuit-switched services. It transports data independently of the service.

See also

Web links

Individual evidence

  1. Network Node interface for the synchronous digital hierarchy (SDH)
  2. Fibercomm 2005 in Munich (PDF; 68 kB)