SDH Networks - An introduction

Synchronous Digital Hierarchy ( SDH Networks)

Digital  Telephony Networks evolved from  Pulse Code Modulation (PCM) & a multiplexing TDM hierarchy called Plesiochronous Digital Hierarchy (PDH). PDH multiplexing had an advantage of higher complexity and cost both as for dropping any signal from the bit stream at the ADD – Drop node each time the entire signal had to be De-multiplexed  - Multiplexed. Also accommodating different kind of traffic such as ATM, Ethernet etc was a problem.

SDH is synchronous implying that a distributed reference clock determines the basic bit frequency in the network.  Multiplexing, grooming, cross connecting etc can be done without buffers unlike PDH.

-          Multiplexing forms higher bit-rate channels out of lower bit rate channels typically in order to achieve more cost-effective transmission.

-          Grooming is done to fill up (in order to reduce the number of) higher bit rate channels in the network.

-          Cross connecting is needed for connection setup, protection, restoration and grooming, among other functions.

SDH networks are optimised for 2 Mb/s and 140 Mb/s tributary channels but can also be used for 1.5, 34 and 45 Mb/s channels.

These channels are then synchronized to the network and put in virtual containers (VC-11, 12, 3, 4)

Channels below VC-4 are called lower-order paths while VC-4 is called higher-order path.

SDH add/drop multiplexers (ADMs) and cross connects (DXC) exist for different path levels.

DXCs are labelled with m/n with m the highest access level and n the lowest cross connection level. Typical DXC are 4/1 or 4/4.

For transport, one or more VC-n’s are put into transmission frames called Synchronous Transfer Modules (STM). These frames or transport modules come with bitrates in multiples of 4: STM-1/4/16/64/256 at 155/622/2488/9953/39813 Mb/s, according to the current standard.

STM -1 Electrical versions exist for STM-1 while the rest are only defined for optical transmission. In reference to the DXC labeling above, an OXC with port-speeds and cross connection at STM-16 would be called a 6/6 DXC (with the difference that the OXC does not alter the overhead. Thus, there’s no fundamental difference in moving from a 4/1 DXC to a 4/4 DXC compared with moving from a 4/4 DXC to an OXC handling STM-16 channels.

The OAM features of SDH are extensive including QoS parameters, in-band management and identifiers for all channels.

Since all the features of SONET and SDH is the same in the context of this text, the term “SDH” is collectively used for both SONET and SDH to avoid writing SONET/SDH whenever used.

The main clock is normally (dual) redundant. 

SDH Networks - Frame Structure and Channel Mappings

SDH Networks - Frame Structure & Channel Mappings

Figure below  shows a simple SDH network and summarizes some Channel mapping and multiplexing in SDH. 

PDH equipment with various interfaces (2, 45 and 140 Mb/s in this example) may use SDH muxes and regenerators for transport. 

The PDH signals are adapted to the synchronous frequency by justification and forms a C-n. Adding lower-or higher-order path overhead (LO/HOPOH) forms a VC-n. 

Pointers are used to indicate the start of the signal in the VC-ns. Lower order VC-ns are multiplexed to form a VC-4, which finally forms an STM-n (n=1, 4, 16, 64, 256) signal.

The SDH regenerator terminates a regenerator section and the multiplex section is formed between the endpoint SDH muxes. The location of the overhead of the different paths and sections are indicated in the SDH frame in the middle.

Fig. below shows the structure of a a SDH frame with payload area that carries data and the overhead sections that are used for proving information to payload data, alarms, protection and signaling.

Section Overhead (SOH): The first 9 bytes in each of the 9 rows are called the overhead. 

G.707 makes a distinction between the regenerator section overhead (RSOH) and the multiplex section overhead (MSOH). 

The reason for this is to be able to couple the functions of certain overhead bytes to the network architecture. The table below describes the individual functions of the bytes.

Path overhead: The path overhead (POH) plus a container forms a virtual container. The POH has the task of monitoring quality and indicating the type of container. 

The format and size of the POH depends on the container type. A distinction is made between two different POH types:

The heterogeneous nature of modern network structures has made it necessary that  all PDH and ATM signals are transported over the SDH network. 

The process of matching the signals to the network is  called mapping. 

The container is the basic package unit for tributary channels. A special container(C-n) is  provided for each PDH tributary signal. These containers are always much larger than the payload to be transported. 

There main capacity is used partly for justification (stuffing) in order to equalize out timing inaccuracies in the PDH signals. Where synchronous tributaries are mapped, fixed fill bytes are inserted instead of justification bytes. 

A virtual container (VC-n) is made up from the container thus formed together with the path overhead(POH). This is transmitted unchanged over a path through the network. The next step towards formation of a complete STM-N signal is the addition of a pointer indicating the start of the POH. 

The unit formed by the pointer and the virtual container is called an administrative unit (AU-n) or a tributary unit (TU-n). 

Several TUs taken together form a tributary unit group (TUG-n); these are in turn collected together into a VC. One or more Aus form an administrative unit group(AUG). 

Finally, the AUG plus the section over-head(SOH) forms the STM-N.

Transmission at the higher hierarchy levels: To achieve higher bit rates, AU-3/4s are multiplexed into STM-N frames.

The following hierarchy levels are defined in SDH:

STM-1 155.52Mbit/s

STM-4 622.08Mbit/s

STM-16 2488.32Mbit/s

STM-64 9953.28Mbit/s

The STM-N frame structures are basically N times the STM-1 structure. For example, the STM-4 overhead is four times the size of the STM-1 overhead. The SOH content is specified for each stage individually. 

SDH Networks - Error & Alarm monitoring

Error and Alarm monitoring (SDH alarms)

Large numbers of alarm and error messages are an integral part of SDH networks. In SDH, these  are referred to as defects and anomalies, respectively.   They are coupled to network sections and the corresponding overhead information. The advantage of this detailed information is illustrated as  follows:

Complete failure of a circuit results, for example, in a LOS alarm (loss of signal) in the  receiving network element. This alarm triggers a complete chain of subsequent messages in the form of AIS (alarm indication signals; see figure below .

The transmitting side is informed of the failure by the  return of an RDI alarm(remote defect indication). The alarm messages are transmitted in fixed bytes in the SOH or POH. For example, byte G1 is used for the HP-RDI alarm.

If the received signal contains bit errors, the sensor indicates BIP errors. Since this is not the  same as a complete failure of the circuit, the alarm here is referred to as ananomaly that is indicated back in the direction of transmission. 

The return message is called a REI (remote error indication). Table below is  a list of all possible defects and anomalies and the corresponding bytes and their meanings.

One of the weak points of SDH is its inflexibility and coarse divisions of bandwidth channels: for a 5 Mb/s service a 2 Mb/s (VC-12) channel is not enough while a 34 Mb/s channel (VC-3) means low utilization. Virtual concatenation, makes it possible to allocate several VC-n’s into a group (VCG) of bundled channels giving 2 Mb/s granularity. The term “virtual” comes from the fact that the transport network (ADMs/DXCs) need not know about or identify the different channels in the VCG, which only is defined at the end-nodes. Dynamic channel re-sizing end-to-end can be done using Link Capacity Adjustment Scheme (LCAS) signaling .

SDH Networks - Protection Mechanism

SDH Networks - Protection Mechanisms 

Network failures, whether due to human error or faulty technology,  can be very expensive for users and network providers alike. As a result, the subject of so called fall-back  mechanisms is very important. A wide range of standardized mechanisms are incorporated into synchronous networks in order to compensate for failures in network elements.

Protection Mechanism  evolution:

  v  80' s –

-          DCS-based Mesh Restoration of DS3 Facilities

-          Centralized (EMS/NMS)

-          Path-based, failure-dependent, after fault detection and isolation

-          Capacity-efficient but slow (~ minutes)

  v  90' s –

-          ADM-based Ring Protect ion of SONET/SDH Facilities

-          Distributed

-          Path-based (UPSR)  or span-based (BLSR) , pre-determined

-          Fast (“ 50 msec” )  but capacity-inefficient

  v  2000' s  –

-          OXC-based Mesh Protect ion/Restorat ion

-          Distributed

-          Path-based, failure-independent, pre-determined and pre-provisioned

-          Capacity-efficient AND fast  (10’s – 100's msec)

Automatic protection switching (APS):

Two basic types of protection architecture are distinguished in APS.

One is the linear protection mechanism used for point-to-point connections. 

The other basic form is the so called ring protection mechanism which can take on many different forms. 

Both mechanisms use spare circuits or components to provide the back-up path. Switching is controlled by the overhead bytes K1 and K2.

Linear protection : The simplest form of back-up is the so-called 1+1 APS. Here, each working line is protected by one protection line. If a defect occurs, the protection agent in the network elements at both ends switch the circuit over to the protection line. The switch over is triggered by a defect such as LOS. Switching at the far end is initiated by the return of an acknowledgment in the backward channel.

1+1 architecture includes 100%redundancy, as there is a spare line for each working line. Economic considerations have led to the preferential use of 1:N architecture, particularly for long-distance paths. In this case, several working lines  are protected by a single back-up line. If switching is necessary, the two ends of the affected path are switched over to the back-up line.

The 1+1 and 1:N protection mechanisms are standardized in ITU-T Recommendation G.783.

The reserve circuits can be used for lower-priority traffic, which is simply interrupted if the circuit is  needed to replace a failed working line.

Ring protection : A ring is the simplest and most cost-effective way of linking a number of network elements. Various protection mechanisms are available for this type of network architecture, only some of which have been standardized in ITU-T Recommendation G.841.

A basic distinction must be made between ring structures with unidirectional and bi-directional connections. Figure below shows the basic principle of APS for unidirectional rings. Let us assume that there is an interruption in the circuit between the network elements A and B. Direction y is unaffected by this fault. An alternative path must, however, be found for direction x. 

The connection is therefore switched to the alternative path in network elements A and B. The other network elements (C and D) switch through the back-up path. This switching process is referred to as line switched. 

A simpler method is to use the so-called path switched ring (see figure). Traffic is transmitted simultaneously over both the working line and the protection line. If there is  an interruption, the receiver (in this case A) switches to the protection line and immediately takes up the connection

In unidirectional path-switched rings (UPSRs), two redundant (path-level) copies of protected traffic are sent in either direction around a ring. A selector at the egress node determines which copy has the highest quality, and uses that copy, thus coping if one copy deteriorates due to a broken fiber or other failure. UPSRs tend to sit nearer to the edge of a network, and as such are sometimes called collector rings.

Because the same data is sent around the ring in both directions, the total capacity of a UPSR is equal to the line rate N of the OC-N ring. For example, in an OC-3 ring with 3 STS-1s used to transport 3 DS-3s from ingress node A to the egress node D, 100 percent of the ring bandwidth (N=3) would be consumed by nodes A and D. Any other nodes on the ring could only act as pass-through nodes.

The SDH equivalent of UPSR is SubNetwork Connection Protection (SNCP); SNCP does not impose a ring topology, but may also be used in mesh topologies.

Bi-directional rings : In this network structure, connections between network elements are bi-directional. This is indicated in figure by the absence of arrows when compared with first figure.The overall capacity of the network can be split up for several paths each with one bi-directional working line, while for unidirectional rings, an entire virtual ring is required for each path. If a fault occurs between neighboring elements A and B, network element B triggers protection switching and controls network element A by means of the K1 and K2 bytes in the SOH.

Bidirectional line-switched ring

Bidirectional line-switched ring (BLSR) comes in two varieties: two-fiber BLSR and four-fiber BLSR. BLSRs switch at the line layer. Unlike UPSR, BLSR does not send redundant copies from ingress to egress. Rather, the ring nodes adjacent to the failure reroute the traffic "the long way" around the ring on the protection fibers. BLSRs trade cost and complexity for bandwidth efficiency, as well as the ability to support "extra traffic" that can be pre-empted when a protection switching event occurs. In four-fiber ring, either single node failures, or multiple line failures can be supported, since a failure or maintenance action on one line causes the protection fiber connecting two nodes to be used rather than looping it around the ring.

BLSRs can operate within a metropolitan region or, often, will move traffic between municipalities. Because a BLSR does not send redundant copies from ingress to egress, the total bandwidth that a BLSR can support is not limited to the line rate N of the OC-N ring, and can actually be larger than N depending upon the traffic pattern on the ring. In the best case, all traffic is between adjacent nodes. The worst case is when all traffic on the ring egresses from a single node, i.e., the BLSR is serving as a collector ring. In this case, the bandwidth that the ring can support is equal to the line rate N of the OC-N ring. This is why BLSRs are seldom, if ever, deployed in collector rings, but often deployed in inter-office rings.

The SDH equivalent of BLSR is called Multiplex Section-Shared Protection Ring (MS-SPRING).

Despite predictions of SDH reduction in favour of lower cost transmission, such as Gigabit  Ethernet, still the vast majority of traffic in the operators transport networks runs of SDH.

Apart from the need for well-defined protection switching strategies, SDH on reconfigurable optical networks is quite straightforward. When the cost for an OXC port is lower than for a DXC port an optical bypass layer can be formed reducing the total network cost, especially if protection is done in the optical layer.

ASON

Automatically Switched Optical Network (ASON) is a network management feature which enables dynamic control of transmission networks through an automated management of network resources. ASON functionality is typically made available in higher capacity SDH/SONET and DWDM networks. It enables fast end-to-end service provisioning, re-routing and restoration. As per ITU-T G.8080, the key goals for ASON are:

-          Facilitate fast and efficient configuration within transport layer network

-          Reconfigure or modify connections

-          Perform a protection/restoration function 

SDH Networks - Synchronisation & Clocking

Synchronization 

Synchronous is the first word in the term SDH for a very good reason.If  synchronization is  not guaranteed, considerable degradation in network function, and even total  failure of the network can be the result. To avoid this worst case scenario, all network elements are synchronized to a central clock. This central clock is generated by a high precision primary reference clock (PRC)  unit conforming to ITU-T Recommendation G.811. This specifies an accuracy of 1X 10^-11. 

This Clock signal must be distributed through out the entire network. A hierarchical structure is used for this; the signal is passed on by the subordinate synchronization supply units(SSU) and synchronous equipment clocks(SEC). The synchronization signal path scan be the same as those used for SDH Communications.

The clock signal is regenerated in the SSUs and SECs with the aid of phase-locked loops. If the clock supply fails, the affected network element switches over to a clock source with the same or lower quality, or if this is not possible, it switches to hold-over mode. In this situation, the clock signal is  kept relatively accurate by controlling the oscillator by applying the stored frequency correction values for the previous hours and taking the temperature of the oscillator into account. 

Clock “islands” must be avoided at all costs, as these would  drift out of synchronization with the passage of time and the total failure disaster would be the result. Such islands are prevented by signaling the network elements with the  aid of synchronization status messages (SSM, part of the S1byte). The SSM informs the neighboring network element about the status of the clock supply and is part of the overhead.

Special problems arise at gateways between networks with independent clock supplies.SDH network elements can compensate for clock offsets with in certain limits by means of pointer operations. Pointer activity is thus are liable indicator of problems with the clock supply.

SDH Networks - Total Management Network (TMN)

TMN in the SDH network

The principle of telecommunications management network(TMN) technology was established in 1989 with the publication by the CCITT(now ITU-T) of Recommendation M.3010. 

The functions of a TMN are summed up in the expression “Operation, administration, maintenance and provisioning” (OAM&P). This includes monitoring the network performance and the checking of error messages, among other things. To provide these functions, TMN uses object-oriented techniques based on the OSI reference model. The TMN model comprises one manager handling several agents. 

The agents in turn each handle several managed objects(MO). The manager is included in the operating system (OS) which forms the “control center” for the network as a whole or in part. In a SDH network, the agents are located in the  network elements (NE),  such as switches, etc. AMO may be a physical unit (e.g. a plug-in card, multiplex section, etc.) but can also occur as a logical element (e.g. a virtual connection). 

TMN also distinguishes between logical management units. For  example, one management unit operates at network level, handling individual NEs. Another management unit operates at the service level, e.g. for  monitoring billing charges. These tasks are performed in modern telecommunications networks by using the common management information protocol (CMIP). 

The simple network management protocol(SNMP) is often mentioned in this context; this is basically a simplified form  of CMIP.  SNMP is mainly used in data communications, however, and cannot cope with the requirement of large telecommunications networks. The Q3 interface,  which is where the exchange of data between manager and agent takes  place, is the point of reference for CMIP.  

CMIP is also used where several TMNs or their  managers are linked  together  via the X interface. Since large quantities of data are not generally required for exchanging information in the TMN, the capacity of the embedded communication channels(ECC) or data communication channels(DCC) is sufficient when managing SDH networks. 

Channels D1 to D3 with a capacity of 192kbit/s(DCCP) are used for SDH-specific NE management. Channels D4 to D12 with a capacity of 576kbit/s(DCCM) can be used for non SDH specific purposes.

To distinguish the implementation in the section overhead (SOH) data channels from the Q interface, the term QECC protocol is used. To summarize, such networks are called SDH management networks (SMN); these are primarily responsible for managing network elements. SMNs can also be subdivided into sub-networks(SMS).

SDH Networks : Performance analysis standards

G.821,G.826 and M.2100 performance analysis

The quality of digital links is determined with the aid of bit error ratio tests (BERT).The results of such measurements must, however, be classified in someway, not least because the quality of a transmission Path is often the subject of a contract between the network provider and the telecommunications user. 

For this reason, an objective means of classifying a line as either  “good” or “bad” is required. The ITU-T Recommendations G.821, G.826 and M.2100 are internationally recognized standards that specify these parameters.

G.821 :  This Recommendation was originally specified for international circuit- switched nX64kbit/s connections and expanded to include higher bit rates as time went on. A hypothetical reference connection is the basis used for determining quality parameters; this comprises an international long distance segment, a national segment and a subscriber access segment.

G.821definitions:

*Errored second(ES):  A one second time interval in which one or more bit errors occurs.

*Severely errored second(SES): A one-second time interval in which the bit error ratio exceeds 10^-3

*Unavailable second(US):  A circuit is considered to be unavailable from the first of at least ten consecutive SES. The circuit is available from the first of at least ten consecutive seconds which are not SES.

The original version of G.821 also included:

*Degraded minute(DM): A one-minute time interval in which the bit error ratio exceeds10^-6

Derived parameter:

*Error-free second(EFS):  A one-second time interval in which no bit errors occur. The disadvantage of this method is that it relies on the evaluation of bit errors and so the test channel must be taken out of service to perform the measurement.

G.826:

 This Recommendation, issued in 1993,takes higher bit rates into account and allows in-service measurement as it relies on the evaluation of block errors.

G.826 definitions:

* Errored second(ES): A one-second time interval containing one or more errored blocks.

*Errored block(EB): A block containing one or more errored bits.

*Severely errored second(SES): A one-second time interval in which more than 30% of the blocks are errored or which contains at least one severely disturbed period(SDP).

*Background block error(BBE): An errored block that is not a SES.

*Unavailable second(US):  seeunderG.821, above.

The results are referred to the measurement time. 

This gives the following error parameters: Errored seconds ratio (ESR), severely errored seconds ratio(SESR) and background block error ratio (BBER). The specified quality requirements refer to  a particular path.

There commended measurement time for G.821andG.826is 30days.

M.2100 Recommendation:

M.2100 specifically applies to commissioning and maintenance. Commissioning consists of a 15-minute line up phase followed  by a 24-hour in-service measurement.  Once the line up phase is completed successfully, errors may occur within certain limits. If this is the case, the line remains in service, but must continue to be monitored for a further 7days.The measurement procedures are defined in M.2110 and M.2120. The limit values are derived form the performance parameters specified in G.821 and G.826.

Jitter:

The term jitter refers to phase variations in a digital signal. Put another way, the edges of the digital signal may differ from the expected ideal positions in time. Jitter is described in terms of its amplitude(expressed in unit intervals, UI)and its frequency. If the jitter frequency is below 10Hz,the term used is wander. 

Signals that are affected by jitter can not be sampled accurately; in an extreme situation, this might result in misinterpretation of the input signal. This results in single errors or error bursts and a corresponding degradation in transmission quality. 

Jitter and wander can also be the cause of buffer under flow or overflow, which leads to bit slips. The theoretical limit of correct sampling at high jitter frequencies is half the bit width. Distortion and additive noise means that the actual limit must be set much lower than this.

What causes jitter? The clock sources for network elements such as regenerators and add/drop multiplexers are one possible cause. Various types of jitter are differentiated as shown in the following table:

Other causes of jitter are interference signals and phase noise. Jitter caused by interference signals is  also called non-systematic jitter. Phase noise occurs despite the use of a central clock as a result of thermal noise and drift in the oscillator used. Various measurement methods have been developed for the different causes of jitter.

This Page is under construction

This Page is under construction