Vous êtes sur la page 1sur 9

SDH frame

An STM-1 frame. The first nine columns contain the overhead and the pointers. For the
sake of simplicity, the frame is shown as a rectangular structure of !" columns and nine
rows #ut the protocol does not transmit the #ytes in this order.
For the sake of simplicity, the frame is shown as a rectangular structure of !" columns
and nine rows. The first three rows and nine columns contain regenerator section
overhead $%S&'( and the last five rows and nine columns contain multiple) section
overhead $MS&'(. The fourth row from the top contains pointers.
The STM-1 $Synchronous Transport Module, level 1( frame is the #asic transmission
format for S*'+the first level of the synchronous digital hierarchy. The STM-1 frame is
transmitted in e)actly 1, -s, therefore, there are .,""" frames per second on a
1,,., M#it/s &0-1 fi#er-optic circuit.
2n# 13
The STM-1 frame consists of overhead and
pointers plus information payload. The first nine columns of each frame make up the
Section &verhead and Administrative 4nit 5ointers, and the last 61 columns make up
the 7nformation 5ayload. The pointers $'1, ', '1 #ytes( identify administrative units
$A4( within the information payload. Thus, an &0-1 circuit can carry 1,".116 M#it/s of
payload, after accounting for the overhead.
2n# 3
0arried within the information payload, which has its own frame structure of nine rows
and 61 columns, are administrative units identified #y pointers. Also within the
administrative unit are one or more virtual containers $80s(. 80s contain path overhead
and 80 payload. The first column is for path overhead9 it is followed #y the payload
container, which can itself carry other containers. Administrative units can have any
phase alignment within the STM frame, and this alignment is indicated #y the pointer in
row four.
The section overhead $S&'( of a STM-1 signal is divided into two parts: the regenerator
section overhead $%S&'( and the multiplex section overhead $MS&'(. The overheads
contain information from the transmission system itself, which is used for a wide range of
management functions, such as monitoring transmission ;uality, detecting failures,
managing alarms, data communication channels, service channels, etc.
The STM frame is continuous and is transmitted in a serial fashion: #yte-#y-#yte, row-
Transport overhead
The transport overhead is used for signaling and measuring transmission error rates, and
is composed as follows:
Section overhead
0alled %S&' $regenerator section overhead( in S*' terminology: ! octets
containing information a#out the frame structure re;uired #y the terminal
<ine overhead
0alled MS&' $multiple) section overhead( in S*': =, octets containing
information a#out error correction and Automatic 5rotection Switching messages
$e.g., alarms and maintenance messages( as may #e re;uired within the network.
A4 5ointer
5oints to the location of the >1 #yte in the payload $the first #yte in the virtual
Path virtual envelope
*ata transmitted from end to end is referred to as path data. 7t is composed of two
5ayload overhead $5&'(
?ine octets used for end-to-end signaling and error measurement.
4ser data $!!= #ytes for STM-"/STS-1, or ,1=" octets for STM-1/STS-1c(
For STS-1, the payload is referred to as the synchronous payload envelope $S5@(, which
in turn has 1. stuffing #ytes, leading to the STS-1 payload capacity of !,6 #ytes.
The STS-1 payload is designed to carry a full 5*' *S1 frame. Ahen the *S1 enters a
S&?@T network, path overhead is added, and that S&?@T network element $?@( is said
to #e a path generator and terminator. The S&?@T ?@ is line terminating if it processes
the line overhead. ?ote that wherever the line or path is terminated, the section is
terminated also. S&?@T regenerators terminate the section, #ut not the paths or line.
An STS-1 payload can also #e su#divided into seven virtual tributary groups $8TBs(.
@ach 8TB can then #e su#divided into four 8T1., signals, each of which can carry a
5*' *S1 signal. A 8TB may instead #e su#divided into three 8T signals, each of
which can carry a 5*' @1 signal. The S*' e;uivalent of a 8TB is a T4B-9 8T1., is
e;uivalent to 80-11, and 8T is e;uivalent to 80-1.
Three STS-1 signals may #e multiple)ed #y time-division multiple)ing to form the ne)t
level of the S&?@T hierarchy, the &0-1 $STS-1(, running at 1,,., M#it/s. The signal is
multiple)ed #y interleaving the #ytes of the three STS-1 frames to form the STS-1 frame,
containing ,=1" #ytes and transmitted in 1, -s.
'igher-speed circuits are formed #y successively aggregating multiples of slower
circuits, their speed always #eing immediately apparent from their designation. For
e)ample, four STS-1 or A4= signals can #e aggregated to form a 6.". M#it/s signal
designated &0-1 or STM-=.
The highest rate commonly deployed is the &0-!6. or STM-,6 circuit, which operates
at rate of Cust under 1.., B#it/s.
Ahere fi#er e)haustion is a concern, multiple S&?@T
signals can #e transported over multiple wavelengths on a single fi#er pair #y means of
wavelength-division multiple)ing, including dense wavelength-division multiple)ing
$*A*M( and coarse wavelength-division multiple)ing $0A*M(. *A*M circuits are
the #asis for all modern su#marine communications ca#le systems and other long-haul
SONET/SDH network management protocols
Most S&?@T ?@s have a limited num#er of management interfaces defined:
@lectrical interface
The electrical interface, often a ,"-ohm coa)ial ca#le, sends S&?@T T<1
commands from a local management network physically housed in the central
office where the S&?@T network element is located. This is for local
management of that network element and, possi#ly, remote management of other
S&?@T network elements.
0raft interface
<ocal DcraftspersonsD $telephone network engineers( can access a S&?@T
network element on a Dcraft portD and issue commands through a dum# terminal
or terminal emulation program running on a laptop. This interface can also #e
attached to a console server, allowing for remote out-of-#and management and
*ata communication channels $*00s(
S&?@T and S*' have dedicated data communication channels $*00s( within
the section and line overhead for management traffic. Benerally, section overhead
$regenerator section in S*'( is used. According to 7T4-T B.!!1, there are three
modes used for management:

75 -only stack, using 555 as data-link
&S7 -only stack, using <A5-* as data-link
*ual $75E&S7( stack using 555 or <A5-* with tunneling functions to
communicate #etween stacks.
To handle all of the possi#le management channels and signals, most modern network
elements contain a router for the network commands and underlying $data( protocols.
The main functions of network management include:
?etwork and network-element provisioning
7n order to allocate #andwidth throughout a network, each network element must
#e configured. Although this can #e done locally, through a craft interface, it is
normally done through a network management system $sitting at a higher layer(
that in turn operates through the S&?@T/S*' network management network.
Software upgrade
?etwork-element software upgrades are done mostly through the S&?@T/S*'
management network in modern e;uipment.
5erformance management
?etwork elements have a very large set of standards for performance
management. The performance-management criteria allow not only monitoring
the health of individual network elements, #ut isolating and identifying most
network defects or outages. 'igher-layer network monitoring and management
software allows the proper filtering and trou#leshooting of network-wide
performance management, so that defects and outages can #e ;uickly identified
and resolved.
Aith advances in S&?@T and S*' chipsets, the traditional categories of network
elements are no longer distinct. ?evertheless, as network architectures have remained
relatively constant, even newer e;uipment $including multi-service provisioning
platforms( can #e e)amined in light of the architectures they will support. Thus, there is
value in viewing new, as well as traditional, e;uipment in terms of the older categories.
Traditional regenerators terminate the section overhead, #ut not the line or path.
%egenerators e)tend long-haul routes in a way similar to most regenerators, #y
converting an optical signal that has already traveled a long distance into electrical format
and then retransmitting a regenerated high-power signal.
Since the late 1FF"s, regenerators have #een largely replaced #y optical amplifiers. Also,
some of the functionality of regenerators has #een a#sor#ed #y the transponders of
wavelength-division multiple)ing systems.
Add-drop multiplexer
Add-drop multiple)ers $A*Ms( are the most common type of network elements.
Traditional A*Ms were designed to support one of the network architectures, though
new generation systems can often support several architectures, sometimes
simultaneously. A*Ms traditionally have a high-speed side $where the full line rate
signal is supported(, and a low-speed side, which can consist of electrical as well as
optical interfaces. The low-speed side takes in low-speed signals, which are multiple)ed
#y the network element and sent out from the high-speed side, or vice-versa.
Digital cross connect system
%ecent digital cross connect systems $*0Ss or *G0s( support numerous high-speed
signals, and allow for cross-connection of *S1s, *S1s and even STS-1s/1c and so on,
from any input to any output. Advanced *0Ss can support numerous su#tending rings
Network architectures
S&?@T and S*' have a limited num#er of architectures defined. These architectures
allow for efficient #andwidth usage as well as protection $i.e. the a#ility to transmit traffic
even when part of the network has failed(, and are fundamental to the worldwide
deployment of S&?@T and S*' for moving digital traffic. @very S*'/S&?@T
connection on the optical 5hysical layer uses two optical fi#ers, regardless of the
transmission speed.
Linear Automatic Protection Switching
<inear Automatic 5rotection Switching $A5S(, also known as 1+1, involves four fi#ers:
two working fi#ers $one in each direction(, and two protection fi#ers. Switching is #ased
on the line state, and may #e unidirectional $with each direction switching
independently(, or #idirectional $where the network elements at each end negotiate so
that #oth directions are generally carried on the same pair of fi#ers(.
nidirectional path-switched ring
7n unidirectional path-switched rings $45S%s(, 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 ;uality, and uses that copy, thus coping if one
copy deteriorates due to a #roken fi#er or other failure. 45S%s tend to sit nearer to the
edge of a network, and as such are sometimes called collector rings. Hecause the same
data is sent around the ring in #oth directions, the total capacity of a 45S% is e;ual to the
line rate N of the &0-N ring.
For e)ample, in an &0-1 ring with 1 STS-1s used to
transport 1 *S-1s from ingress node A to the egress node D, 1"" percent of the ring
#andwidth $NI1( would #e consumed #y nodes A and D. Any other nodes on the ring
could only act as pass-through nodes. The S*' e;uivalent of 45S% is subnetwork
connection protection $S?05(9 S?05 does not impose a ring topology, #ut may also #e
used in mesh topologies.
!idirectional line-switched ring
Hidirectional line-switched ring $H<S%( comes in two varieties: two-fi#er H<S% and
four-fi#er H<S%. H<S%s switch at the line layer. 4nlike 45S%, H<S% does not send
redundant copies from ingress to egress. %ather, the ring nodes adCacent to the failure
reroute the traffic Dthe long wayD around the ring on the protection fi#ers. H<S%s trade
cost and comple)ity for #andwidth efficiency, as well as the a#ility to support De)tra
trafficD that can #e pre-empted when a protection switching event occurs. 7n four-fi#er
ring, either single node failures, or multiple line failures can #e supported, since a failure
or maintenance action on one line causes the protection fi#er connecting two nodes to #e
used rather than looping it around the ring.
H<S%s can operate within a metropolitan region or, often, will move traffic #etween
municipalities. Hecause a H<S% does not send redundant copies from ingress to egress,
the total #andwidth that a H<S% can support is not limited to the line rate N of the &0-N
ring, and can actually #e larger than N depending upon the traffic pattern on the ring.

7n the #est case, all traffic is #etween adCacent nodes. The worst case is when all traffic
on the ring egresses from a single node, i.e., the H<S% is serving as a collector ring. 7n
this case, the #andwidth that the ring can support is e;ual to the line rate N of the &0-N
ring. This is why H<S%s are seldom, if ever, deployed in collector rings, #ut often
deployed in inter-office rings. The S*' e;uivalent of H<S% is called Multiplex Section-
Shared rotection !ing $MS-S5%7?B(.
0lock sources used for synchroniJation in telecommunications networks are rated #y
;uality, commonly called a stratum
. Typically, a network element uses the highest
;uality stratum availa#le to it, which can #e determined #y monitoring the
synchroniJation status messages $SSM( of selected clock sources.
SynchroniJation sources availa#le to a network element are:
<ocal e)ternal timing
This is generated #y an atomic 0aesium clock or a satellite-derived clock #y a
device in the same central office as the network element. The interface is often a
*S1, with sync-status messages supplied #y the clock and placed into the *S1
<ine-derived timing
A network element can choose $or #e configured( to derive its timing from the
line-level, #y monitoring the S1 sync-status #ytes to ensure ;uality.
As a last resort, in the a#sence of higher ;uality timing, a network element can go
into a holdover mode until higher-;uality e)ternal timing #ecomes availa#le
again. 7n this mode, the network element uses its own timing circuits as a
Timing loops
A timing loop
occurs when network elements in a network are each deriving their
timing from other network elements, without any of them #eing a DmasterD timing source.
This network loop will eventually see its own timing Dfloat awayD from any e)ternal
networks, causing mysterious #it errors+and ultimately, in the worst cases, massive loss
of traffic. The source of these kinds of errors can #e hard to diagnose. 7n general, a
network that has #een properly configured should never find itself in a timing loop, #ut
some classes of silent failures could nevertheless cause this issue.
Next-generation SONET/SDH
S&?@T/S*' development was originally driven #y the need to transport multiple 5*'
signals+like *S1, @1, *S1, and @1+along with other groups of multiple)ed 6= k#it/s
pulse-code modulated voice traffic. The a#ility to transport ATM traffic was another
early application. 7n order to support large ATM #andwidths, concatenation was
developed, where#y smaller multiple)ing containers $e.g., STS-1( are inversely
multiple)ed to #uild up a larger container $e.g., STS-1c( to support large data-oriented
&ne pro#lem with traditional concatenation, however, is infle)i#ility. *epending on the
data and voice traffic mi) that must #e carried, there can #e a large amount of unused
#andwidth left over, due to the fi)ed siJes of concatenated containers. For e)ample,
fitting a 1"" M#it/s Fast @thernet connection inside a 1,, M#it/s STS-1c container leads
to considera#le waste. More important is the need for all intermediate network elements
to support newly-introduced concatenation siJes. This pro#lem was overcome with the
introduction of 8irtual 0oncatenation.
8irtual concatenation $80AT( allows for a more ar#itrary assem#ly of lower-order
multiple)ing containers, #uilding larger containers of fairly ar#itrary siJe $e.g.,
1"" M#it/s( without the need for intermediate network elements to support this particular
form of concatenation. 8irtual concatenation leverages the G..6 or Beneric Framing
5rocedure $BF5( protocols in order to map payloads of ar#itrary #andwidth into the
virtually-concatenated container.
The <ink 0apacity AdCustment Scheme $<0AS( allows for dynamically changing the
#andwidth via dynamic virtual concatenation, multiple)ing containers #ased on the short-
term #andwidth needs in the network.
The set of ne)t-generation S&?@T/S*' protocols that ena#le @thernet transport is
referred to as @thernet over S&?@T/S*' $@oS(.