SONET/SDH

SONET is an acronym for Synchronous Optical Network. SONET is widely

used in telephone network and is one of the first large scale optical transmission
systems Digital information is sent through optical fibers using a LED or a laser
source. However, most of data processing, switching etc. are done electronically.
SONET is primarily used in the North America while Europe and Japan use a
modified version, called the Synchronous Digital Hierarchy (SDH). SONET
arose out of a need to find a solution of the problem of inter-operability among
various vendors and technologies that existed in sixties and seventies.
Multiplexing :
In telecommunication, the phrase multiplexing is used to denote the process of
combining two or more channels into a single channel. For instance, in encoding
a video stream, we need to multiplex audio and video on single channel. One of
the ways of doing it is known as time division multiplexing. The basic idea of
the process is as follows.
Suppose we have three individual users in Mumbai who wish to send low stream
data to three users in Delhi. Assigning physical lines for each will be expensive
and such expense would grow exponentially with increasing number of users. If
instead, we used a higher bit rate channel, we could use different time slots for the
different data set. The figure below illustrates how this is achieved.
In sixties, the International Telecommunication Union (ITU) defined what is now
known as a T-1 carrier. The analog voice data was digitized by sampling at a
rate which is twice that of the maximum frequency component in the signal. This
is known as Pulse Code Modulation (PCM). The T-1 standard was an universal
agreement on a sampling rate of 8 kHz and a channel rate of 64 kilo-bit per second.
The single voice channel is known as DS-0 signal (DS= Digital Signal). Thus, if
we return back to the example of three users, with each transmitting at a rate of 64
kbps, we could transmit them as a sequential stream of data over a single channel
capable of transmitting at 192 kbit/sec.
Channel 1 Channel 1

Channel 2 Channel 2
MULTIPLEXER DEMULTIPLEXER

Channel 3 Channel 3

1
The way to achieve this is to divide the high rate channel into a series of time slots
and the time slots could be assigned to the invividual data stream.
Example :
The T-1 line (T= Transmitted) is a channel capable of transmitting at a speed of
1.544 Mbit/sec. The voice is sampled at 8 kbits/sec, so that the time occupied
by a bit is 125µs. The interval is subdivided into 24 time slots with each time
slot coding 8 bits of data for a channel. Thus in each 125 µs there are 192 bits.
Adding one bit to mark the beginning, there are 192 bits transmitted in 125 µs
which gives the speed to be 1.544 Mbits/sec. The signal formed by interleaving
24 DS-0 signal is known as DS-1 signal and the corresponding transmitted signal
is T-1. Proceeding further, four DS-1 signals are interleaved to give a 6.3 Mbit/sec
DS-2 signal (transmitted as T-2) and seven DS-2 give rise to a 45 Mbit/sec DS-3
transmitted over a T-3 line.
Asynchronous, Plesiochronous and Synchronous Systems : Electrical signals
which are used in transmission use system clocks which are very accurate. Gen-
erally, they use clocks which are based on natural frequency of vibration of some
crystals, such as quartz crystals. Though such clocks are accurate, there can be
differences between different clocks used by different systems and such clocks are
not synchronised. Such systems are called asynchronous.
A plesiochronous system also uses different clocks but they are accurate to within
a specified tolerance. Thus the clocks have phase differences. Plesiochronous
Digital Hierarchy (PDH) was the most commonly used digital transmission sys-
tem before the advantage of SDH. To accommodate delays between different
clocks, extra bit(s), known as justification bits may be added to the multiplexed
stream. This is known as bit stuffing. These bits will have to discarded while
demultiplexing.
The disadvantage of PDH is that at higher speeds, it does not allow adding or drop-
ping channels without having to first demultiplexing the channels. For instance,
consider a service provider who has to provide a 2 Mbps line to a customer from
his bandwidth of 155 Mbps.He has to locate and identify frames in the 2 Mb line.
For this, the service provider will have to first demultiplex the line into its 64 kbps
constituents, add the customer and remultiplex the system. The Digital telephone
hierarchy for the North America and Europe is shown in the following table. Japan
uses a scheme which is close to that of Europe but is not identical with it. The
data rates are in Mbit/sec.

2
 North America Europe
0.064 (DS0) 0.064
1.544 (DS1) 2.048 (E1)
6.312 (DS2) 8.448 (E2)
44.736 (DS3) 34.368 (E3)
139.264 (DS4) 139.264 (E4)

Principle of SONET/SDH

• In a synchronous device the clock transitions occur precisely at the same

rate. All signal transitions are fixed with reference to a very accurate atomic
clock (such as Cs clock), called Primary Reference Clock (PRC). The
accuracy of such a clock is one part in 1011 or better. The advantage of a
synchronous system is that multiple signals can be stacked without any need
for bit stuffing.

• In SDH, data from different sources are multiplexed in a way that the chan-
nels have fixed locations with respect to the framing byte.

• As the location is fixed, it is not necessary to demultiplex while dropping a

single channel from the stream.
The basic signal of SONET is synchronous transport signal, called STS-1 which
operates at 51.84 Mbit/sec. After conversion to optical signals, STS- 1 is known
as optical carrier, or OC-1. The higher level signals are multiples of STS-1 signal
and operate at multiples of base frequency. Thus STS-3 (or its optical equiva-
lent OC-3) operates at a bit rate of 155.52 Mbps interleaving frames from three
STS-1 signals. This STS-3/OC-3 is the base signal for SDH and is known as the
synchronous transport module or STM-1. The hierarchy for SONET/SDH hiher-
archy is shown in the following table.

Signal Name Data Rate

SONET SDH Optical (Mbps)
STS-1 – OC-1 51.84
STS-3 STM-1 OC-3 155.52
STS-12 STM-4 OC-12 622.08
STS-48 STM-16 OC-48 2488.32
STS-192 STM-64 OC-192 9953.28
STS-768 STM-256 OC-768 ∼40 Gbps

3
Each rate is an exact multiple of the lower rate ensuring that the hierarchy is syn-
chronous. The network hierarchy is organized in a master-slave relationship with
the lower level nodes receiving time signal from the higher level nodes. All clock
level can be traced back to the primary reference clock (PRC), which is very sta-
ble.

FRAMES IN SONET
As the bit rate of information is moving at very fast rate, a convention has to be
applied distinct digital channels that have been multiplexed together can be dis-
tinguished. SONET uses the concept of framing to achieve this.
A framing bit can be thought of as a pointer or an address. As the line is mov-
ing fast, it would be easy to skew it a little to left or to right and the information
would then get out of sequence. The extra bit of information creates a locator for
the system.
SONET organizes data into 810-byte blocks called frames. The bytes are ar-
ranged as a two dimensional array of 9 rows and 90 columns. The data transmis-
sion is serial, i.e. starting with the extreme left byte ( first row, first column) we
proceed to right, byte by byte, and reach the 90th element on the first row. The
next byte assessesd is the first column of the second row and so on. Frames are
sampled at the rate of 8000 frames per second, i.e., one frame is transmitted every
125 micro second. This determines the speed to be as follows :

8000 × 810 = 6, 480, 000 bytes/sec

= 6.48 × 106 × 8 bits/sec
= 51.84 Mb/sec

If one compares the above data rates with that given earlier for the telephone hi-
erarchy, one observes that the SONET/SDH rates are about 10% higher. This is
because of overheads required to administer and monitor signal transmission.
The frame has two main constituents. Each row has 3 bytes of transport overheads
and and 87 bytes of payload. Thus there are 9 × 3 = 27 bytes of transport over-
heads and 9 × 87 = 783 payload bytes. Payload is the traffic that is transported
through the SONET network.

4
 3 cols. 87 cols.

OVERHEAD
PAY LOAD
SECTION

OVERHEAD (POH)
LINE OVERHEAD

PATH

TRANSPORT SYNCHRONOUS PAYLOAD ENVELOPE (SPE)

OVERHEAD

Path overhead :
Out of the 783 bytes of payload, which occupy 87 columns, only 756 bytes in 84
columns actually carry payload capacity of STS-1. The first column, consisting
of 9 bytes, is known as Path overhead (POH). POH is transported point to point
with the payload till the latter is demultiplexed. It carries service provider infor-
mation and monitors end-to-end transport of the payload. Columns 30 and 59 are
not used for payload.
Transport overheads :
Out of the 27 transport overheads, 9 bytes are section overhead and the remaining
18 bytes are line overhead. Transport overhead provides transport information.
The section overhead contains information required for communicating between
successive network elements in the network, such as, repeaters. It also provides a
voice communication channel for maintenance personnel.
The line overhead contains information to monitor line performance. This over-
head provides for the following functions:

• line performance monitoring and line maintenance

5
 • automatic protection switching

• multiplexing signals

SONET Topology :
SONET and SDH standards specify that the network topology be in the form
of a ring. The ring contains fiber redundancies which allow traffic both unidi-
rectionally and bidirectionally. In case of an accidental snapping of a fiber, the
multiplexers can reroute the traffic along an alternate path using redundant fibers.

SONET
NODE

Primary Fiber

SONET
NODE
SONET Redundant
NODE Fiber

SONET
NODE

The SDH frame format is based on the synchronous transport module STM-1.
The frame is organized in 9 rows and 270 columns consisting of 2430 bytes.

6
 270 X N bytes

9 X N bytes 261X N bytes

SECTION STM − N PAYLOAD

9 rows
OVERHEAD

Out of the above 9 × 9 = 81 bytes are section overheads while the remaining 2349
bytes form the payload. The higher rate frames are multiples of STM-1 and are
labelled STM-N.