transmission technology fundamentals SDH & PDH
Many years ago there was a need to make telephone calls digital.
Since our speech is analog, it has to be converted into a digital bit stream.
To do this, samples of the analog signal are taken and represented as
numbers. Expressed in binary digits, every sample uses 8 bits. To gather
all information in a normal conversation, phone services normally use
8000 samples per second. This way, we need a data stream of 64 kbit/s to
transport one telephone call.
That is the reason why 64kbit/s data streams are frequently used in
telecommunications.
So you have a 64 kbit/s data stream that carries one phone call.
The transport medium that you have available is usually a cable or a microwave link
with a much higher data rate, so you can transport many phone calls.
To combine many slow streams of data into one faster one, you use a multiplexer. The
multiplexer puts data into so-called frames that are sent over and over again. How
many 64 kbit/s slots there are in a frame depends on which standard is used. There is
the American standard, ANSI, that uses 24 of these slots for a total data rate of 1.5
Mbit/s. This multiplexed data stream is called a DS1. The European standard, ETSI,
uses 32 slots of 64kbit/s each for a total data rate of 2 Mbit/s. This is called E1. These
frame formats are standardized and also the interfaces are standardized, so PDH
equipment from different manufacturers works together.
We will in this course mainly talk about the ETSI standard and use that in the examples.
The examples are in general applicable to the ANSI standard as well. We will show
when there is a difference in the ANSI standard.
So the 2 Mbit/s frame consists of 32 slots and acts like a container that you fill up
with different types of traffic, like speech or data.
The totally filled slots to the left can represent one telephone call in fixed
networks, while the others can represent traffic coming from GSM networks. In
GSM, the data rate per phone call is lower so that one slot can contain up to four
mobile telephone calls. The traffic in the different slots is independent of the
others and the whole frame will be transported to its final destination.
A DS1 or E1 frame that is used for 64 kbit/s slots is called “structured”. Frames
can also be “unstructured” – in this case the whole frame payload carries just one
data stream. This can be used for transporting Ethernet traffic over PDH
connections.
No matter how the frame is structured, it is transported further in the network.
When several frames are to be transported you put them together in a
multiplexer. This multiplexer takes the incoming signals and puts them together
into only one bit stream out of the multiplexer, with a higher data rate. In the far
end of the transmission chain the multiplexed signal comes to a demultiplexer
that separates the bit stream into the original frames.
This is what is called multiplexing and demultiplexing.
As we go further up into the network to trunk connections with higher data rates,
four E1s, or 2 Mbit/s signals and some stuffing bits become one 8 Mbit/s frame,
E2.
Then four of the E2 signals become one E3 and four of those become one E4.
There are no higher multiplexing levels of PDH in ETSI. In every multiplexing step
some stuffing bits have to be added to compensate for timing differences
between the multiplexers. This is because all multiplexers in PDH have their own
clocking – they are not in synchronization with each other. That’s where the name
Plesiochronous comes from in PDH, almost in synchronization.
In ANSI we have the same kind of steps, but the number of signals and
capacities differs. Four DS1s combine to one DS2. Seven of those make one
DS3 and six DS3s become one DS4.
This step-by-step multiplexing to higher standardized data rates and frame
formats is called the PDH multiplex hierarchy.
There is also a possibility to take all E1s and instead of multiplexing them into
higher multiplexing orders in several steps, you do something that is called flat
multiplexing instead.
This means that you keep the original structure and send the E1s next to each
other across for example a microwave radio link. However this is not following the
PDH standard, so it will only work between systems of the same manufacturer
that are using the same proprietary frame formats.
Synchronous Digital Hierarchy, SDH
We have seen that the data rates in PDH don’t go too high.
When you need higher capacities, you need to change technology. This means
using Synchronous Digital Hierarchy, SDH. SDH is the ETSI name when the
transmission network is synchronous. The corresponding ANSI technology is
called Synchronous Optical Networking, SONET.
Apart from higher capacities, another advantage in SDH is its standardized
management and control functions. In PDH, this is treated differently by each
equipment vendor.
A common way to go from lower data rates at PDH into the higher rates of SDH is that you
take 63 E1s and map them into a Synchronous Transport Module level 1, STM-1 with a
capacity of 155 Mbit/s.
Here we’ll show how this is done in a slightly simplified way.
The SDH STM-1 frame contains a standardized overhead for management in addition to
the actual traffic. The overhead is for management of the traffic itself as well as for the
network elements that send the information.
So when you map an E1 into an STM-1 you start by adding some justification bits and a
low-order path overhead for traffic management. This forms an SDH “virtual container” for
one 2 Mbit/s data stream, called a VC-12. The management of the traffic follows the E1
until you terminate it in the far end. You add a pointer to the VC-12 and then multiply VC-
12s into different groups along the way, first by three, then by seven and then by three
again, in a simplified way. This way you can fit 63 E1’s into what is called the Virtual
Container-level 4, VC-4. Then you add Section overhead bytes for the physical section and
a pointer showing where the payload of the VC-4 starts. This becomes the STM-1.
So the smallest building block even in SDH is the 64kbit/s data stream.
This is not how the frame is sent, but the picture that best represents how
the STM-1 frame looks like is this, where you have nine rows and 270
columns of 64 kbit/s channels. Of course, most of the frame is the VC-4
on the right hand side. The first nine columns to the left are bytes used for
management and frame control. Just as we saw for the 2 Mbit/s E1 frame,
there is also a possibility to erase the structure of the VC-4 and thereby
make it possible to load it with traffic that does not have to be structured,
like Ethernet traffic. It is most common to map Ethernet over SDH by
using the standardized Generic Framing Procedure, GFP.
In ANSI it looks the same, but the smallest frame here is called
Synchronous Transport Signal level 1, STS-1.
It is built up the same way with rows and columns, but here it is only 90
columns instead of 270 and it is only the first three columns for the
overhead and the pointer. The ANSI payload is called Synchronous
Payload Envelope, SPE, instead of the VC-4 in ETSI. You take three STS-
1 to make one STS-3 that has the same capacity as the STM-1, 155
Mbit/s. One common application is to map 84 DS1s into one STS-3.
We have seen examples of how the STM-1 can be organized, either as 63
VC-12s or one VC-4.
There are also other alternatives; instead of using one VC-4 for traffic, you
can divide it into three VC-3s. Each VC-3 can either be used for traffic or
be further divided into 21 VC-12s, so you can have these alternatives for
the STM-1. Any virtual containers can be filled with Ethernet traffic. You
pick which ones and how many of these containers you want to use.
If you use smaller containers, each has a lower data rate, but you can
send them in different directions using multiplexers.
Going to higher capacities, in SDH you can multiplex four STM-1s into one STM-
4 and then four STM-4s into one STM-16 and so on.
This looks like the multiplexing in PDH, but the big difference here is that all
multiplexers use the same clock source. This means that the capacities in SDH
are exact multiples of the four previous capacities since no extra stuffing bits are
needed. Another advantage when all nodes are in synch is that you can now
extract lower level signals directly from higher levels without de-multiplexing. So
you can extract an STM-1 or even an E1 directly from an STM-16 for example.
You don’t need to go through all multiplexing steps as in PDH.
In ANSI you take 84 DS1s and map them into an STS-3. Then you take four of
those to make up one STS-12. Four STS-12 become one STS-48 and so on.
With this, we have covered the basic main points of PDH and SDH and we end
this Transmission Technology Fundamentals, PDH & SDH-course.
Thank you for participating!
Well done
ReplyDelete