Good day. In this lecture we will discuss about SONET. The word SONET stands for Synchronous
Optical Network, SONET in the USA, Canada, and Japan, Synchronous Digital Hierarchy elsewhere.
For example in India we will be calling SDH. So this is a time division multiplexing system
that transmits a constant stream of information.
SDH is actually a successor of PDH. Few years back we used to have a PDH gear in our telecom
infrastructure in the wide area network part that is plesiochronous multiplexing (nearly
synchronous). This business of being a nearly synchronous introduces us a lot of problems
and complications. So, when from this nearly synchronous we went to synchronous, that was
a major achievement as well as improvement of services as we will see later on. In the
PDH multiplexing in which two or more signals are transmitted and nominally in the same
digital way and the significant instance occur at nominally the same time. This was PDH.
When SONET was introduced, it had a number of achievements to its credit. Firstly, it
is a standard multiplexing using multiples of 51.84 mbps that is an STS-1 rate and STS-N
– we will look at these rates. This was used as building blocks. This is something
that is to be understood that why it is that particular rate value is so important. The
point is when you are multiplexing the original source may come from various sources, and
these signals will travel, will get together, will separate out and then mixed with others,
etc. That is possible only when we have in international multiplexing standard and this
international multiplexing standard was first achieved in SONET. What happened was previously
of course the rate which people used were must less and they had all kinds of differences.
As technology grew and different sort of countries on different things came together, they came
together at a certain rate of transmission and this is a basic building block of SONNET.
There is a standard multiplexing using multiples of this particular day. That in itself was
a big achievement.
Secondly, it also first stated the optical signal standard for interconnecting multiple
vendor equipment. the point was previously of course at lower rates they were all electrical
signal standards, SONET has both has electrical signal standard as well as optical signal
standard and in this optical signal standard it was possible to bring together multiple
vendors to agree on to some particular format.
And the third achievement in SONET was extensive OAM & P capabilities. So what are OAM & P?
O is for operation; a is for administration; and m is for maintenance; P is for protection.
Maintaining the system, administrating the system, operation of the system etc., are
much more flexible in SONET when compared to others. What kind of flexibility, etc.,
that we will see. Regarding protection also, as a matter of fact it is so strong in SONET,
that we will specifically discuss this aspect in a separate lecture when we talk about protection.
These are very strong points in SONET and that’s not all.
The fourth one was multiplexing formats for existing digital signals. It’s not that
such a development can take place in vacuum – that means they had some history and the
trouble was that different countries have different kinds of histories. It is not feasible
for a technology to come and say throw away whatever you have been doing and put this.
There will never be forklift upgrade; it is never possible because of cost, practical
considerations, and all kinds of things. So an evolving technology, in order to be successful,
has to bring together previous technologies so that they can merge into this new technology
and that was another SONET achievement. These existing digital signals – these DS1, DS2,
etc., are different multiplexing standards at the low end. By the way there is also DS0
and DS0 rate is our venerable 64 kbps line rate. Do you remember once again that for
voice channels, we require a 64 kbps because of PCM, etc. we have already discussed. So
that is a DS0 rate and these several DS0 get together to form DS1 and so on, and that way,
there is a hierarchy of rates.
Then the fifth achievement of SONET was that it supports ITU hierarchy: E1, etc. so this
ITU hierarchy was more popular in Europe, India, etc. and they had rates like E1, E2,
E3, etc. E1 was something like 2 mbps, and then E2 was 1mbps, and E3 was 34 mbps, whereas
the development in USA was on a different track. They had a tone rate T1, T2, T3, etc.
Their rates were as above: DS1, DS2, etc. What happened was that when SONET got introduced,
these two sort of came together and although they are not perfectly identical – these
SONET and SDH – for most of the part, they are identical; they interoperate with only
very slight modification at the boundaries, which is not very important. That is a great
thing and that means that the same standard is being adopted worldwide, so that any signal
can be transported in any way. If there is an infrastructure, we can transport in any
way to another part of world; there is no problem. So bringing together of these, that
means bridging the Atlantic Ocean of these two standards, which was another good achievement
of SONET.
The next is that it accommodates other applications. The other applications which were not a part
of this kind of hierarchy, like BISDN, that is broadband ISDN, also can be accommodated
in SONET and that way you see SONET was quite flexible; and how this flexibility is achieved
we will see that later on.
Finally it allows quick recovery from failure, talking about protection, etc. So if there
is a failure like a line failure or if there is a terminal equipment failure you can deploy
a SONET in a particular fashion and SONET can recover from this failure and this retransmission,
etc., can take place in a short period of time.
That is very important when you want to give the so-called career great service, where
arbitrary down time is absolutely not acceptable. As I said, we will discuss this separately
in another lecture.
Some of the broad features of SONET and SDH: it was first standardized by ANCI/ECSA, SDH
by ITU–T. So SONET was by this ANCI/ECSA and SDH by this ITU–T. SONET is time division
multiplexing, pure. We know what time division multiplexing is, and we will see later on
how frames, etc. are made up. It is a pure time division multiplexing system. SONET encompasses
optical and electrical specifications, so there are optical specifications as well as
electrical specifications. You know that usually at the user end, quite often things start
at the electrical level and the rates are low.
But as you go more towards the backbone of the network, the rates that are needed at
the backbone start becoming higher and higher and finally at the real backbone it has to
be very high-speed network, and such high-speed networks are only possible through optical
communication and optical networking. Once again, we will see about optical networking
in the next couple of lectures. Our specification, the SONET specification, spans both the electrical
side as well as the optical side, and that is a very good feature of SONET.
SONET uses octet multiplexing, octet means the same thing as a byte that means 8 bits,
so sonnet uses octet multiplexing. They are multiplexed byte by byte. SONET uses extremely
precise timing, something like in 30 years, maybe; SONET has very precise timing and that
is why things are synchronous. And if things become synchronous, then we derive a lot of
advantages out of that. And SONET provides support for operation, maintenance, and administration
(OAM) as we have already mentioned
SONET is actually superior to T3 and T4, etc. with improvements over the T carriers; these
T3, T4 are still in use but they feed into SONET nowadays.
But earlier, they were used to feed into this PDH and these T3, T4 have particular rates
which existed, and their specification left something to be desired. Because of this lack
of synchronicity, handling the signals from different sources is not easy. What could
happen is that when things are not synchronous, but just almost synchronous, then to handle
this “almost” part, you have to do something; you have to incur some overhead; and you have
to incur some complexity. That was the difficulty with PDH; in SDH or SONET, this is eliminated,
and we get better transport performance. Then, we have the ability to identify sub streams.
This was another advantage of SONET over PDH, which is that a particular user uses may be
using a very small kind of bandwidth – small in relative sense – and then, as more and
more users, as I said as all these data streams or communications streams come towards the
backbone of the network, the pipes tend to get fatter. That means, we need faster and
faster communication. So between say two points in the backbone, there may be a very fast
communication going on and then after going to some other hops, this will again diverge.
SONET has this ability that different streams can get together, travel for some time, and
then again diverge. So the ability to identify sub streams is very important, and that is
also allowed in SONET, which was more difficult in the RDR system. And of course international
connectivity, as I said that it breached Atlantic and that was great. It enhanced control and
administrative function that was also very good from the point of view of service providers.
We have talked about this seven-layer OSI protocol; where does a SONET SDH really fit
in? SONET SDH goes to the bottom of this. If you remember, starting from the application
layer, we go right up to the physical layer. There are several layers in OSI model, and
there are other models. Anyway, usually the bottom-most layer is always the physical layer.
So SONET really fits into the physical layer in some sense.
So what would happen is that the layer just above the physical is the data link layer,
may be, or layer two. So after all this encapsulation, etc. is over through all these six other layers
including the data link layer, SONET takes it over for transporting it from one point
to another. So SDH is placed at the bottom of the protocol stack in the physical layer
along with the fiber. Any IP traffic even if it is the IP traffic of a packet oriented
traffic – and remember that SONET is a TDM system – it can sort of travel within a
sort of TDM transport as they quite often do. So any IP traffic that is destined to
be transmitted across a fiber-based SDH network will be framed by a layer two protocol before
being ready to take its orders from the SDH equipment.
These are some of the multiplexing standards – I have not given all of them I just indicate
some of them. If you remember as I mentioned DS0 is a 64 kbps channel and 24 of them constitute
a T1 line. So T1 rate is approximately about 1.5 mbps; 4 T1 gives T2 and 6 T2 gives T3
and so on. Similarly 30 DS0 –this is a European system – gives E1 line. So E1, if you remember,
is about say 2 mbps: 4 E1 gives E2; E3 is a 34 mbps line. And then I suddenly jump right
up to this thing called OC3; this o is for optical. So this way, this 155 mbps is 3 of
the basic STS 1 rates that I mentioned earlier; I will come to this later on. So these are
some of the standards. There is a whole hierarchy of standards; for example, this name SDH is
also synchronous digital hierarchy, this is a hierarchy. For the SONET, the basic rate
is STS 1 that is synchronous transport signal level 1, and the speed is 51.84 mbps. This
is designed to carry what was DS3 RDR or a combination of DS1 c, and DS2 etc. As I said
a combination of different streams can flow through a SONET pipe or SONET infrastructure.
So that is good and that means DS3 is a fat pipe or DS3 is almost the same as STS-1. So
it is a fat pipe through which multiple pipes, say may be DS2 or DS1, etc. may travel.
And this net goes up to STS-N, whereas synchronous transport signal level is N; so this has a
speed of N into 51.84 mbps designed to carry multiple STS -1. I mentioned that these are
byte multiplexed STS-1 means 1 byte from one source and another byte from another source
and so on.
Fundamental SDH frame is STM -1; SDH if you remember is the other standard, which came
from Europe and they sort of came together and that is what we are talking about. SDH
frame is STM -1 synchronous transport module and the SONET version is OC -3, that is, optical
container, each providing 155 mbps. So when we come to this rate these 155 mbps
OC 13, different rates etc. and different systems are culminated here, at this 155 mbps,
almost 155. STM 4 provides four times the STM -1 capacity, STM 16 provides a further
fourfold increase, which means STM4 may be about 620 mbps, and then, if you go to STM16,
which is four times that of about 2.5 giga bit by s, then you have STM 64, which is about
10giga bit by s. So all these rates are there; that means, from this point onwards, these
two streams have converged and we are going to higher and higher rates in a sort of universal
fashion, which makes things easy across the world.
It is worth noting that the internet working between SDH and SONET systems is possible
at matched bit rates; for example STM4 and OC12; so they interoperate. A slight modification
to the overhead is required as they are structured little differently so there will always be
a little something; but anyway that is not very serious. So they do interoperate.
We have seen the SONET electrical hierarchy; now we look at the SONET optical signal hierarchy:
OC-1 is the optical career, level 1; it carries STS-1; OC 3 carries STS-3 or STM -1 at 155
mbps; OC-N optical career level N.
OC - N as I mentioned is an optical carrier, which uses N into 51.84 mbps, so OC - 48 is
about 2.4 gbps; overhead percentage is about 3.45%. OC signal is sent after scrambling
to avoid a long string of 0s and 1s to enable clock recovery. This is a small technical
point; that means in order to keep the whole thing synchronized, the SDH units use the
transitions which happen when there is a 1. So the point is that if there is no 1 for
a very long period in the data stream, then the clock on one side may drift relative to
the clock on the other side; that is always possible. So we try to avoid long streams
of 0s in this SONET or SDH, and we do that by scrambling the data from various streams,
etc., or descrambling them. The idea is that even if one of them is sending a long stream
of 0s, there will be quite a few 1s from the other streams and then the clock will be maintained.
An STS -N is synchronous transport signal electronic equivalent of optical carriers.
OC 3, OC12, OC 24 and OC 48 rates are common in telecom circuits – if you remember OC
48 is 16 times of OC 3; that is, 16 times 155 mbps, which is about 10 gbps. Up to 10
gbps is very common these days. Actually right now, with DWDM systems, OC 192 rate is already
in operation, and OC 768, which is 40 gbps, is being talked about. So that was another
disadvantage earlier that this digital hierarchy of standard rates did not exist beyond a very
small rate – I mean small in today’s comparison. But now we have an extended and open system
where, as technology improves, we can always go for higher and higher rates; so from OC
3, which is 155 mbps, we can go to maybe OC 192, which is 10 gbps or OC 768, which is
40 gbps that we are talking about now.
How do you use these high-speed links? These high-speed links of course have to be on fiber
– we can look at details of fiber later on, but please note that in practical application,
an SDH line system will have a multiplexer that takes its inputs from a variety of sources
in different layer 2 data formats. So here we are talking about these different signals
coming in the electronic domain, and they are coming from a variety of sources, may
be coming with different layer 2 data formats. These are aggregated up to form frames at
a line rate of system, for example up to STM 64 for a 10 gbps bit rate system.
Now these frames at 10 gbps cannot be pumped anywhere. It is very difficult to pump it
on a copper. So these frames are transmitted out onto optical fiber links. There is a possibility
of multiple SDH multiplexers to each give out one wavelength of a WDM system. As we
will see later on, this WDM stands for Wave Length Division multiplexing; this is some
form of frequency division multiplexing. I mentioned about it when I talked about frequency
division multiplexing. In fiber optics, we talk about wavelength multiplexing so it is
possible that one multiplexer is feeding into one wavelength, another multiplexer is feeding
into another wavelength, and all these different wavelengths are traveling together in the
fiber.
At the end of the system, there will be an SDH demultiplexer on the other end, just as
we have a multiplexer on one side. Naturally, you have to have a demultiplexer on the other
side that now accesses the individual data streams from the STM 64 frames as required.
So STM 64 is carrying lots of frames in a very short time; they are sort of separated
out and then fed into slower streams down the line. So there may also be an SDH add
drop multiplexer with the ability to remove and insert lower bit rate streams from the
signal.
Alternatively a digital cross connect may be present with the ability to switch individual
VC4s. Well, this is virtual container four, which is another concept, we will talk about
later. So between different fiber links there is a digital cross connect; if you have the
digital cross connect in the optical level, the advantage is that you need not go into
the electronic domain at all. So the advantage of not going into electronic domain is that
you are handling a huge, very fat, pipe; that means, a large number of channels, and you
can just switch them from one fiber to another fiber simply in the optical domain without
doing any kind of processing; and that is always an advantage.
We will talk about some SONET terms now; for example, envelope. This envelope is the payload.
Basically, after all encapsulation, etc., you remember that finally near the bottom
we have this layer 2 and this layer 2 protocol will encapsulate it and then hand it over
to SONET at the lower level, maybe at the physical level. So whatever this layer 2 hands
over to SONET is the payload; the rest of it are kind of system overheads – payload
plus some end system overhead also goes into this payload. So these together form what
is known as the envelope; this is a SONET term. Other bits and bytes which are used
for management, that means OAM and P portion, goes as the overhead of SONET. Then there
is the concept of concatenation; that means, unchannelized envelope can carry super rate
data payload, for example, ATM, etc. So, the method of concatenation is different from
that of T carrier hierarchy; we need not bother about it at the moment.
Then there are some nonstandard functional names in SONET, like
TM is for terminal multiplexer, also known as line terminating equipment or LTE. These
are ends of point-to-point links. ADM is for add drop multiplexer; we have mentioned this.
DCC is for digital cross connect wideband and broadband; MN is for matched nodes and
D plus R means drop and repeat, etc. Anyway, these are just some terms.
Now let us come to some important concepts in SONET namely: section, line, and path.
What is a section? I will just show you figure first and then come back to this.
Please look at this figure: we have some multiplexers. So as the figure shows, we have a multiplexer
in this side, another is an output that fits to another multiplexer. This multiplexer is
going in this direction and after some time, the signal becomes weak. So we want a repeater;
what is a repeater? A repeater is something which boosts the signal strength.
So there is a repeater, then it travels some more distance then there is a repeater again
and then it travels some more distance and then on other side we have the corresponding
demultiplexer and then it fits into the other de-multiplexer. From repeater to repeater,
we call it a section. So from repeater to multiplexer, this is also a section. So multiplexer
to repeater, repeater to repeater, these are called sections. And then, from multiplexer
to multiplexer, we call it a line. At the repeater, nothing happens excepting the signal
is cleaned up.
The signal may be boosted or there may be other cleaning operation, synchronizing operation,
etc., that may be done at the repeater; but as such, the signals which are traveling here,
the same set of signals are traveling here. At the multiplexer, of course, some of the
signals may go off in another direction; some signals may go in some other direction, etc.
So at the multiplexer, there may be a convergence or divergence, depending on which way the
signal is flowing. That may happen at the multiplexer, so from multiplexer to multiplexer,
we call it a line; and then from the end user point to end user point, we call it a path.
Look at this once; the portion from a multiplexer to a repeater is known as a section or it
could be a repeater to a repeater also; the portion from a multiplexer to another multiplexer
is a line. The portion from source to destination multiplexer is a path; below path line and
section is the photonic sub layer; that means photonic sub layer is whatever is happening
in the optical domain, and we are not discussing that at the moment.
Sections are bounded by repeaters or multiplexers that terminate the line; lines may carry several
tributary signals and are bounded by multiplexers, a path goes end to end between terminating
multiplexers.
Each STH frame lasts 125 microseconds. As I mentioned, this 125 microseconds time period,
time epoch, is sort of sacred in this whole domain because 125 microseconds is what is
required for a DS0 channel. Remember this is a time division multiplexing, which means
that if you have a 125 microsecond kind of slot, then some of the DS0 bytes can take
these bytes. Actually if you have to take it as 8 kbps
and if it is 8 kbps, inverse of that is 125 microsecond. So if you have a 125 microsecond
slot, if 1 byte travels in this frame, then that is enough for 1 DS0 channel. In SONET
we have very sophisticated and very fast equipment; that means this is a time division multiplexing
system; within this 125 microseconds, not only 1 byte can go but lot of other bytes
can go. That means a lot of channels can travel together in this 125 microseconds frame. This
is the idea. So each STH frame lasts 125 microseconds; how many bytes are going in there depends
on whether it is STS -1 or STS -2 or STS –N, etc.
So 125 microseconds as I mentioned is 8000 frames/s. STS -1 frame has 6480 bits or 810
bytes. That means in one, 125 microsecond slot or frame, we are putting in 810 bytes.
Theoretically, of course, that means it can carry 810 DS0 or voice signals; actually it
is not 810, it is lesser than that because a number of these bytes are used for different
types of overheads. We will talk about this. We have these 810 bytes; the octets are understood
in terms of a table of 9 rows and 90 columns; so let us look at this figure.
We have a SONET frame or an SDH frame, which has 9 rows; you can see the 9 rows on this
side and then 90 columns, total 90 columns. Out of these 90 columns, 3 columns have been
shown in yellow. These are sort of used for overhead and these 87 columns are used for
payload or for envelope. If you remember, the envelope contains the payload as well
as little bit of overhead, which we will come to later on. This is how after every 90 bytes,
we come back to again another 3 bytes of this overhead. This is how it is to be understood:
the first 3 columns contain transport overhead and TOH has 9 rows by 3 columns, that means
27 bytes, which is subdivided into section overhead SOH (section overhead), 9 bytes,
3 rows of 3 columns; LOH, that is, line overhead, which is 18 bytes, that is, 6 rows of 3 columns.
So we have section overhead and we have a line overhead – remember we have these three
concepts like section, line, and path. We have not talked about path overhead.
There is some path overhead and it goes into the envelope; so there is some path and as
far as these things as line and section are concerned, these are the overhead bytes. Just
to clarify why do we require the over bytes – the point is that the multiplexers or
the repeaters have to have some communication between them in the control plain so as to
give you this OAM capability. For that some information needs to be sent or exchanged
between the two points; anywhere there is a section, the section overhead would consider
those things which are central to the section about the signal strength and other kind of
things; line overhead maybe would contain something else and similarly path overhead
would contain something else. But these are required for these OAM capabilities that we
have in SONET.
Let us look at these overheads separately; first section overhead, which defines and
identifies frames and monitors section errors and communication between sections terminating
equipment. So these are its functions: it identifies frames; monitors section errors
– if there are errors, it monitors section errors; and communication between section
terminating equipment, maybe two repeaters or a repeater and multiplexer, and so on.
Line overhead locates first octet of SPE and monitors line errors and communication between
terminating equipment. We will come back to this locating of the first octet of SPE. This
is a very interesting feature and we will talk about this separately. Previously we
talking about section errors; so line errors and communication between terminating equipment,
etc., is taken care of by the line overhead. Apart from that, line overhead contains this
pointer, really, which points to the first byte of the SPE.
And then there is a path overhead; and as I said path overhead is really inside the
envelope and we will look at all these later. Path overhead verifies connection path; you
remember path means from end to end; that means from the end to end multiplexer is a
path. Whether the connection has been established or not, it monitors path errors, receivers’
status, communication between path termination equipment, and so on. This is the POH . We
talked about the synchronous payload envelope or SPE that I was talking about. That is,
the other 87 columns hold the SPE (synchronous payload envelope). So SPE has 9 waves by 87
columns, which are divided into path overhead and payload, which means the path overhead
goes along with the envelope that is in the SPE, whereas other overheads have separate
bytes or separate columns associated with them as shown.
Now this SPE does not necessarily start in the column 4, which means that the SPE does
not necessarily stay within one frame; these are two very important points in SONET. The
point is that although you have these 87 columns, actually the data may start getting transmitted
at some arbitrary points inside those 87 columns. What is the idea? I mean why do you want to
leave something and then only start from the middle? The point is that if there are some
kind of mismatches of late, etc., if everything in the world were absolutely synchronous,
all activities and all equipments, etc., then you could have started from the beginning.
But that is not the case and this is where we absorb this kind of variation and this
gives great flexibility to SONET, which was not there earlier. And the other interesting
thing is that the SPE does not necessarily stay within one frame, which means that the
SPE may start in one frame and then end in another. We will just look at a diagram of
this; let us have a diagram of this.
You see the SPE in light green color; it really starts from somewhere. I mean somewhere after
leaving some of the rows: it starts here, and the path overhead is somewhere here, and
there are two frames here. So SPE is really spanning both the frames.
SPE is not frame aligned; it overlaps multiple frames; avoids buffer management complexity
and artificial delays. Whenever there is something to send, you can just send it in the envelope;
just put that pointer to that yellow edge, so that yellow edge will point to the first
byte of the SPE. It allows direct access to byte synchronous
lower level signals, for example, DS1, with just one frame recovery procedure.
These are the advantages of the SONET frames. This is one frame coming in may be 125 microseconds;
this is the next frame; and SPE, as I have already shown, can overlap. I mean it may
start somewhere within the first frame and then continue in the second frame in this
fashion and then be over here. Actually after this, some other envelope may come in over
here.
Now of course where is the path overhead? There are two fields, H1 and H2 in LOH; LOH
means line overhead, which points to the beginning of the path overhead. Path overhead beginning
floats within the frame; 9 bytes that is one column may span frame along with the SPE;
it is originated and terminated by all path devices; and this gives you end-to-end support.
These are the features of path overhead. The point is that if you remember the path is
end to end, that means it is close to the end users; just as the end user may start
somewhere arbitrarily in-between, a path overhead also goes along with the SPE and it starts
over there and at LOH, we keep a pointer to this path overhead.
Just as some of the equipment that we use in SONET, one of the most important of these
is the add drop multiplexer. They are important because at certain point in the network, what
might happen is that there are some sources which want to send into the network. They
will sort of go so there is this SONET equipment, which is ADM let us say, and SONET stream
is flowing let us say like this. There may be something that wants to upload and travel
along with this thing. At the same time, this may be the destination location for some of
the other signals which originated elsewhere; they have to be dropped here. So some signals
have to be dropped, some signals have to be added. So this multiplexer can handle that
and that is very important.
That is why they are called add drop multiplexers. This stream is itself of course flowing at
a tremendous rate, whatever that rate is. So SONET SDH is a synchronous system with
the master clock accuracy of 1 in 109, which you will see is highly accurate. It shows
when you come in some kind of CCM clock somewhere and then there is a protocol for distributing
and maintaining this clock over the entire network. Frames are sent byte by byte and
ADMs can add drop smaller tributaries into the main SONET SDH stream and I have explained
how that is done. Within that frame you can send lot of bytes; you can take out some of
the bytes and add some of the bytes. That is how you take out some of the smaller tributaries
and add some of the smaller tributaries.
Digital cross connect, which is an optical layer equipment, is also very important. It
cross connects thousands of streams and software control, so it replaces patch panel; that
is a good thing about the digital cross connect and a software control is coming where the
control is coming from the control plane of the switches. You can connect the streams
from may be one fiber to another; it handles performance monitoring, PDH SONET streams,
and also provides ADM functions; that means add drop multiplexing functions.
Finally we have this concept of grooming in SONET. Grooming means, we group the traffic
in some format. So you want to keep this group in one particular way; it could be that there
is a one group of streams for whom you want to give higher priority or you want to give
higher quality of service. So you have to group them together. Similarly there may be
multiple groups; so it enables grouping traffic with similar destination, Quos, etc., which
is a part of grooming. It enables multiplexing or extracting streams also – that is also
part of grooming. Narrow wider broadband and optical cross connects may be used for grooming.
If you look at this figure, you have this narrow band, this SONET layer and optical
layer. In the narrow band, we have this DS0 grooming and then in the DS1 grooming, there
is a white band and then the broadband DS3 grooming – so the rates are going up, starting
from the 64 kbps, it is going up. When you are going up for the STS 48, you are in optical
domain; that means STS 48 is STM 16, so that is a high rate. The point is that, at that
rate, most probably, you are well in the optical domain. Then, finally, you can go to all optical
domain; that means wavelength, waveband, and fiber grooming – there are different levels
of grooming, depending on what you want to do. Lastly we will just talk a little bit
about virtual tributaries or containers. We have already talked a little bit about it.
This is the opposite of STM; actually in some sense this is called sub multiplexing; that
is, different streams coming together to form one very fat or very fast stream.
This is the other thing – how do we, sort of, differentiate these sub streams within
this, which has to do with sub multiplexing? STS -1 is divided into 7 virtual tributary
groups, SDH uses the term virtual containers or VCs. We talked about VCs; we just mentioned
what are called VTs or virtual tributaries in SONET lingo. So we have 7 virtual tributaries,
12 columns each, which can be subdivided further. You see that there are 12 columns each, with
7 virtual tributary groups – we have got 84 columns and these 84 columns are out of
the 87 you have in STS -1.
VT groups are byte interleaved to create a basic SONET SPE. So this VT groups are byte
interleaved. They may be again extracted from each other. VT 1.5 is the most popular, quickly
accessed, T1 line within the STS-1 frame. So the idea is that you have a T1 line, which
is approximately 1.5 mbps line, which is coming out of your small business, and you have a
1.5 mbps line. So that is your bandwidth requirement, you want to connect it to a distant location
somewhere. And you do not want your thing to get mixed up with others. At the same time,
as a small business you cannot have infrastructure of connecting to another location which is
wide apart. So you will go with this public infrastructure or public switched tele PSDN
network or whoever is maintaining this communication equipment.
Usually telecom people maintain it in most of the places. Anyway, they have a sort of
fiber going from one place to another, which contains very high-speed links. What you want
is your T1 line should join them, sort of get transported over the distance and then
go and feed into another T1 line at the destination. That is what you want. You want your T1 line
to sort of have a separate sort of existence – just like in a compartment, we have different
passengers. Passengers have their own individual entity but together they are packed into one
compartment and then they travel. Similarly your T1 line is going to ride onto to this
very fast stream and travel to the destination. So VT 1.5 gives your T1 line.
How do you find out about the difference? How do you separate them in the SP? The point
is, you require one more level of pointer used to access it. You can access a T1 with
just a 2-pointer operation, first from the LOH – you remember, you go to the SP, just
like that. Similarly, you go to the different tributaries or different containers using
just one more level of pointer. This flexibility was not there earlier; so it was very complex
to do the same function in DS3. For example, accessing DS0 within DS3 requires full demultiplexing,
stacked multiplexing, etc. So you require full demultiplexing; that is not required
in SONET. The point is that the other streams may go; you know where in that frame your
bytes really are traveling for the stream or for the container or for the tributary
that you are interested; you just extract it, others keep on traveling as they are.
So you do not demultiplex the whole thing and that gives a great advantage of add drop
multiplexing.
This is just a figure showing that you can have various types of lines, all feeding into
the same infrastructure. You may have what we have put over here: DS1, which is 1.544
mbps, E1, 2.048 mbps, DSIC DS2, DS3, ATM .48.384, E4, which is 139.264 mbps, ATM is about 150
mbps, etc. They are sort of traveling; they are getting in different containers. From
VT 1.5, different tributaries, that is 1. 5236 etc., form a VT group and ride on a higher
strength or higher speed stream.
Just as I said, these are sort of identified through a pointer; so we have this transport
overhead. We use some bytes for that out of those 87 columns we have. So we use some columns
of that and then we put a pointer, which gives to the STS payload pointer. Then there is
a VT pointer, virtual tributary pointer, and this much is the VT SPE within the overall
STS-1 SPE, which is the payload. Even now SONET is the most widely used technology in
wide area networking that is existing today. Of course, as you know, as technology grows,
maybe we will go out of SONET.
People are already talking about going out of SONET because one disadvantage of SONET
is that its equipment tends to be expensive. Well, expensive compared to what we think
today. What is cheap today and what we think is cheap today may sound very expensive tomorrow;
that is how the technology grows. So people are talking about direct transport over the
optical layer, etc. May be we will touch those aspects later on. But all that is still in
a sort of experimental stage and on the field, actually, SDH or SONET equipment is almost
everywhere; all types of telephone companies are connected through that and major service
providers use this as a means of transport. Thank you.
Good day, so today we will be speaking about fiber optic components and fiber optic communication
as might of heard this lecture as well as the next couple of lectures, we will concentrate
on fiber optic components. We have looked at some of the physical layer components of
fiber optic systems before so we will sort of quickly review that, some of the stuff
we will be talking about today is going to be common and then from that point.
we will take out take it up into WDM systems, how wavelength division multiplexing is done
and how systems are handled in fiber optic domain, this fiber optic domain happens to
be very crucial because a lot of traffic in terms of volume may be as much as forty to
fifty percent, actually goes through the fiber as days are going by and as more and more
demand for bandwidth is coming up fiber optics is becoming more and more important , we will
be talking about fiber optic components today.
In fiber optic component of course the basic fiber is there we have already talked about
it, so we will talk little bit more about this then we have light source and receivers
on two end because we know that in fiber optic cables light is the carrier of information
then we require these different components like amplifiers, couplers, modulator, multiplexer
and switches so we will look up at these components one by one and then we will start our discussion
on wavelength division multiplexing.
The next set of components are multiplexers filters gratings, just talk little bit about
it ,if you look at this wavelength , these are all; wavelength selective , devices multiplexers,
filters , these are wavelength selective devices in a wavelength filter and what we want is
suppose ?1, ?2 etc so many are coming, I want only ? 1 out 2 ?3 ? 4 etc are absorbed or
something where as if you are a multiplexer I want the difference this ? coming in different
lines, I want all to be mixed together and use the same line, these are wavelength multiplexer
so application could be particular wavelength or a particular wave band selection.
Wave band is nothing but some contiguous operating wavelengths which all are side by side, if
you remember that in the operating window what ever be that 1550 what ever may be the
window you are using there you can have a number of ? all side by side, there is a guard
band between each of these operating ? so where the guard band that is given by the
i q t has specified, how much guard band etc you have to have but so you can have large
number of ? all group together in the same window. Aband out of that means a bunch of
sequence is out of that you can short select instead of selecting only, that is wavelength
band selection static wavelength cross connects and OAM is optical add drop multiplexers,
you have come across this term optical add drop multiplexers in the context of Sonet
but optical domain we require optical add drop multiplexers, we will come to that. Equalization
of gain so that is another application filtering of noise ideas used in laser operation and
dispersion compensation modules etc, these are the different applications one of the
standard wavelength selective component is arrayed waveguide gratings ,we have seen this
before…
