Computer Networks

Prof: Sujoy Ghosh

Department of Computer Science and Engineering
Indian Institute of Technology, Kharagpur
Lecture-18
Satellite Communication

Good day! In this lecture we will be discussing one very important component of worldwide
communication, which is satellite communication.

(Refer Slide Time: 00:59)

Communication through satellite has been going on for quite some time now. Satellite
Communication is fairly mature technology although some improvements are taking place.
Satellite Communication is an important component of worldwide communication
infrastructure. We will also see how specifically MAC, which is Media Access Control, is
handled in Satellite Communication. This is important, since we obviously need MAC
because the communication medium is the electromagnetic field around us, which is common
to everybody.

We now talk about satellite communication. Satellite can be looked upon as a big microwave
repeater. Repeater is something, which repeats the signal. It takes in the incoming signal,
amplifies it and then sends it back. It contains several transponders. Transponders listen to
some portion of the spectrum. Each transponder is listening to some portion of the spectrum
so that several transponders together can listen to the wider band of the spectrum. It amplifies
the incoming signal and broadcasts in another frequency back to earth. Satellites are up in the
space; they take the incoming signal, amplify it and broadcast it back. So satellite has to
broadcast in a different frequency so as to avoid interference with the incoming signals.
(Refer Slide Time: 01:59)

(Refer Slide Time: 03:32)

This can relay signals over long distance – this is one strength of satellite communication,
because, if you send it all the way up and then when it is sending all the way down, and if it
does so at an angle, the signal can reach a very long distance. There are different kinds of
satellites and the most commonly known ones are geostationary satellites because they are
above the equator at a distance of 2300 miles approximately and are in the geosynchronous
orbit. They travel around the earth in exactly the time the earth takes to rotate.
(Refer Slide Time: 03:57)

We have an uplink station from where some signal is going and then it is being listened to by
one particular transponder in the satellite. It amplifies signal and sends it back to earth where
the downlink station (another dish) which is facing the satellite and tracking the satellite
receives the signal over here. This is the just reflected part which then gets on the horn. This
signal is taken down and amplified. Then we can use it.

Earth stations communicate signals to satellite on an uplink. They are called earth or base
stations on the ground. The satellite then repeats these signals back to the down link. This
feature is attractive for distribution of TV programmes. It is known that many of our
television channels(unless coming through cable) are coming from satellite, which is beaming
back the signal over a very wider area and all the receivers around can receive the signal and
then amplify it and use it. So it is very popular for such kind of applications like distribution
of television programming.

There are other applications. These signals are used to transmit signals and data over long
distances for weather forecasting, television, internet communication and global positioning
system. These are the various applications of a satellite. The space orbit allows more surface
coverage.
(Refer Slide Time: 05:26)

(Refer Slide Time: 05:37)

The spectrum is usually used by the satellite and is divided into sections; each of these bands
has names like C band, Ku band or ku band and Ka band or ka band. The C band may have
four downlinks and six uplinks; C band is from 3.7 to 4.2 GHz. This part of the spectrum is
reserved for the C band communication through satellite; that is, 3, 7 to 4.2 GHz for
downlink and 5.925 to 6.425 GHz for uplink. There is a 0.5 GHz bandwidth for downlink and
another 0.5 GHz bandwidth for uplink. Since the uplink is at a higher frequency, you can
have more channels over there.
The capacity is not very high by today’s standard, especially, if you compare with fiber, this
is really low. But at the same time, at one point of time, when transoceanic fibers were not
there, satellite was the chief medium of communication across continents. The capacity is
low but still it is useful; however, terrestrial interference is a problem because when the
weather is bad or when there are other kinds of extraneous sources of some electromagnetic
noise, etc., these interfere with the signals. But still it has got a lot of strong points; that is
why it is still an important component of communication these days.

(Refer Slide Time: 08:08)

Next set of transponders come in Ku band. They can accommodate greater number of
transponders – 12 on the downlink to 14 on uplink. Rain interference is the problem here. Ku
band is from 11.7 to 12.2 GHz and 14.0 to 14.5 GHz. In this band, rain is a problem though it
has higher capacity and is less crowded than C band. C band is very crowded, if you consider
people with a sort of dish antenna for receiving satellite signal – I am not talking about the
base station – those using 1-meter antenna are possibly using C band. For Ku band, the
antenna size is smaller and pizza shaped at something like 18 to 20 inches. They have a
higher bandwidth; however, rain and terrestrial interference are the problems faced in C band.
Different parts of electromagnetic spectrum are most susceptible to interference and noise,
etc., but Ku band has higher capacity. The Ka band is at an even higher frequency – 19
downlink and 29 uplink transponders are needed. The equipment needed for this is quite
expensive. It is from 17.7 to 21.7 GHz and 27.5 to 30.5. It covers a greater bandwidth, but it
is still not very widely used.
(Refer Slide Time: 10:03)

Satellite can be used for point-to-point transmission, to transfer large volume of data; voice
data and communication and for video conference. Satellite is not just for broadcast; satellite
can be used for point-to-point communication. You may look it as point-to-multi-point
communication. So we can say that point-to-point and point-to-multi-point communication
are supported. Also, standard broadcast is supported. Point-to-multi-point communication can
be data communication, internet, and video conference. Broadcast services include television.

(Refer Slide Time: 10:53)

The advantages are that you can reach a large geographical area and have a high bandwidth;
it is cheaper over long distances. It is certainly cheaper than wiring it up, but if you are
covering long distance, because of its inherent high capacity, the economy is in favour of the
fiber optic cable and it is almost even in some cases. But in many situations satellite retains
its advantage. It can transmit to places where the cable cannot reach. These are applications
where satellites have an undoubted advantage over any wired kind of system. It may be a
remote area, which is not very easily reachable and road or other kind of communication is
not well established to take cable over there. But the satellite, since it is sitting high up in the
sky, can cover any area without any difficulties. Hence this is a very strong point in favor of
satellite. It is especially useful for technology deployed at multiple sites regardless of
location, like mobile technology. Nowadays the first thing people think about to be mobile is
the cellular phone, which uses the nearest base station, which is usually wired. But in some
places where it cannot be wired, it may go through the base station and may be connected to
the backbone through a satellite. That is one thing; and if you want 100% roaming all over
the earth wherever you go, then you may not have a base station around the place where you
are. In that case, some satellite or the other would be visible to you and you can communicate
through the satellite. So mobility and accessibility to places which are difficult to reach
otherwise are very strong points of satellite and of course standard communication is also a
point.

(Refer Slide Time: 13:35)

The disadvantage is the high initial cost. You have to build the satellite and then to put it up
in the space through some launch vehicle involves high investments. But laying fibers all
over is also costly so we have to work it out on a case by case basis and determine which one
comes out to be better. We cannot do much at the physical level about susceptibility to noise
and interference because the physical media is such that all kinds of noises are being
generated all over the place and crowding this shared medium.

So we have to handle it in some other way at a higher layer. Then there is a propagation
delay; and this is a significant disadvantage compared to terrestrial communication for many
satellites, especially the geostationary satellites. We will come back to this point later on.
Geostationary satellites are thousands of miles above earth. So although your electromagnetic
signal is travelling at the speed of light which is very high but even then to go all the way up
to the satellite and then come all the way back down takes significant amount of time and this
has lot of implications. It has the implications on the MAC and it has the other implications
like delay and may be quality of service in some cases so this is an issue, a potential
disadvantage and security can also be an issue since this medium is open to everybody.
Whatever you are communicating, anybody else can listen on to it. If you are trying to send
some very sensitive data through satellite and you do not want other people to listen to what
you are sending, then you have to take some other measure like encryption, etc. We will talk
about encryption much later in the course.

(Refer Slide Time: 156:31)

There are different types of satellites and they have different types of orbits, circular or
elliptical orbits. The circular orbits will centre at the earth’s centre. There is the elliptical
orbit with one focus at earth’s centre. There are some equatorial orbits above earth’s equator.
This is quite common and necessary, especially for geostationary satellite. It is necessary that
they go around the equator but then you could also have a satellite pass over both the poles.
Other orbits are referred to as inclined orbits.

(Refer Slide Time: 17:21)

The altitudes of the satellites and their distance from the earth have significant implication in
terms of the time it takes for the signal to travel. We have three different classes of satellites:
geostationary orbit satellites, GEOs; medium earth orbit satellites, MEOs; and low earth
orbits, LEOs. Out of these, GEO is the most common one. GEOs go around the equator and
have a high bandwidth. But they also require high power – this is an important point – and
long latency. These are the important issues. When you are trying to communicate with some
satellite, which is very far away, then your transmitter has to be strong enough so that the
signal reaches the satellite. If the transmitter power has to be large then you require lot of
power. In order to power this, it may be difficult to power it with a battery. That kind of
battery power may not be enough to reach a geostationary satellite. The mobility becomes
difficult from this power angle. So this is an important issue.

(Refer Slide Time: 17:46)

MEO has a high bandwidth; medium kind of power; and medium kind of latency. LEO has
low power and latency, but you require more number of satellites. They have smaller
footprint. We have VSAT, which means very small aperture. When we talk about VSAT, we
are not actually referring to any satellite. We are referring to the ground equipment that we
use and VSATs have small apertures, which private WANs can use with smaller antenna. But
if they are using C band or Ku band they use antenna which has a dimension of either 1 meter
or 18 inches. But for the main base station of the satellite, usually a much larger antenna is
put in place. We will first talk about geostationary satellite. This is the most common type of
satellite today and it is in a circular orbit. We say it in miles about 23,000 to 22,000 odd; and
in kilometres it is 35,000 odd kilometres above the earth in the equatorial plane these
satellites remain in the same position over the earth as it rotates.
(Refer Slide Time: 20:42)

(Refer Slide Time: 21:07)

As the earth is moving, the satellite is moving along with that. From some point on the earth,
it would appear that the satellite is stationary over its head at all times. That is why this is so
sacrosanct about this geostationary satellite, this distance of 35,785 kilometres. You can
always calculate this distance by finding out about the centripetal force and the earth’s
gravitation etc. You can put it in that equation and calculate that this is the exact distance at
which, if you put a satellite with a particular angular velocity, which is the same angular
velocity as that of the earth, what will happen is that to the people directly under it, it will
appear as if the satellite is stationary. So this is the good thing about this geostationary
satellite and this is why this distance is so fixed.
(Refer Slide Time: 21:57)

From such a long distance, if you send a beam which has a reasonable solid angle over here,
it will cover a large portion of the earth. As a matter of fact, it has been calculated that with
three or four satellites you can cover the entire surface of the earth; but of course, three or
four satellites would not be sufficient to handle the bandwidth that we require these days.
Furthermore since this distance is so fixed, there is only one band in space where the
satellites can be parked. The other point is that if two such satellites are very close to each
other, the signals will start interfering with each other. So there has to be a minimum distance
between two geostationary satellites. The geostationary orbit and these parking slots are
internationally decided, but these parking slots are quite crowded. Some nations may not put
the satellite but would have reserved some parking slots. These parking slots are quite
crowded today with so many geostationary satellites up in the space.

(Refer Slide Time: 23:38)

These have coverage of about a fourth of the earth and have good tracking properties. That
means you can track it very easily but their signal weakens over great distance and the
propagation delay can be large. The propagation delay we are talking about is of the order of
0.24 seconds. Usually we talk in terms of milliseconds but here we are forced to talk in terms
of seconds. It may also be hard to get coverage at the Polar Regions because this
geostationary satellite has to be over the equator; it is difficult to get coverage at the far
northern and southern hemispheres.

(Refer Slide Time: 24:53)

GEO satellite provides universal connectivity in its footprint. Its footprint means the area of
earth, which is covered by one particular transponder; let us say on one particular satellite. So
it covers all that area at the same time. So now within its footprint it covers universal
connectivity. From any particular point, you can communicate with the satellite and possible
satellite parking slots are quite crowded. Wide beams are circular, whereas spot beams, which
are more focused, are elliptical. Apart from broadcasting, they may also be used with VSATs
for point-to-point communication. We will come back to this point later.

So they are at this kind of distance that requires large transmitter power making them large
and expensive. There is considerable space delay and large cell size, which means smaller
number of channels. This is one particular point and we will come back to this point in
greater detail when we discuss terrestrial wireless communication. But the point is that a
satellite has some amount of bandwidth which is assigned to it which it can handle; now
within that bandwidth for communication, namely, the voice channels require about 64 kbps
rate. One particular transponder can handle 800 such channels.
(Refer Slide Time: 25:08)

If the footprint of the satellite is very large, that means it is covering a large number of
people. But all these large number of people are constrained with those 800 channels. A large
cell size also necessarily means a smaller number of channel densities on earth.

(Refer Slide Time: 26:48)

A typical satellite has 12 to 20 transponders, each with a 30 to 50 MHz bandwidth; a
transponder can carry about 800 voice channels. FDM, Frequency Division Multiplexing,
was used in early satellites. Nowadays TDM is also used; as a matter of fact, a mixture of
TDM and FDM is also used. The cost of transmission is independent of distance. Sometimes
this is an advantage. When you are communicating via a communication satellite with
somebody who is, say, 10 kilometres away or with somebody who is 100 kilometres or 1000
kilometres away, the cost is constant over the entire footprint. That may be an advantage in
some cases.

(Refer Slide Time: 27:38)

For some applications this may be useful. Security and privacy pose a problem as mentioned
earlier. So encryption is essential. Mobility is easily achievable and setup time is not
required. If the satellite is in space, then you do not require any setup time. If the satellite is
not there, you have to send it there. GPS is an interesting application we will see.

(Refer Slide Time: 28:14)

We now come to the middle earth orbit satellites. These are used for global wireless
communication coverage. They maintain orbit about 8000 miles from earth. The moment you
come out of the geostationary orbit, the satellite cannot remain stationary over somebody’s
head any longer. Now the satellite will necessarily move over your head so in MEOs and
LEOs what will happen is that you will find that the satellite is coming over your head just as
other stars move. So the satellite will move over in the sky and it will be there for some time
and then it will be out of your reach because it will go down below the horizon. This is what
is going to happen with MEOs as well as LEOs. How soon will they come back depends on
how far they are and what is the speed, etc. So it orbits around earth once in about every 2 to
12 hours depending on its parameters. So more satellites are needed and some handoffs are
required as the satellites orbit now. What is this handoff we are talking about? You are
communicating through the satellite to somebody else but the satellite has moved away. So
now the communication link will break. In order that the communication link does not break
what will have to happen is that another satellite will have to come and take its place and
there has to be a handoff of communication from this satellite to this satellite before this
satellite disappears altogether so that this communication between the two end points can
continue. So these handoffs are required as satellite’s orbit and transmit data rate at 9.6 Kbps
to 38.4 Kbps. Transmission delay is less than 0.1 seconds. So this is considerably less than
the 240 milliseconds we had earlier.

(Refer Slide Time: 31:10)

Taking it still further down, we have the low earth orbiting satellites, launched like a large
flock of birds. You require a lot of satellites because each of these satellites is visible to one
particular point in earth only for a short duration of time. So in order to give continuous
coverage, you need a large number of satellites. This cell size or the footprints will also
become smaller. So in order to cover the entire earth, you need a very large number of
satellites orbiting at constant altitude of 400 to 1000 miles. They must travel very fast to
avoid gravity forces because in that case they will fall down to lower orbit, which allows for
transmission of the 2.4 to 9.5 Kbps. It travels at 17000 miles per hour, circles earth
approximately every 90 minutes. So this goes around very fast. It is used for mobile voice
low- and high-speed data; internet access via mobile phones and PDAs and GPS, etc. So the
low earth orbiting satellites are what are used for the so-called SAT phones. It is dual linked
due to some reason. Since this is low earth orbiting, the power that you require to reach the
satellite is much less. So now you have a handheld device, which may be a little bigger than
your standard cell phone, but not very much bigger. You can still carry it in your hand and
with that now you can phone from anywhere from the earth because you are just
communicating through the satellite.

(Refer Slide Time: 32:36)

Low earth orbiting satellites are individually cheaper. They also give lower space delay,
which is much less because we are that much nearer. However a large number of satellites
need to be deployed. As the satellites keep moving, ground stations which are communicating
will have to switch from one satellite to another and quick handoffs will be required.

(Refer Slide Time: 32:51)

One of the examples was the IRIDIUM set of satellites. There were 66 satellites, which
offered mobile telephony, paging and data communication. Unfortunately by 1990 they went
out of service. What happened was that, SAT phone services became very costly. By that
time, other kinds of technology like fiber technology and the terrestrial wireless technology
developed to such an extent that a major part of the market was captured by those, which
were giving the same service to the end user at a much cheaper rate. Furthermore we require
66 of them, all of them moving very fast and you have to have this complicated handoff from
satellite to satellite to make it low enough so that you can communicate from a handled
device etc. This could not be supported in the market since there are very few people going to
such remote places where there cannot be any other communication infrastructure. Majority
of the users would have some base station for some terrestrial wireless nearby and that is why
this price became very uncompetitive. There were some efforts to revive this company, but
finally it did not work out and they became bankrupt. At present there is the ambitious
ongoing project called TELEDESIC in LEO, which includes 288 LEO satellites to provide
low-cost high-speed internet access, networking and teleconferencing across the globe.

(Refer Slide Time: 35:07)

Here, as I mentioned, the space delay is lower and becomes comparable to terrestrial lines in
the L band. It is possible to communicate with it using battery powered handheld devices.
However cell sizes are still too large compared to terrestrial cells.
(Refer Slide Time: 35:47)

IRIDIUM system had 66, so each satellite was to have 48 spot beams, giving a total of 1600
cells, each with 174 channels. But 174 channels or total of 183,000 channels are not so many
globally because terrestrial service providers give millions of channels over the area. So
finally, the cost turned out to be too high and the project went bankrupt.

(Refer Slide Time: 37:13)

Now we talk about VSATs, which are small terminals with about 1-meter antenna and 1 watt
transmitter power. Often the downlink capacity is more than the uplink speed for point-to-
point communication usually goes through a hub. If two persons want to communicate the
communication will go from here to the satellite and from the satellite to a central hub, which
will have a very big antenna and it can handle all the MAC part of it. We are not going into
the details of this but anyway this will go to the hub and from the hub again it will go to the
satellite and then go the point B. Let A be the sender and B be the receiver. So it goes in two
hubs; so the space delay is doubled because we are going in two hubs, sometimes with the
help of the satellite MODEM kind of thing, you can go in one hub. So those channels are
more costly. Otherwise, with the usual two hubs, there is considerable delay of the order of a
half a second for this communication between A and B, but then there is no setup time. So
various combinations of TDM, TDMA, FDM, and FDMA are used for handling the MAC
and all this is controlled by that central hub.

(Refer Slide Time: 37:52)

As I mentioned, one important application was the GPS satellite constellation. This is a
global positioning system and this is operated by US Air Force. Another GPS system is on
the drawing board; it will be deployed by some other consortium. So there are 28 satellites in
this and it has six orbital planes at a height of about 20,200 kilometres. Since this is not the
geostationary orbit, they keep on moving and a minimum of five satellites are visible at all
times. So with this what you can do is that you can locate any position on the ground, if you
have a GPS terminal somewhere. A very common application of GPS is to put that antenna
on a car so the car can be tracked anywhere on the earth.
(Refer Slide Time: 39:19)

For GPS we do a Trilateration. Suppose there are a number of satellites which are visible to
any of the ship, plane or a car as shown in the picture. Suppose we are measuring the car’s
distance from a number of satellites. Since for the satellites we know relevant parameters,
from three different readings and three different distance measurements, if measured quite
precisely and accurately, you can get the accurate position of the car. There are so many
different applications, which are possible. For example, there is a central database, if you give
information about where you want to go, then it can tell you the path to take or the alternative
paths and which of them is crowded. In many places in Europe, you can get very precise
information about where you are, which road you should take to reach the final destination.

(Refer Slide Time: 40:59)

The advantages of AVL (automatic vehicle location) are fast despatch; customer service;
safety and security; digital messaging; dynamic route optimization and driver compliance.
Slide 32 shows a mobile GPS unit located on a truck. With GPS satellites, you can give faster
dispatch of goods. For example, some company has to deliver a lot of goods to a lot of
warehouses. So it may ship something and now the customer wants to know where it is. You
can immediately know where it is because you know into which truck you have put these
goods and where it is exactly. So, for courier service and similar services, you can give
enhanced and sophisticated level of service through GPS.

(Refer Slide Time: 41:24)

In conclusion, we can say that satellite communication will continue to serve where
broadcasting is essential or where terrain is hostile or very sparsely populated. It also has a
niche where rapid deployment is extremely critical. Let us consider battlefield kind of
situation where you want to rapidly set up some communication network. If you have a
satellite up in the sky already, you can take that transponder. May be there is a disaster where
you want to quickly set up disaster recovery and relief operations. You can use this satellite
because satellite communication will still go on whatever may happen. Even if there is a
flood and everything is flooded, the satellite communication will still go on moving with your
mobile units. In other combination with terrestrial radio links fiber is likely to hold the
advantage
(Refer Slide Time: 42:14)

Now we come to an interesting part of the satellite communication. We see how this MAC
works for satellite communication. This is another class of media access control technique,
which we mentioned earlier. This is the first example that we see. It is Random Access
Protocol.

(Refer Slide Time: 42:54)

We mentioned Random Access Protocols earlier just as we talked about token based
protocols, etc. We also talked about random access protocols. This is the first example of a
Random Access Protocol. Random Access Protocol is actually very simple. When a node has
a packet to transmit at full channel data rate, so just do not bother whether somebody else is
transmitting or not transmitting or wants to transmit or it’s going to transmit. Right away, you
just do not bother about anything. You just try to put in whatever you want to transmit on to
the channel. Now what will happen? If you are lucky since you are doing the random thing
without any knowledge or consideration about what other people are doing, your
transmission may get go through. So there is no a priori coordination among the nodes but
your transmission may still go through. What will happen is that two or more stations may try
to communicate at the same time or very nearly the same time. What may happen is that the
packet or the frame sent by one station A and the frame sent by the station B may collide and
what will happen is that both the frames will be lost and will become garbled. Since you are
talking about a medium, which is shared by everybody, these two transmitting stations A and
B would also be able to listen to this and find out that both of their messages are garbled. If
they find that both are garbled, they will retransmit as a sort of backup for a random amount
of time.

That is very important once again the second random is also very important this backing a
random amount of time because if the protocol says no both of them backup for a fixed
amount of time and after the lapse of that fixed amount of time both of them will start
communicating again. It is important that the back off for a random amount of time and
hopefully the random number generated by one station and the random number generated by
the other station happen to be two different numbers. So now they are going to stay
communicating at two different points of time but they will not collide. They may collide
with other stations. That is a different issue. To ensure the random access protocol, if two or
more transmitting nodes start transmitting at the same time we have a collision. Random
access MAC protocol specifies how to detect collisions and how to recover from collisions;
for example, via delayed retransmissions. These are the two parts of any random access
protocol.

(Refer Slide Time: 46:01)

This satellite protocol, which is called ALOHA or slotted ALOHA, is the simplest kind of
random access protocol. We will see other examples like CSMA CSMACD CSMACA etc
later on. So these are the examples of Random access MAC protocols ALOHA which was
the original grandfather of all these protocols first came, then slotted ALOHA and then all
these CSMA CSMACD and CSMACA etc were used
(Refer Slide Time: 46:40)

The ALOHA protocol was originally developed for packet radio in 1970s. This was
applicable to any system with uncoordinated users competing for a shared channel. There was
no carrier sensing. Carrier sensing means you do not try to find out whether or not somebody
is already communicating. That may sound a little strange at first; because if you could find
out that somebody is already communicating, then by trying to transmit at that point of time
you are not only sure to fail because your own message will get garbled and somebody else’s
frame will also get garbled, but then you will also have a more crowded media. For satellites
there is a problem.

You cannot always do this carrier sensing, due to space delay. Since we have a 240
milliseconds space delay, whatever you are listening to now was actually transmitted 240
milliseconds earlier. So if you find that the channel is quiet now, there is no guarantee that
after 240 milliseconds also it will also remain quiet. That is why it is difficult to do carrier
sensing in the case of a satellite. That is why this latency, space delay, has a very significant
impact on how we handle the MAC. Here we do not do it; 240 milliseconds is a lot of time in
which a number of frames may be sent. So we do not do any carrier sensing. We simply send
whenever we have to send something and then later on listen to the medium to find out
whether or not there has been a collision. This is the pure ALOHA protocol.
(Refer Slide Time: 48:46)

Users transmit whenever they have data. If there is collision within a time frame, dictated by
the space delay, it backs off for a random amount of time and sends it again. If the first bit of
a new frame collides with the last bit of a frame that is just finishing, both are taken as
garbled. So there is no such thing that the frame has gone through 95%. Either the frame has
gone entirely, or even if one bit has collided, both the frames are taken as garbled and lost.

(Refer Slide Time: 49:11)

Collision detecting by listening to broadcast channel or by absence of acknowledgement: if

collision occurs, each user waits random length of time as already mentioned. Various
collision resolution algorithms are available. Station does not transmit new frame until old
frame has been successfully transmitted. The station is stuck with the frame until it can
successfully send it.
(Refer Slide Time: 50:01)

The slotted ALOHA has better performance. All frames are of the same size. Time is divided
into equal size slots and each slot is to transmit one frame with some distance. If the
transmitters and the receivers are synchronised they try to send frames only at the beginning
of a slot. So nodes start to transmit frames only at the beginning of slots. Nodes are
synchronized; if two or more nodes can still try to transmit in a slot, all nodes detect collision.
There may be collisions even after making the slots. For example, suppose within the span of
the fifth slot three different stations are ready to transmit something. What will happen is that
at the beginning of the sixth slot, all three will be communicating and all three will collide.

(Refer Slide Time: 50:51)

When node obtains fresh frame it transmits the next slot. As soon as it gets a frame to send, it
sends it in the next slot. No collision node can send new frame in the next slot. If there is a
collision, the node retransmits the frame in each subsequent slot with probability p until it
attains success.

(Refer Slide Time: 51:17)

This is a picture of a slotted ALOHA. There is a collision over three nodes – nodes 1, 2, and
3. All three of them wanted to transmit. They start transmitting and then collision occurred.
Then all of them backed off. It may so happen that the second slot went empty. Then in the
third slot, what happened was that number 3 had backed off randomly for a long period of
time but node 1 and 2 may have decided on the same number over here. So 1 and 2 collide
again. What might happen is given as an example: may be in the fourth slot, 2 have tried
again and it has been successful. So there is a success over there. Then next slot is empty.
Node 1 and node 3 are trying and then they collide again. Then may be in the next slot, 3
succeed and so after 9 slots, the three have been able to communicate these three frames. So
you see because of this, collision affected efficiency.

(Refer Slide Time: 52:32)

In slotted ALOHA pros a single active node can continuously transmit at full rate of channel.
If nobody else is transmitting one active node can go on transmitting. This is highly
decentralized. Only slots in nodes need to be in sync; so that may be taken care of from the
hub or through the satellite etc.; it is a very simple kind of protocol. (Refer slide 43)The cons
are that there are idle slots and hence low efficiency.

(Refer Slide Time: 53:08)

Now what is the efficiency? Efficiency is the long run fraction of successful slots. There are
many nodes each with many frames to send. Suppose N nodes with many frames to send each
transmitting slots with the probability P. So probability that first node has success in a slot is
equal to p (1 − p*(N − 1)). So it is the probability P that that particular node sends and it is
the p1 − p that all the other N − 1 do not send. So only then you will have a success; and
probability that any node has a success is Np(1−p) N−1.

(Refer Slide Time: 53:47)

So for maximum efficiency with N nodes, we have to find a p* that maximizes Np(1 − p)*(N
− 1).Hence this value. For many nodes, we take the limit of Np*(1 − p*) N − 1 as N goes to
infinity and this gives a value of 1/e, which is 0.37. So this is the efficiency – 37% is the
maximum theoretical efficiency that you can achieve in slotted ALOHA. So at best, channels
have useful transmissions; 37% of time, which may not look like a very high figure, but then
you are absolutely uncoordinated; but if you have low load then that may be alright.

(Refer Slide Time: 54:38)

For pure or unslotted ALOHA, is simpler because there is no need for synchronization. When
a frame first arrives, it is transmitted immediately. So collision probability actually increases.
A frame sent at t0 collides with other frames sent in t0 − 1 assuming that each frame takes one
unit of time to send up to t0 plus 1; so there is a big interval from t0 − 1 to t0 plus 1, where
things may collide.

(Refer Slide Time: 55:14)

So this will overlap. So this frame will overlap and now there is no question of any slots, so
anybody can start transmitting at any point of time.

(Refer Slide Time: 55:39)

What will the efficiency be like? P of success by a given node is equal to P that the node
transmits multiplied by the probability no other node transmits in [t0 − 1, t0] into P, a
probability that no other node transmits in t0, t0 plus 1. So this is p(1 − p) N − 1 into (1 − p)N − 1
so p (1 − p)2(N − 1) . So choosing optimum p and then letting N tend to infinity gives as an
efficiency of 1/(2e) or about 18%. So this is a rather low efficiency for transmission.

(Refer Slide Time: 57:15)

(Refer Slide Time: 57:19)

When a computer network developed there were a number of competing LAN technologies
but today Ethernet has a come to dominate a LAN almost totally excepting for the newly
emerging wireless part which we will discuss later. So this Ethernet is not only a LAN.
Nowadays people are talking about Ethernet in the MAN that is metropolitan network area
also. So Ethernet, as you understand is very important. So we will talk about Ethernet. It’s a
dominant data link layer technology the multiple access scheme of Ethernet is CSMACD
now

(Refer Slide Time: 57:29)

(Refer Slide Time: 57:44)

So this minimum time is which may vary from system to system, that means a network to
network. So this is how it is calculated so far. Let us see how the minimum time would
comes. Suppose this is the shared bus we put two nodes at the two extreme ends. Suppose
now packet starts at over here at time t equal to zero. Packet is almost reached b at time t-ε ,
at which point of time, it does not transmit. Suppose the propagating time of the packet for
moving from one end to the other is τ, at τ- ε it has almost reached b. At that point of time b
starts transmitting, because we have been finding that the bus has been very quiet. There is no
signal at end so b can start .So b starts and if there is a collision and this collision is at one
this end of the of the medium .So the collision bus starts transmitting back the jam signal.
.So it again take s almost tau to reach .Soa so the total time is τ noise. So collision detection
can take as long as 2 τ. (Slide 52)