 If we remember back to our topic on data transmission, we said there are analog and digital data, and we want to transmit that data as either analog or digital signals. And one of the examples we used was if the digital data we want to send is zeros and ones, and if we want to use a digital signal, then one approach we used was if we have a bit one, send a positive voltage, a high signal. If we have a bit zero, send a negative voltage or a low signal. That's a signal encoding technique. It's how do we encode data as a signal. Another example is if we had an analog signal, if we had a bit one, we use one shape of the sine wave. If we have a bit zero, we use a different shape, maybe go down and up. That was another signal encoding technique. And that's what this topic is about. But there are a few more signal encoding techniques that we need to talk about, and then the trade-offs between them. To get started, we'll just give a quick introduction and then start on digital data as digital signals, and then later we'll look at how do we send digital data as analog signals, analog data as digital signals, and analog data as analog signals. There's four combinations there. We'll just start on the first one. The signals that we generate, we'd like to have some properties such as they don't use much bandwidth. The more bandwidth, the higher the cost. That they don't allow many errors. They minimise the chance of errors. So the shape of the signal should be such that errors are reduced. We can have either digital or analog signals. The terminology we use, when we have analog or digital data, we say we encode that data into a digital signal. So we talk about an encoder. When we have an analog signal to send, the data is transmitted. The analog signal we say is a carrier signal. It carries the data. It will have a particular frequency, and we say that we put the data on the carrier signal by modulation. That's maybe best illustrated from this diagram. The top diagram shows an example when we have a digital signal. Highs and lows. Maybe plus one volt for a high, minus one volt for a low. A square wave, for example. No matter the type of data at the transmitter, we encode that data onto the square wave. So the process is called encoding. And we send the digital signal. The receiver receives it and takes that digital signal decodes and gets the data back. So a device that does both the encoding and decoding, we call a coder and decoder, or short a codec. So you may have heard of the word a codec. A codec is converting our data into a digital form. And the decoder part does the opposite. If we have an analog signal to transmit, say a sine wave or a combination of sine waves, then the data, we take some input carrier signal, maybe a sine wave of a particular frequency, and we change the shape of that carrier signal depending upon the data. This process is called modulation. So we have a modulator here. We send a signal, the demodulator from the received signal gets, extract the data out. And a device that does both is called a what? A device that does both modulation and demodulation is called a, what's the name? A device that does modulation and demodulation is called what? A modem. A modem deals with analog signals. A codec deals with digital signals. Digital data. Let's go direct to some examples of digital data sent using a digital signal. We will not get much time to explain why there are different approaches. Let's just go to show you some of the different approaches. We'll talk about the reasons later. Tell me a scheme. How do I send digital data as a digital signal? What have we used in the past? Bit 1, high voltage. Bit 0, low voltage. A simple scheme. So that's the first scheme we can use. Or we could switch it. Bit 1, low voltage. Bit 0, high voltage. Effectively the same. And where we, to transmit bit 1, we transmit a voltage for a fixed duration and to transmit another bit, we transmit a voltage for the same duration. This duration we've mentioned as the signal element duration. But there are other schemes. So we send a series of pulses. High or low. We've seen some of that terminology. Let's go to some examples. And the examples of the schemes, the six schemes that we'll look at and there are others as shown here on the left. And this is an example where the data that we want to send is at the top and the digital signal that we generate for that data for the given scheme is shown. The first one is the one that we know. Look closely, we see if we send a bit 0, send high voltage, send a bit 1, low voltage. 0, high, high, low, low and so on. So that signal is generated from the data from that simple scheme of one bit is one voltage level, the other bit is the other voltage level. In fact, it's opposite to some of the examples we've used. We used 1 for high, 0 for low. But it's okay. It's just switched. The voltage levels in this case, the high must be a positive voltage and the low is a negative voltage. Neither of them are 0 volts and the name NRZ means non-return to 0. The signal never returns to 0 volts. It's always a positive or a negative voltage, whether it's plus 1 or plus 5, doesn't matter from our purpose, but it's positive and negative. So 0 is in the middle, 0 volts. Non-return is 0 level, the L is at level is the name of this scheme. We've seen that before, but there are others and 1, which is similar, is called non-return to 0 invert on 1s. Whenever we want a bit 1 to be transmitted, we change the level. Let's say we start at low. For bit 0, we maintain the level at low. The next bit is a 1, so we switch to high. We make a transition. We invert the levels. 0, we maintain the level. Maintain bit 1. Transition. Invert. Next bit 1. Invert. 3 zeros. 1, 1. A different scheme. Invert on 1s. In the last 5 minutes, we will not explain the advantages and disadvantages yet. We'll do that next lecture. Let's just look at the last one, one or two other examples. Allow you to do some quiz questions. Bipolar AMI has three levels. The rules are bit 0, 0 volts in the middle. Bit 1 is either positive or negative, non-zero, and we alternate the level for each bit 1. Bipolar alternate mark inversion, that is the first bit 1, it's high. The next bit 1 is low. The third bit 1 is back to high, low, and so on. Bit 0 is always 0 volts. Bit 1 is either positive or negative, and it alternates for every bit 1. Some quick examples to finish that. I'll draw a signal, you tell me the data. And we'll start with an easy one. NRZ level. Here's NRZ level. What's the data received if you receive that signal? 0. High is 0 in this case. Low is 1. And we must take note of the signal element duration, so there's two ones here. That one's the easiest one. Low is 1. 3 zeros and a 1. Let's do one with NRZ invert. That's the 0 volts, so let's to be clear, this is 0 voltage here. Non-return to 0, neither of them will return to 0 volts. What's the data in this case? And we need an assumption. I'll let you try. What's the first bit? 0. Now we'll assume in this case that I think the previous bit was 0, so we maintain the level at the start. Maintaining the level mean we have a bit 0. We don't change. Next one, we maintain 0. Alternate or in invert, 1. Same. Invert. 1. 1. 1. 0. Easy. Bipolar AMI to finish. So far, what's our data? Bipolar AMI, 0 volts, means a bit 0. Non-zero will be 1. I didn't hear everyone. Let's finish that. Add a few more bits. This bit. 0. Next bit. What's this bit? We have an error. Why? This one was low voltage. 0 volts, 0. The next one is non-zero, so it's not a 0, but it should be high voltage. We know the transmitter always alternates when it sends a 1. But it didn't when we receive. What does it tell the receiver? Something's gone wrong. Maybe this previous one, it wasn't really 0 volts. There was an error in the system such that this should have been positive. Or maybe this one should be 0. But the receiver knows there's an error. And that's the feature of this encoding scheme that the others don't have. It has this in-built error detection. So that's one trade-off between different signal encoding schemes. Some schemes, if something goes wrong, the receiver can detect that. And that's a useful feature because then it can take some other action to try and correct that error. So bipolar AMI has this in-built error detection if we have levels at the wrong, like two negative ones in a row there. Have a look at the others. Pseudo-ternary is the opposite of bipolar AMI. Bit 1, it's 0 volts. Bit 0, it alternates. That's easy. Manchester. And on the previous slide that I described, bit 0, in the middle of the bit interval, we transition from high to low. Bit 1, in the middle of the interval, we make the transition from low to high. Bit 0, high, low. Bit 1, low, high. Bit 0, high, low. So we see the signal when we use Manchester encoding. And we'll see that has other advantages compared to the previous systems. So have a look at the descriptions for them on the previous slide. It describes the rules for those six. You should be able to give in the signal, find the data, or give in the data, draw the signal. And I'll give you a few quiz questions just on that conversion for the next quiz. And then next week we'll discuss why are some better than others? What are the trade-offs between them? We'll do that next lecture. We'll stop there. Our plan in this topic is to look at the four combinations of how we can send digital data and because we have digital data and analog data and we can send it as either digital signals or analog signals. We get these four different cases. We've given some examples in the last lecture about how to send digital data as digital signals. So we went through five or six schemes or actually we went through in the lecture three schemes and you've studied I think some of them in the quiz. Questions about the schemes in digital data as digital signals. So today I'll just summarize that. I will not go through more examples of them if you do not understand them. Then look at those quiz questions. We'll summarize and look at the trade-offs. Then we'll quickly go through digital data and analog signals, analog data and analog signals and finish on the third one. The one that takes the longest. So we may jump through a bit. With digital signals we send pulses. You think we send say a high voltage for some period of time and then a low voltage, a pulse which we've been calling a signal element. So we have signal elements which have some defined duration. We can talk about and this comes from one of our earlier topics. We talk about data rate the number of bits per second we send but we can also talk about signalling rate the number of signal elements per second which are transmitted. Sometimes they are the same sometimes they are different. We'll see with our different encoding schemes. An encoding scheme is one way to try and improve the reception of our data improve the chance that the receiver will get the data successfully with no errors. There are other ways to try and avoid errors like increasing the signal strength relative to the noise or reducing the noise and so on. So there are a number of things that impact upon errors. One of them is the encoding scheme. So we're looking at some different encoding schemes and we got to an example where one can actually detect errors. This defines those common encoding schemes. There are others but they are the ones that we see in this course from non return to zero level the simplest low is bit one high is bit zero through to some other ones bipolar AMI and pseudo ternary are almost the same they're just like the inverse of each other and Manchester and differential Manchester are similar as well and the last two which we will not spend much time on are really extensions of bipolar AMI. Have a look at them and make sure that given some data you can draw a signal or given a signal you can interpret what the data is and you can distinguish between the different schemes. These six schemes specifically. So I will not go through more examples because you should see that in the quiz. Very easy once you try one or two examples. What's the difference? Why do we have multiple different schemes to send data using digital signals? Well let's look at some of the trade-offs the advantages and disadvantages of some of the schemes. We we saw one advantage last lecture we saw with bipolar AMI if there's an error in some cases the receiver can detect that error. That is with bipolar AMI for every bit one we should transmit positive negative. Positive negative. They must alternate. But if the receiver receives two negative signals in a row that indicates an error. Something's gone wrong. So this is what we call error detection. The receiver through the encoding scheme can detect if something's gone wrong. So that's a nice feature which some schemes have others don't. Inbuilt error detection. What else? What other features do they have? Well if you look at for long sequences of bits, zeros and ones look at the shapes of the signals you can do analysis and see what bandwidth they occupy. The spectrum of the signals when we use different encoding schemes differs. And the key point is that maybe captured by this picture some occupy more bandwidth than others. Before we explain the picture normally to send the same amount of data we'd like to use as small bandwidth as possible. Bandwidth is a resource that many people want to use so it has some cost involved. If I want to send my data then a scheme that uses a small amount of bandwidth is a good scheme. Which of these schemes use the smallest amount of bandwidth? Well this plot gives some view where on the horizontal axis we can think that's the measure of bandwidth. It's normalized against data rate so we can compare each of them. What we can interpret here is that Manchester and differential Manchester, the solid black line have a larger bandwidth occupied to send the same amount of data as a bipolar AMI and pseudo ternary, this dashed line. You can see it's narrower compared to Manchester slightly narrower saying that with AMI and pseudo ternary we use less bandwidth than Manchester encoding and that's a good thing. Non-return to zero although it starts here AMI is better than non-return to zero as well. It's hard to see but it uses more bandwidth with NRZ and more. So there's differences in the amount of bandwidth they consume so we need to select the one that achieves our goal of how much bandwidth we want to use. So that's another trade-off here. There's also some things about the heights of these signals so bipolar AMI has a smaller bandwidth but requires higher signal strengths and that sort of makes sense if you think of bipolar AMI there are three levels needed. With bipolar AMI there are three levels needed high, zero and low. The others just need two so under the same conditions we normally need to transmit at a higher signal strength to have the same separation between the levels. Think of the ones with two levels to get the same separation in bipolar AMI we'd need a higher positive and lower negative voltage so the middle two require more power to transmit and that's one disadvantage of them and maybe captured by this diagram they're higher in this plot so the different trade-offs in terms of the bandwidth consumed. Another feature is some of the encoding schemes. Here's a signal using NRZ level the normal non-return to zero what's the data? Quickly tell me the data with NRZ zero remember high is bit zero low is bit one zero one how many ones? Why seven? Well to be precise there's a exact duration here so visually sometimes it may be hard to see is it six? Seven? Where does the seventh bit end? Well the receiver and the transmitter must know what is the signal element duration so let's say the duration had a number let's say this duration was one millisecond what the receiver needs to do is every one millisecond indicates the next bit one millisecond and so on and that would determine how many bits that's fine the problem is in practice the receiver must have a clock that is very accurate or synchronized to the transmitter's clock because to count on a very small time basis milliseconds, microseconds the clocks in the hardware may not be accurate much that even though the receiver thinks this is one millisecond and checks okay that's the first bit the second bit and so on it may be slightly wrong it may think it's one millisecond but it's only 0.9 seconds compared to the transmitter's clock what can happen especially when we have a long sequence where the signal is the same level like in this case the receiver gets to a point where it thinks it's maybe the eighth bit but it was only the seventh bit transmitted because the clocks are slightly different then they can be out of sync and that mainly happens when we have a scheme that allows a long sequence or a signal at one level for a long period of time I think it was seven, I think you were correct when there were seven bits there two, three, four, five six, seven but in practice that the clock may be just slightly wrong such that it cannot accurately measure when the next bit starts and that's a problem with non-return to zero level, if you have a long sequence of ones the signal stays the same similarly if you have a long sequence of zeros the signal stays the same and the receiver can be out of sync consider for example bipolar AMI if we had, what do we have a long sequence of ones with bipolar AMI what would it be well the zero is at zero volts a long sequence of ones with bipolar AMI we alternate so in this case with a long the same data which was zero one, one when we use the bipolar AMI encoding scheme we don't have this problem of a signal at the same level and therefore there's not much of a chance that the receiver will be out of sync because the receiver knows every time it sees a change in the signal level ah signal levels change it must be a new bit it must be a new bit so it can synchronize its clock so some schemes have that feature some don't inbuilt synchronization it's good to see the signal changing it helps them to synchronize what if we have bipolar AMI but with a long sequence of zeros what do we get that is the inverse sequence one followed by seven zeros with bipolar AMI the one may be high a long sequence of zeros we get zero volts again that's a problem bipolar AMI is okay when there's a long sequence of ones but when there's a long sequence of zeros and zeros here again we have this issue of the signals at the same level all the time it's hard for the receiver to determine when is the actual next bit start so it works for ones but not zeros so in fact there are extensions of bipolar AMI such that when there is a long sequence of zeros replace that sequence with a special sequence instead of having seven zero volt signals there use a special sequence and we will not go through the details that's not important for this course but the schemes B8ZS and HDB3 do this they take one of the previous schemes and when they have just one special case when there's a long sequence of a particular bit zeros in this case instead of transmitting that expected sequence it will transmit a special sequence with bipolar A0 substitution B8ZS whenever there's eight zeros in a row which would produce a signal which is always at zero volts instead of transmitting that signal at always zero volts it transmits a special sequence of say three zeros zero, low, high and note in the special sequence it has a violation of the rule we'd expect with bipolar AMI positive, negative positive, negative but this special sequence has this violation of positive, negative negative, positive so the transmitter uses this to indicate to the receiver here is a violation of our rule what it really means is I transmitted to you eight zeros even though we see positive and negative signals here HDB3 is slightly different it considers sequences of three zeros the details of how they implemented are not important well the point is that there are extensions of existing schemes to try and solve this synchronization problem so we don't get this flat signal that we always have variations every few bits that's desirable let's try and finish so you don't need to remember those two extensions just remember the point that synchronization of the receiver is an issue and it arises when we have a signal that is maintained at one level for a long period of time what else synchronization, error detection some schemes can detect errors others cannot some schemes work better when there's noise like Manchester and differential Manchester bipolar AMI is not so good with noise because it has three levels and some are more complex to implement the transmitter and receiver and hence we can think they cost more to use especially the Manchester schemes with the first four schemes there's a transition in the worst case every signal element duration every bit there's a transition but with Manchester encoding our transmitter must change the level in the worst case two times per bit so it must change in the middle and it may have to change at the start of the bit duration so here in Manchester encoding there must change the voltage faster than the others and that's more complex to implement so results in a more complex transmitter and receiver so there's a trade-off in terms of complexity and that's all I think we'll say about these different encoding schemes there's no one best encoding scheme they're used in different applications what applications we jump through that shows the B8ZS compared to bipolar AMI used in different cabling, RS232 serial cables, USB uses non-return to zero Manchester encoding is used in wired lands when you plug a LAN cable in you're sending digital data using a digital signal using Manchester encoding and some long distance links used by bipolar AMI and variants of that so just some examples questions before we move on to the next two approaches who hasn't done the quiz everyone's tried the quiz once so you should be experts at converting the data to signals and vice versa there were four or five quiz questions let's then look at we've got digital data zeros and ones but now we want to transmit analog signals not pulses but analog signals think of sine waves continuously varying signals remember back to our first lectures a sine wave or the sine equation there are three parameters what are they what are the three things that we can vary to change the shape of a sine wave amplitude we can change the height of the sine wave the amplitude what else the frequency it can either be changing quickly or it can change slowly so the frequency of the sine wave and the phase the phase is the offset of that sine wave and we've seen some cases where a phase of zero follows that normal structure and a different phase means it starts by going down so we can change those three parameters and that's what we'll do here when we have a bit zero we'll transmit a sine wave or a signal with a particular say amplitude if we want to transmit bit one we'll transmit with a different amplitude we'll change that parameter of amplitude or we can change frequency or phase we can send digital data over systems that support analog signals like your telephone network carries analog signals through the telephone lines some radio or microwave systems use analog signals the three techniques that we have available is either change the amplitude of the analog signal change the phase or change the frequency and the names of these schemes amplitude shift keying phase shift keying frequency shift keying and we'll see some more complex combinations very easy to understand here's three simple examples here's the data the digital data we want to send a sequence of zeros and ones this is just the NRZL signal of that, we can ignore that for this example the first analog signal we use amplitude shift keying and the rules in this example are to transmit bit zero set the amplitude to be zero when you have a sine wave with the amplitude of zero it's just a flat line when you want to transmit a bit one set the amplitude of the sine wave to be non-zero one or two or whatever the amplitude you want to go up to so we get this analog signal transmitted bit zero zero the amplitude is zero amplitude is non-zero so we get a sine wave that means it's a bit one being sent and so on what amplitude zero one or zero and five or one and two well that needs to be defined but the important point they need to be different amplitudes we shift the amplitude depending upon the bit the other parameters of that sine wave in this case the amplitude changes but the frequency and phase stays the same so just in this example we've got a frequency such that let's say the bit period is one second then in this case there are two waves in one second or a frequency of two hertz and the phase is zero we get just going up let's look at what if we change the frequency of the signal so all right frequency shift keying the rule here is that when we have a bit zero we're going to use a low frequency and a bit one a higher frequency the amplitude stays the same the phase stays the same the frequency changes with bit zero we have this low frequency one hertz one hertz bit one double the frequency two hertz what the receiver does every time every signal element duration it measures the frequency of the signal it received if it's low it must be a zero received if it's high it must be a one received so by changing the frequency we can represent the two bits last we change the phase you can see bit zero we have the the inverted sine wave here we go down first a non zero phase with bit one we have the normal sine wave a zero phase there so we go up first frequency is the same in each signal element the amplitude is the same the phase is changing depending on the bit the other basics of sending digital data as analog signals questions not so hard this one any questions at the back got the calculator out the phone doing some calculations ok then let's go through one example just to make it clear that's an example of frequency shift keying I'll draw a signal and we'll look at what bits that may represent and just to make it easier for me I put some scale to it in this example I'll draw four signal elements and for me to draw sine waves not so easy but we'll try quite to scale we almost got there it'll make sense when we see the the frequencies just to make it clear because my sine wave didn't turn out so well the signal element durations are from here to here the frequency shift keying in this case that is what we do is to send a different sequence of bits we'll send a signal with a different frequency the amplitude is the same in all cases the phase is the same in all cases it's just the frequency changing and what I've tried to draw here is four different signal elements there are four different shapes and the way that I tried to draw it is that we four different shapes are from four different frequencies we have a frequency maybe this is the lowest frequency we have one wave completed in the one signal element duration so let's say a frequency of one could be one hertz one kilohertz but one unit this second signal element we repeat two times so a frequency of two and here I tried to draw three is that right three waves this is four so four different frequencies represented here where what we do each signal element will represent a sequence of bits so what we need to know to work out what the data is is what is the mapping from each signal element there are four to a sequence of bits and that needs to be defined in advance so I'll define the mapping to say this represents zero zero zero one one zero one one here I'm using two bits per signal element as opposed to one frequency one bit because we have four different levels each level can represent two bits we need to have four to represent two bits because there are four combinations if we just had three frequencies we could not do this we'd have to have one bit per level it'd be inefficient but if we have four frequencies we can transmit two bits per signal element so a trivial example in this case the data transmitted lowest frequency or frequency of one corresponds to zero zero for this duration zero one a frequency of two here we have a frequency of three so we get one zero here a frequency of four one one and if I wanted to transmit one zero next then I would transmit a signal with a frequency of three units so depending upon the pairs of bits I want to transmit I choose one of the four frequencies to transmit at so we break it into pairs of bits we have seen this concept before we have seen that we can send multiple bits per signal element by using more than two levels here we have four levels and we saw an equation that relates it to the data rate the Nyquist equation uses the parameter m the number of levels the more levels under the same conditions the higher the data rate we can send two bits per signal element as opposed to one this is what we would call quaternary FSK or maybe simply four FSK in this slide it was called B FSK binary B for binary two levels here we have four frequencies and you can extend that in practice frequency shift keying is used with multiple levels 16 FSK 32 256 and so on so we're not restricted to just two levels one more example not only are we not restricted to two levels so we can use four or more levels for each parameter we change amplitude, frequency or phase we can actually also change two parameters at a time instead of just one let's define a different scheme where we use a combination of amplitude shift keying and phase shift keying and we'll use two different amplitudes so an amplitude of one and an amplitude of two and we'll use two different phases phase of zero phase of zero and a phase of pi 180 degrees so here's our encoding scheme we've got four levels but instead of using say four different frequencies like in the previous example we're going to change two parameters at the same time the amplitude can range between one and two so two different amplitudes and the phase can also change so with those four levels we'll map them to bits and the mapping needs to be defined and known by the transmitter and receiver let's just choose in order given that scheme let's transmit some bits let's draw a signal and determine what the bits are here's our transmitted signal you receive this signal there are four signal elements here split into four parts what's the data received in this case and it is a side wave as much as it may not look like one here so this is the first signal element what's the data received write down the data here we have a small amplitude if we put values of one and up here is an amplitude of two so this signal element the amplitude is one and the phase is zero we've got the normal shape sine wave so the bits for this signal element is zero zero the second signal element we have the flipped sine wave the different phase and an amplitude that goes up to two so the amplitude of two and the different phase gives us one one in the third signal element we have the flipped sine wave so the non-zero phase but an amplitude of one the bits are one zero and the third one is in fact the same as the second so we get one one so this is just combining the change of two different parameters amplitude shift keying and that's common in practice to combine those two together not just with two levels but you can have more than two amplitudes and more than two phases and the name when you combine them together it's called quadrature amplitude modulation QAM that's not important to remember but you may see it in the specs of some equipment QAM it's really just a combination of those two and in practice you may see 16 QAM so there are 16 levels not 4 256 QAM and even larger for some cable systems 256 QAM means there's 256 levels combining the amplitude of the phase any questions on shift keying digital data using analog signals where are they used again there are different trade-offs between them maybe just finish with some examples amplitude shift keying is used in some optical fiber systems some low data rate applications ASK turns out to be very inefficient so it's only used when we need low data rates or we have a large capacity like optical fiber when efficiency is not a problem frequency shift keying is used in some wireless systems like UHF, VHF radio RFID is just a wireless system say you use to identify devices you have a tag on a on some object and you have a reader that will communicate wirelessly and read the tag so very low data rate phase shift keying and phase shift keying combined with amplitude shift keying QAM used in your mobile phone Wi-Fi cable modems ADSL and other systems digital radio and digital TV let's jump to the fourth scheme very easy again so we'll have a bit more time for the third one the first two with digital data now we want to look at well what if the data is analog for example voice someone speaking we just want to record that as he is and send it we'll jump to sending using analog signals and a common example that you know of is AM and FM radio with AM and FM radio there's analog data someone talking or some music and it's sent using analog signals from the radio station as a transmitting tower and you receive it in the car for example analog data using analog signals and we use similar concepts to the shift keying remember with shift keying we could change the analog signal we could change the amplitude frequency or phase the same in this case we can change the amplitude frequency or phase but first some of the details here you can read through but the key point is that often we'd like to transmit analog data at a frequency different from that data and the example is that a wireless transmission system for the audio system here when I speak I'm producing analog data think of my voice as analog data what's the typical frequencies of someone's voice a human's voice anyone remember from about 0 hertz up to about how many hertz more than 50 hertz set not 100 anyone else he's going to guess forever remember we had a plot in one of the earlier lectures it goes up to about 4 kilohertz 4000 hertz when people talk about 4 kilohertz the frequencies range from several hertz or 100 hertz up to about 4 kilohertz so that's my analog data but when I want to transmit it using this wireless transmitter to transmit an analog signal from here to the receiver in the desk I don't want to transmit at that range of frequencies I'd like to transmit at a different range of frequencies such that it can maybe play much better it will not be interfered with by the people near here we need different size antennas so what this device does and it was in the previous lecture but it's not written on the back in the previous lecture I had that this transmits at a frequency of about 900 900 kilohertz so my voice is ranging from about 4 kilohertz this transmits an analog signal about 900 kilohertz so essentially we move the frequency of our analog data to a different frequency instead of from 0 to 4,000 it's around 900,000 hertz and that's one of the roles of transmitting analog data as analog signals and there are three things we can change we can change the amplitude of the signal sent by this device depending upon the data we can change the frequency or phase and we get AM, FM and PM and the two that you know of are quite common are AM radio and FM radio amplitude modulation and frequency modulation this is an illustration of AM amplitude modulation what it shows is that in the middle is the data in this example the data is very simple it's a sine wave think of that as the voice if you measure voice you'll see that there's variations of the signal strength over time and different frequencies this has the same frequency for a simple example the top signal is what we call our carrier signal this is the signal we're going to transmit from the transmitter to the receiver or it's going to be based upon the carrier signal it's not going to be exactly that so for example the middle signal may be my voice and it would range from up to about 4 kHz the carrier signal would have a frequency of about 900 kHz and what we do is based upon the signal strength of the data the height we change the amplitude of the carrier and the result is the bottom signal and that's what's actually transmitted and you can hopefully see the shape it's the transmitted signal it's the same frequency as the carrier but the amplitude of the transmitted signal is changing according to the amplitude of the input data you can see the shape it's going to be high here when we have a a peak, a positive peak on the data when we have a negative trough here the amplitude is low on the carrier so this is the transmitted signal this is the data this is what we call the carrier signal we modulate the data onto the carrier to get the transmitted signal this is amplitude modulation we change the amplitude of the carrier phase modulation we change the phase of the carrier it's a bit harder to see but if you look close at the start the phase is non-zero it's the phase of pi it's going down first and over time the phase of this carrier signal which is zero in the carrier that changes and it changes depending upon the height of the input data that's phase modulation and the last one of course is frequency modulation used for FM radio looks similar to phase modulation but we're actually changing the frequency slightly of the carrier so maybe the carrier has a frequency of 900 kilohertz originally but depending upon input data we make small changes to the carrier frequency maybe plus or minus several hertz or kilohertz such that we get this transmitted signal you can see I think the pattern the frequency is lower slightly lower than the carrier when our input data is high when the input data is low the frequency of the transmitted signal is higher than the carrier so the carrier frequency changes depending upon the input data and that's AM, PM and FM and the two that we see commonly amplitude and frequency modulation AM radio, FM radio for example with AM radio with FM radio on this slide we have the voice or the music which is ranging from frequency up to about 20 kilohertz music and when it's transmitted over say a channel of 107.1 megahertz that means the carrier signal is centered around 107.1 megahertz but it varies a little bit so plus or minus depending upon the input data if the magnitude of the input music is high or low then that carrier signal has its frequency varied the receiver measures the received frequency to determine the original data and that's it, that's all we want to say about the last techniques questions before we go the fourth and final technique