 We've gone through three examples of guided media, wired media, twisted pair, coaxial cable, optical fiber, just as three common examples of what wired media, and then we went through general concepts of wireless transmission. That is the concepts of the use of an antenna, the gain of an antenna relative to an isotropic antenna. When we transmit a signal across some link, we know we lose power. How much power do we lose across some distance? Well, there are different models and one that we covered was called the free space path loss model, which gives us a relationship between if we transmit with power PT across some distance with some characteristics of the signal we can receive with power PR. So the difference is the amount of power lost and we've also in your quiz because you have some questions on that some relationship between power in milliwatts, watts, DBM, DBW, so the relationship on the linear scale on the decibel scale. The last part of this is some examples on wireless media. Give me some examples of wireless systems. Wireless systems. Give me some example technologies that you use or may have used on a regular basis. Wi-Fi, wireless LAN, okay? Uses frequencies of the order of 2 to 5 gigahertz. When we send our signals, they're around 2 or 2.4 gigahertz or 5 gigahertz. Different variations there. Wireless LAN or Wi-Fi. What else? Bluetooth. Uses frequencies of also around 2.4 gigahertz. So the Bluetooth built into your laptop or on an external adapter transmits the same frequencies of common as common wireless LAN. What else transmits around 2.4 gigahertz? I'm not sure. I don't know about that one. Near field communications in the latest Android some of the phones that have this communication between phones we put them near to each other within a few centimetres and they can transfer data wirelessly. So NFC, near field communications, I think they use a higher frequency. I have to check. Okay, other wireless technologies then. Come on. TV remote. Okay, TV remote. People use TV remote. What else? GSM, mobile phones, okay? Everyone has a mobile phone. GSM is the name of one of the standards used for mobile phones. The old, old, still used standard for mobile phones. And then it was upgraded to 3G and different variations. But mobile phone systems transmit signals in the order of 2 gigahertz between 1 and 2 gigahertz depending upon the system. Others? Radio, AM, FM radio. We've used them as an example. So all wireless transmission systems, others? Any more? GPS, okay? Most of you may use GPS in the car or on your phone. Others? Microwave, you cook your food or? That's a good one because in fact microwave that you cook your food before we move that is not of course for communications but uses the same frequency as Wi-Fi, wireless LAN. And in some cases if you, they've done experiments, if you put your laptop very close to the cooking microwave and cook something, you can interfere with wireless LAN. But other, generally microwave refers to a range of frequencies. So we talk about, and I don't want to go through these slides, but we can talk about microwave systems depending upon the frequencies they use. Terrestrial microwave, meaning on the ground. And others that you use, or some may use, not just GPS but satellite TV is common in a number of locations. So you receive TV via a satellite. IP Star is another, is a example of a satellite internet provider. So in places where you can't get land lined in the internet like ADSL or fiber optic into a building there, or even dial up, so in a rural area, then sometimes satellite internet is the only opportunity for internet access. So you have a dish on your home which can transmit up to the satellite and receive from the satellite and you get internet access via that satellite, which then goes to a base ground station. IP Star is the brand name of one of the popular ones in Asia provided in a number of countries in Asia. And it's run by a Thai company. Tycom is the company that runs it. They have several satellites, I think, but one main satellite that they have up, which provides coverage for many people in Thailand, in Indonesia, Australia, parts of India, I think for internet access via a satellite. These other, these slides, we've said the things that we need to say about these slides. So it's just a set of examples about wireless systems and I think you use wireless systems on a regular basis. So we're not going to go through any more details in these slides. I'm not going to ask questions about the details like what is a VSAT and so on. But if you want to look through, look through those remaining slides for a few examples of wireless systems. There's no more that we can add there that's useful for this course. I want to move on to the next topic, signal encoding techniques, which is going back to some things we've already discussed of we have data to send from A to B. We transmit some signal from A to B that carries that data. So there's some mapping from the data to the signal and how we do that mapping is called a signal encoding techniques. We encode our data onto a signal. One example that we've used in earlier lectures because we hadn't gone through these techniques was a simple one like transmit high to represent bit one and low if we want to transmit bit zero in terms of a part of a sine wave. A very simplistic technique where if we have an analog signal that we're transmitting through our communication system and we have digital data to send zeros and ones is the data. This is a technique which says whenever we have a one transmit a high portion of our signal, whenever we have a zero transmit a low portion of the signal. We've used that one before. We're going to look at some more some extensions of this and some other alternatives. So when we choose, when we design a communication system to transmit signals, it's important to generate a signal that is good. And there are different definitions of what is good, but some properties that we look for is that we make good use of the medium we have available. So if I have a twisted pair copper wiring, that's the medium, then that can carry a range of frequencies. So I want to generate a signal that uses what's available. So it's efficient in the use of the medium. When we have a wireless system, I want to transmit us, I need to transmit a signal from a laptop to the access point. What bandwidth or what frequency do I use for that signal? Well, that will have an impact on the data rate. So we need to design the signal to be appropriate for our system. Some things we consider and make sure that our signal doesn't use too much bandwidth. That is, we conserve the bandwidth. We don't use too much because there's only a certain amount of bandwidth in some systems. So we need to pay more to use more. Minimize errors, for example. Choose a signal such that when we transmit it, when the receiver receives that signal, even if there's noise, there's attenuation, there's some interference, even in the presence of those, the receiver should be able to understand the information. That is no errors. So we'll see some examples of different signals can reduce the impact of errors. So we need to design a signal to meet our requirements. Now we will distinguish between two types of signals, digital and analog. So when I transmit a signal from A to B, we would say it can be either analog or digital. And we know the difference. We've seen some plots before an analog signal and say some digital signal. So we think of our digital signal as we transmit some voltage pulses. So if we have an electrical signal, we maintain some voltage at some level, let's say minus five volts and plus five volts. So maintain some voltage for some time period to represent one piece of information. And then a different voltage for the same time period for a different piece of information. For example, bit one, bit zero. So we'll treat them differently, digital and analog signals. Some of the notation that we use here that you see in red, I'll come back and define that after we go through some examples, okay, because we may not even get to that today. We'll focus first on digital signals. In either cases, whether we're transmitting an analog order digital signal, the data can be analog or digital as well. So the information that I'm getting from A to B may be digital, like zeros and ones, text files, emails, any digital representation of our information. Or it may be analog, some audio, some video, some sensor data. So we can have digital signals carrying digital data or analog data. And we can have analog signals carrying digital data or analog data. The signaling coding techniques we go through are ways to map the data to the signal. Like this one, the signaling coding technique is transmitter high for bit one, low for bit zero. There are other alternatives. Mainly today we'll go through examples of them and discuss them more tomorrow. Just before we go through some examples, some more terminology. When we're transmitting a digital signal, that's this part here in the diagram between the two boxes, it doesn't matter if we've got a digital or analog data. So this is our data as the input. We convert that data into some digital signal, transmit the digital signal from A to B, and then at B we convert that signal back into the original data. That's our goal. These steps of converting data to signal, signal back to data are called encoding and decoding. So we have an encoder and a decoder when we're dealing with digital representation here. And together, what are they called? If we talk about an algorithm or a piece of software that does both encoding and decoding, what's the name? You've heard of it, I'm sure. A codec, some of you may have heard of codecs or seen codecs when you play files or media. A codec, just remove the EN and combine the two. Codec, a coder and decoder. That's what codec is. It stands for the combination of the two. A codec is dealing with taking some data and converting it into a digital form. Like an mp3 codec, it takes audio, some analog information, analog data and converts it into a digital form. If we're dealing with analog signals from A to B, doesn't matter about the type of data here, we do a process called modulation. We take some analog input, some analog signal called a carrier signal and vary the carrier signal according to the input data. And we get our analog signal as output. And at the other side, we demodulate and get our original data back. And combined, what are they? What's it called? What's the answer? The device that does both of those things? What's the device that does modulation and demodulation? A modem. Okay, you've all heard of and used modems. A modem is just a combination of a modem. It does the modulation and demodulation. So we're going to go through some examples of different techniques for converting either digital or analog data into a digital signal or an analog signal. So there's four combinations. All of them are used in real life in practice. They have different advantages and disadvantages. Using digital nowadays with electronic devices is usually cheaper to build the devices. So it makes sense to use digital signaling. Sometimes we have analog data. So if we have equipment that sends digital signals, we convert that analog data into a digital form. Some systems, we transmit analog signals. So if we want to transmit our digital data across a system that only carries analog signals, we need to do this conversion or also with analog data. Let's go through some examples starting with how do we convert digital data into a digital signal? Today I will not go through, I will not describe much on the slides. We'll just show you some extra examples and you'll do a few in the class. And then either at the end of today or tomorrow I'll come back to some of the slides and explain the things we've skipped over. So I want to go straight to the examples. This is a list of some common techniques for transmitting digital data as a digital signal. The first one we'll go through is the simplest. It's called non-return to zero level. We'll get a definition in a moment. Here's a digital signal. It's been generated and transmitted across a communication system. How is it generated? Non-return to zero level, NRZL level. The mapping, remember we have digital data, zeros and ones, and we create a digital signal like this one. The mapping is simple. When we have bit zero, we transmit a high pulse. When we have a bit one, we transmit a low pulse. And the pulse in this example is this duration shown by the dashed vertical lines. What's the data? Try and determine the data. How many bits first on the example? How many? Eight bits. What's the first bit? Two options, so you can guess. Let's do a hands up. There's option zero, option one for the first bit, hands up for zero for the first bit, hands up for one for the first bit, hands up for don't know for the first bit. The mapping in non-return to zero level is in fact the opposite of what we've used in some previous examples and maybe the opposite of what we think to be the logical approach is low represents bit one. High represents bit zero. So the first bit is a one. So one, zero, one, one, zero, zero, one, zero. Eight bits in this example. So this is one of the simplest schemes that we'll see for mapping digital data to digital signals. We transmit a pulse for some predetermined duration, this time period, at some level to represent the bit that we're trying to send. And note that non-return to zero, if we give these voltage levels, this is zero volts in the middle. That's what non-return to zero means. We're either transmitting a positive voltage or a negative voltage. We don't return to zero voltage. So this may be minus five volts plus five volts. The value doesn't matter. It's just positive or negative. Easy, that one, non-return to zero level. The next one is, and there's a spelling mistake here, non-return to zero inverted. It's a variation of the previous one where we invert the level when we have a bit one to transmit. So this is the same sequence as the previous example. If we have a bit one to transmit, we change the level from the previous old one to the next one. So we're high when we have a bit zero, we don't change the level. We maintain the level. So bit zero, we maintain as the previous bit high. We have a bit one to transmit, therefore we change the level, transmit for that pulse period. Next bit is bit one, so change the level. Bit zero, don't change, and so on. Non-return to zero, again, it's just negative and positive. It doesn't reach zero. Inverted, or an easier way to remember. NRZI, meaning invert on once. It's a good way to think. We invert the level. We switch the level when we have a bit one. Quickly, what's the data? Try and yourself determine the data in this case, just so you understand what's happening. This is using non-return to zero, inverted. First bit, zero in this case. In fact, the first bit do we invert on... Well, if we have some original value, if it was a one, then I would have shown that, okay, we started here and inverted. So we actually need to know what was the previous value to determine. Zero, you see, whenever we have a bit one, the level changes. Any problems with that one? So we're going to go through some examples, and then after we go on through the examples, we'll come back and discuss which ones are better, why we need multiple different techniques. Almost all of the ones we'll go through are still used in practice. All right, one of the concepts that we may need to introduce. Let's say on this plot, the pulse duration, this period between the vertical dashed lines is 2 milliseconds. So 2 milliseconds, 2, 2, and so on. How many pulses per second in that case? How many per second? Just some sort of mathematics from eight years ago or so, for most of you. Just maybe even primary school mathematics. So if each pulse is 2 milliseconds, then in one second we can have 500 pulses. Okay, now a more general term, instead of calling it pulse, when we deal with any signal, we'll talk about a also a signal element, signal element. You'll see on the slides it's defined, but we can also say this is one signal element, another signal element. And later we'll talk about and do some analysis and say, well, if we're sending 500 pulses per second or 500 signal elements per second, well, that's our signalling rate, the rate at which we send our signal. So in this case, 500 per second is the rate. Another term for that is board. So you may have heard of board, B-A-U-D regarding old modems, not board in terms of the lecture. In this case, what's our data rate? How many bits per second? Louder, I can't hear. How many bits per second? Bits per second. If our signal element or pulse duration is 2 milliseconds, then we're going to have 500 pulses per second. Each pulse in this case is quite simply represents one bit. So we have 500 bits per second in this case. 500 pulses or signal elements per second is our signalling rate. Our data rate is also 500 bits per second. We'll see when we look at other schemes that they may be different. We will distinguish between the data rate, the number of bits per second we send, and the signalling rate, the rate at which our signal changes. That will become apparent when we look at some later examples. But first, we can talk about a signal element, just one unit of that signal. Of course, how long I made up this 2 milliseconds, it will be defined by the standard that technology uses as to what the duration needs to be. Okay, here's another third scheme, bipolar AMI. AMI stands for Alternate Mark Inversion, which means Alternate Mark Inversion. Mark means 1, a bit 1. We invert the signal when we have... Yeah, we invert the level when we have a next bit 1. So on successive ones, we change the level. When we have a bit 0, we maintain a 0 level, 0 volts. That is the middle point here. So I don't have a scale, but say this is 0 volts, negative, positive, with bipolar AMI, bit 0, 0 volts, or no line signal. When we have a 1, we have either positive or negative, and we alternate for each bit 1. What's the data? Write down the data in this case. Okay, and if you didn't catch my definition, they're on the slides. Bipolar AMI, no line signal for bit 0. No line signal means 0 volts. No output there. Positive or negative level, alternating for successive ones. So first bit 1, maybe positive, next bit 1, negative. And we alternate, positive, negative. First bit, 0, this is 0 volts at this level. So 0, 1, back to 0, no line signal, or 0 volts, so it means a bit 0 here. So here's the alternate, positive, negative, okay. And we see positive, negative, for bit 1s, positive, negative, positive, negative, positive, negative, and keep going like that. So this is one where we have three levels in effect. One level for bit 0, two levels for bit 1. Same as the other schemes, we have one bit per signal element, per pulse. Let's look at some characteristics of, using this one as an example, some characteristics of signals, some advantages and disadvantages. Here's using bipolar AMI, but the data is 1, 6, 0s, 1, okay. So in this example, the data has a long sequence of zeros. And we see the signal generated, 1, 0 volts for some period of time and then 1. And if you imagine that there were 100 sequence of zeros there, then it would simply be 1, then a long sequence of zero volts and then negative. We transmit this across our communication system and the receiver receives it. One of the problems with some of the schemes, including bipolar AMI, is that when we have such a long sequence of zeros, it's hard for the receiver to determine where does the next bit start. This is what the receiver receives. So the receiver has a clock trying to determine every 2 milliseconds the next bit, but it may not be exactly synchronized with the clock of the transmitter. The clocks are not perfect. So especially over this duration, where is the next bit? Visually, you cannot obviously see I've removed the lines. So what the receiver does is measures from here, okay, our signal change. That means there's a new bit because by definition, if the level changes, that must be a new bit. Then 2 milliseconds later, but again, if the clock at the receiver is not accurate, where is the 2 milliseconds? Is it here or here? So it may lose some accuracy in determining where is the next bit. And with a long sequence of zeros in this case, it may turn out that it cannot detect, it loses synchronization with what was transmitted from the transmitter. That's a problem. That is with some systems with a long sequence of bits usually of the same value, it can make it difficult for the receiver to determine where is the next bit. Because of course, the receiver has to use a clock, but unfortunately in practice clocks of the transmitter and receiver are not the same. So synchronization is an issue. Some schemes have inbuilt methods to automatically synchronize. This one doesn't. We'll see another one that does shortly. So a disadvantage of bipolar AMI and also with the output the previous two is that synchronization is difficult. We can lose synchronization with the transmitter. This is another example of bipolar AMI. What's the data quickly? Determine the data with bipolar AMI. First bit, zero, zero. What's next? Zero. What's this bit? It's an error. Something's gone wrong, but with bipolar AMI by definition it should be positive or zero and if we had a positive the next non-zero value should be negative. But for some reason if this is the receive signal we received a positive, it tells the receiver there's an error. Something's gone wrong in the transmission or in the communication system. So this is an advantage of some schemes. Including bipolar AMI if there is an error the receiver can detect that. So in this case if the receiver receives this signal it knows at this point something's gone wrong. There's an error in transmission either it will take some action to inform the transmitter that there's an error and try to fix that error. So some schemes have inbuilt error detection including bipolar AMI. We'll see some other advantages and disadvantages soon. A fourth scheme, Manchester encoding. I always forget. Manchester, bit zero transition from high to low in middle of interval and bit one is the opposite. We transition from low to high. So we still have this interval say of our two milliseconds, but now we're making a transition in the middle of the interval. Determine the data in this case. If we use Manchester encoding what's the data if we receive this signal? What's the first bit? So again we need to know the definition. Bit zero we make a transition from high to low and bit one is the opposite from low to high. So here we have a transition from high to low in the middle of the interval. So this is bit zero. So here the next bit we transition from low to high, bit one. And then the next bit again we transition from low to high. So it's a bit one. So we see what happened here that the previous bit this bit one we ended up at high. The next bit bit one by definition we must make a transition from low to high. So in fact we transition to low and then from low up until high. So every bit there's a transition in the middle of the interval. For some bits there's a transition at the start of the interval. In this case we had from low to high bit one. The next bit bit zero must be from high to low. Therefore we maintain the level and then make that transition in the middle. So Manchester encoding in this case. In this case with Manchester encoding the rate at which we send signals would be defined as if this period was two milliseconds we're defined the rate at which we change is every one millisecond because we can have changes of the signal level at a rate of every one millisecond. So because we see here for example we have a change then in the middle of change or a change in the middle of change and so on. So every one millisecond we potentially change our signal level. In this case in our previous cases we said the signal signalling rate was 500 pulses per second gave us a data rate of 500 bits per second. With Manchester encoding in the same conditions we'd say the signalling rate is what 1000 pulses per second a maximum. In some cases we don't change but we define that the signalling rate in this case because every one millisecond we potentially change. So signalling rate is 1000 pulses per second or 1000 signal elements per second. What's the data rate? What data rate do we achieve with Manchester encoding? If our signalling rate is a thousand per second the data rate is 500 bits per second. We're still the same as the previous ones in that we're still sending one bit every two milliseconds so still 500 bits per second except we need to change the signal at a much faster rate 1000 times per second which is more complex that is to send the same amount of data our transmitter needs the electronics the circuitry to change the signal at a faster rate which is harder to do. So here we have a higher signalling rate but the same data rate. Ideally we'd like to have a high data rate and a low signalling rate. What if we have a long sequence of zeros with Manchester encoding? Will the signal be the one level all the time? In bipolar AMI we saw a case if we have a long sequence of zeros our signal stays at one level that's bad for synchronization because we can get out of sync. With Manchester encoding that's not a problem because every bit it doesn't matter if it's a bit one or a bit zero there's a change in the level so we always determine okay here's a change in the received signal that indicates that it must be a new bit or it's part of a new bit. So synchronization is not a problem with Manchester encoding. The lecture notes define some others as well. Looking at the first six we've gone through NRZ level non-return to zero inverted bipolar AMI. We just went through Manchester encoding. There's a variation of Manchester called differential Manchester and there's another one similar to bipolar called suit eternity encoding. We'll mention the last two on some later slides. Those first six I don't ask you to remember the definitions you should remember the definitions of the first two okay. So in the exam if I say here's some data draw the signal if non-return to zero inverted was used you should be able to draw the signal. If I ask you here's a signal encoded with Manchester encoding then I'll give you this definition I don't want you to I don't need you to remember these these other ones. Remember the first two the simple ones the others I would define them for you but you need to be able to understand what it means by no line signal means zero volts. These ones at the end B8ZS HDB3 extend upon bipolar AMI and solve this problem of a long sequence of zeros. So when we have a long sequence of zeros we get a flat signal. These two introduce some special signal when we have a long sequence of zeros. We're not going to cover them you can have a look and see some examples in some later slides. There are others as well this gives examples of each of them slightly different than ours. Which one's best what are the advantages and disadvantages or quite quickly there are issues with synchronization some encoding schemes can help with synchronization others cannot. Some encoding schemes when we transmit the signal using that encoding scheme we use more spectrum more bandwidth than others. We want to use as little bandwidth as possible but some use more bandwidth than others. That point about bandwidth is captured in this diagram. Without looking at the details think of this as our frequency spectrum plot. Here's our frequency and our signal strength the wider the signal the more bandwidth and the lines are for the different encoding schemes NRZ bipolar AMI Manchester. Generally we want a signal that consumes a small amount of bandwidth so is narrow in this plot and is not at zero hertz because that can introduce more errors. So in terms of the the signal spectrum the best ones in this plot are B8ZS HDB3 they are the narrowest and therefore can find the energy in a narrow bandwidth. Manchester uses a larger bandwidth NRZ uses a large bandwidth and also has this zero frequency component a DC component. So there are some trade-offs in terms of the signal spectrum for each of them. Some can detect errors some cannot so this concept of this built-in error detection. Some work better in the presence of errors in the presence of noise fewer errors and some are more complex than others. For example Manchester encoding we need to generate the signal twice as fast as the others to transmit the same data so our transmitter and receiver need to be more complex when we implement them so the cost goes up. The other details here we're not going to go through you can look through but some description of the schemes but I think what we've gone through with the examples is sufficient. This describes these two of B8ZS and HDB3 but we'll not cover that in the exam we don't need to remember it. Some more examples. Where are they used? RS232 USB so the technologies for sending cross cables use non-return to zero in its variants. Manchester encoding used in LANs so I plug the LAN cable into my laptop I transmit a digital signal to the receiving device encoded using Manchester encoding. Some telecommunication companies to connect cities across cities between countries in the past use what was called T carrier and E carrier systems used the bipolar AMI and these others HDB3 and so on. So they are used in practice in real communication systems so we vary the digital signal level depending upon the input bit and there are different techniques for that. Let's look at some examples of if we still have digital data zeros and ones but now need to transmit an analog signal some continuous waveform. Well we've got three basic approaches because our analog signal we can think of as a a sinusoid a sine wave or a combination of sine waves and if we recall our signal equation from one of our first lectures A of the signal as a function of time equals the peak amplitude times sine of 2 pi times the frequency times time plus the phase phi and there are three parameters we can vary to change the shape of that signal the peak amplitude a the frequency f and the phase phi so three parameters so we vary them to carry our different bits 0 and 1. This diagram shows some example of data we want to send it's at the top the sequence of bits ignore the digital signal there that's not useful for this example then it shows an output analog signal if we're varying just the amplitude and the scheme in this example is with bit 0 transmit with amplitude a to be 0 and with bit one transmit a signal with amplitude of some positive value one for example and we see with the first two bit zeros we get zero signal out then with the bit one we get some sine wave because we have a positive amplitude and so on bit zero one and so on so here this is a signal with two different amplitudes one amplitude of zero and another of say one so we vary the parameter a in our equation of course we can vary the other parameters as well frequency and phase the second case is when we vary the frequency here we have one frequency a low frequency signal for bit zero and a higher frequency for bit one so you can see our low frequency sine wave for bit zero bit zero and then for bit one it's higher frequency double the frequency in this example so we change the signal of our output analogs we change the frequency of our output analog signal depending upon the input bit and what the receiver does as they receive this signal they measure that characteristic if they can measure and determine okay over this period what is the frequency and if they determine it's a low level then they know it's bit zero if it's a higher frequency a bit one what are the exact values what are the exact frequencies they are defined in the different systems so whether it's one hertz one megahertz or whatever that would be defined in advance last one vary the phase remember the phase shifts the signal in time a phase of pi for example you think of our sine wave usually goes up and then down here with a bit zero we have a phase such that it goes down and then up and then with this bit one here the normal phase where we go up and then down so this is varying the phase here's a definition of a very simple case let's say we use a variation of the amplitude so a high amplitude sorry a low amplitude means bit zero a high high amplitude bit one in fact we'd define what exact amplitude here I'm just saying high and low but we could say okay plus one plus three given that definition what's the data well my diagram doesn't join greatly but you can see at least the amplitudes the low and high amplitudes here we need to define the time period of each bit or each signal element so I put some vertical dash lines to indicate this is the duration of one signal element in this scheme so when I want to transmit a particular bit I send my signal for that period of time so from that you can work out the the data in this signal what are the first three bits first three bits are zero zero one okay and I think you should be able to work out low low high low low low high high high and so on okay so simple mapping from the over this duration measure the amplitude and by the definition we said low amplitude bit zero high bit one it can be different okay but we need to have this mapping defined between the signal element and the bit these pictures will be on the website or actually already on the website so you can have a look at them in detail let's say our time period we still have two milliseconds so between the dash lines two milliseconds what's the frequency of our signal in this case what's the signal frequency if between the dash lines the vertical dash lines is two milliseconds what's the signal frequency what's the frequency of the this sine wave 500 watt 500 watt the frequency 500 hertz okay because we see if this is two milliseconds we have one repetition of our sine wave and the definition of frequency is how many repetitions per second you see it's in two milliseconds we have one in the next two another one repetition one cycle okay so in one second we'd have 500 cycles so our frequency is in fact 500 hertz in this example and we'd say that the amplitude we can say the phase in both cases is zero because we just follow the normal shape of the sine wave in the when we have a bit zero we have some amplitude which is a small value all right we don't show the the value here let's say one and we have a bit one we have an amplitude of two so twice if it's twice the value so we just see that we're varying one of the parameters to indicate the different bits how many bits per second in this case data rate how many bits per second to help that's the data how many bits per second 500 bits per second still because each t of t two milliseconds we're sending one bit so we still have 500 bits per second and in fact we'd say the signaling rate is also 500 signal elements per second all right as se per second 500 signal se signal elements per second we'll come back later for a definition that is 500 uh yeah this is one signal element a second signal element and so on let's look at another scheme let's vary two parameters of our sine equation at the same time remember we've got three parameters amplitude frequency phase in this scheme I'm going to vary both the amplitude and the frequency and the definition I've given is shown in this diagram when I want to transmit the bits zero zero I'll transmit a signal with a high frequency and a if you can compare to this one a low amplitude so the phase is going to be zero in all cases so we'll ignore the phase when I want to send bits zero zero I'll set the frequency to a high value and from memory it will be two times the original one so 1000 hertz and the amplitude to some low value let's say one that's the top left part of the diagram when I want to send a bit zero one I'll send it the same amplitude but a lower frequency in fact the same frequency as before and I'll not keep writing when we want one zero we'll have a high frequency we'll have more cycles per second and a high amplitude one one low frequency low high amplitude given that encoding scheme of data to signal what's our data and there's the hint there's the signal element duration so that again the dashed lines are our two milliseconds one signal element what's the data in this case try and work it out what's the first bit zero in fact now we have four signal values so we look at pairs of bits every two bits at a time so in this signal element duration we transmit a signal what is it it's a high frequency and it's a low amplitude compared to the others we'll see it's in fact this one which is a higher frequency than these two and a lower amplitude than these two which means that over this duration of two milliseconds the two bits sent are zero zero the next two bits if you do some pattern matching you'll see I think it's the same high frequency but also a high amplitude which corresponds to one zero zero zero one one one zero zero one so we don't have to vary just two we don't have to have two different levels or values we can vary multiple given the same conditions as before our t equals two milliseconds that is this period is two milliseconds what's our what's our first our signaling rate this is one signal element if this is two milliseconds and one signal element we have the same signaling rate of 500 per second what's our data rate now how many bits per second are we sending if this is two milliseconds how many bits per second use your phone to calculate two milliseconds what's the bit rate data rate can you help help them out what's the data rate one thousand bits per second okay because we're sending two bits now every two milliseconds or one bit every one millisecond one thousand bits per second so with this scheme compared to the previous one our signaling rate is the same but our data rate has gone from 500 up to one thousand bits per second why because we now have four different levels or four different types of signal to carry our bits that is we can carry two bits per signal element you've seen this concept before when we looked at the Nyquist capacity we saw the data rate was b to b times log base two of m m was the number of levels when we increased m from two to four it doubled our data rate when we increased it up to eight we another doubling of the data rate this is the same concept that if we represent our data using more values more levels here we have four levels we get a higher data rate when all other conditions are the same go back to our slides to define some of these so we're sending digital data as analog signals the process of converting the digital data to the signal is performed by modem modulator and demodulator and the three techniques that we've described we vary either the amplitude phase or frequency and they're called amplitude shift keying phase shift keying frequency shift keying ask psk fsk the signal that we generate has some bandwidth according to what our input carrier signal was so in my examples okay what is the frequency of this signal well it's called the carrier signal it's whatever we define it to be so in my case i define my frequency in the first case to be 500 hertz but it can be five megahertz 10 gigahertz that's called the carrier signal that carries the data by varying this signal according to our parameters in this case we varied the amplitude of that signal to represent the different data in this slide ask amplitude shift keying b fsk what's b what's b mean there so fsk is frequency shift keying we were shifting the frequency b means binary it's commonly used to refer to binary frequency shift keying because we use two different frequencies here binary phase shift keying two different phases we don't have to be limited to two we can increase it to four and more so instead of just two frequencies we have four different frequencies one frequency represents two bits another frequency another two bits and so on or we can have 16 frequencies each representing four bits and so on and like we saw in our example here we can even combine them together where we have a combination of amplitude shift keying and frequency shift keying that is different frequencies and different amplitudes some examples of where they're used uh turns out in practice amplitude shift keying is the most inefficient of the techniques it uses the most bandwidth it's used in very slow systems or when we have a lot of bandwidth like in optical fiber it's simpler than the others frequency shift keying is used in some wireless systems radio and in coaxial cable say cable tv phase shift keying used in also many wireless systems like wireless LAN mobile phones and so on and instead of just using for example two phases b psk for binary you can use say four phases and you get q for quaternary meaning four phases and common thing is to combine amplitude shift keying with phase shift keying and one of the technologies is called quadrature amplitude modulation quam so in mobile phones and in and other wireless systems the signals are generated using these two techniques combined and in adsl is another example so these systems are used when we want to send digital data but using an analog signal the more frequencies amplitudes or phases we use with the same signaling rate the higher the data rate but the higher the complexity more chance of errors things going wrong so you see things with 256 levels and higher for some systems let's not go through differential phase shift keying some examples or i think we've mentioned most of them wi-fi cable modems mobile phones for phase shift keying some radio systems for frequency shift keying optical fiber rf ideas for between small sensors for amplitude shift keying some examples so we've gone through two both sending digital data using first digital signals then analog signals last 10 minutes let's skip to the last one we'll skip over analog data digital signals we'll come back to that tomorrow let's go to analog data analog signals and just quickly go through the examples because they are easy here we have some analog input data which in this case is the middle line in this example let's say this is our input data of course it would be more complex than that it may be some audio someone talking and it's varying over time and what we do is we change what's called our carrier signal the top one again varying either the amplitude frequency or phase of that carrier signal in this case we're using amplitude modulation we change the amplitude of our carrier according to the amplitude of our input so if this is the data the output is the one at the bottom where you see we have the same frequency as the carrier the signal here but the amplitude of this is varying we have a high amplitude here when we have a positive input and a low amplitude with this negative input this is am as in am radio with am radio we have say the input voice the input data that we want to send someone talking we have a carrier signal at say 900 kilohertz that's the frequency or the channel and we vary that carrier signal by changing the amplitude according to the input amplitude modulation fm as in fm radio frequency modulation same concept except now change the frequency of the carrier according to the input so this is our data in the middle this is the output signal at the bottom and we see the carrier signal we have small variations in the frequency of the carrier signal according to the input data we have a high input here we'll see the frequency is lower it's spread out it's not so rapid a negative input we've got a higher frequency and we vary the frequency of the carrier according to the input data so fm radio when someone's talking there's a carrier signal at say 105 megahertz and that frequency varies a little bit in either side so it's not exactly 105 megahertz the the signal that's transmitted is varying according to the input audio that's being sent so this is the output signal which is transmitted to your car your car receives this signal and the variations in the frequency if it's a low frequency then it knows that the output audio has a high amplitude if it's a high frequency then it has a negative amplitude here and the one before that was phase modulation which turns out looks as the same as frequency modulation because when you look at varying the phase it's here we see we start we go down and we have a different phase as we change the phase continuously we get the same effect of changing the frequency so the two common ones are am and fm amplitude and frequency modulation why do we need them why would we send an analog data like audio using an analog signal why not just send the input analog directly the reason we use these schemes is to shift the frequency of the that we use in the transmission system so if this is your voice so go back to am this is your voice the typical range of voice is around 300 hertz to 3 kilohertz that's the range of frequencies if we want to send a signal at say 3 kilohertz from some radio tower to your car then to transmit it that very low frequency we need either a very big antenna or a very short distance so what we do for practical reasons is we instead of transmit the frequency direct as say 3 kilohertz we shift the frequency of that voice using amplitude modulation up to say 900 kilohertz that's a typical range of am radio so we carry that analog data on an analog signal carry a signal the last one is we have digital we have an analog data and we want to send a digital signal and we'll go through that tomorrow we'll see that we need to use some way to to sample the analog data to get some digital output let's stop there and we'll cover the last part section three hopefully finish that tomorrow