How to Draw a Constellation Diagram

Digital Modulation

Walter Ciciora , ... Michael Adams , in Mod Cablevision Television Technology (Second Edition), 2004

Furnishings of Interference on the Constellation Diagram

The constellation diagram is useful for understanding the effect of interference on the betoken. In Figure 4.12, quadrant I illustrates the effect of a continuous wave (CW) interfering indicate in the passband of the receiver. An interfering carrier creates a beat note, which forces each point in the constellation to spread out in a circumvolve around the desired indicate. The radius of the circle is proportional to the relative amplitude of the interfering carrier. Unlike analog video signals, the impact of the interfering carrier tends to exist independent of its location inside the passband, except for some effect caused by nonflatness of the receive filter.

It is worth noting that composite 2d order (CSO) and composite triple beat (CTB) are not CW interferences. Their amplitudes change constantly, with peaks reaching fifteen dB to a higher place the average. The variation is due to phase alignment of the multiplicity of carriers. The occurrence of a peak is rare, but the peak can last from a microsecond to a few hundreds of microseconds (longer than the burst protection menstruum of the Interleave. ⁴

Quadrant II illustrates the event of phase dissonance on the signal. Phase noise is introduced when a local oscillator has significant phase dissonance. It tin can also be shown to occur when the transmission aqueduct does not have a response that is symmetrical well-nigh the carrier. Phase noise causes the constellation to rotate about the origin.

Random noise (quadrant III) causes each signal in the constellation to "blur," somewhat as with CW interference, except that the spreading of the signal tends to be more than about uniform within the circumvolve, and the circle does non have a well-defined radius. Again, recall that the spreading of points in the constellation becomes a problem when the spread points cross a threshold to another state. With QPSK, that threshold is defined past the axes of the constellation diagram. With more dumbo modulation, the thresholds are closer together, and so lower levels of interference volition crusade the threshold to be crossed. As a practical matter, an fault may occur when a point is forced close to a threshold purlieus: Real demodulators don't work quite likewise every bit the theory predicts.

Finally, reflections, which cause ghosting in analog transmission, cause a replication of the constellation effectually each point in the constellation, every bit illustrated in quadrant IV. The constellation is rotated with respect to the "direct path" constellation, due to phase shift between the directly and reflected RF carriers. Usually, systems that employ denser modulation formats include adaptive equalizers, which can recoup for echoes if the echo doesn't modify as well chop-chop.

If the peak aamplitude of the signal is i, as shown in quadrant Iv, then the distance from the nominal location of whatever of the 4 states to the nearest state boundary is 0.707. This result is easily obtained via elementary geometry, where the amplitude vector, of length i, is the hypotenuse of a right triangle and the distance to the nearest conclusion threshold is ane of the ii other sides of the triangle. Bear in mind that it is more mutual to measure the amplitude of a digital betoken by using the average amplitude, which concept is explored later.

Read full chapter

URL:

https://www.sciencedirect.com/science/article/pii/B9781558608283500060

Optical performance monitoring of optical phase–modulated signals

Bartłomiej Kozicki , in Optical Performance Monitoring, 2010

12.3.1.seven Asynchronous amplitude histogram

The monitoring techniques employing constellation diagrams provide ample information about the modulator characteristics and OSNR level affecting the signal. Nevertheless, no methods employing the constellation diagrams to evaluate the influence of transmission impairments in phase-modulated channels take been presented to engagement. In gild to obtain data about the level of CD or PMD affecting the signal, the samples need to be digitally processed. One of the methods to analyze the samples is to build a histogram of indicate amplitude values. Such an approach is used to obtain a synchronous center diagram, which can be used to extract information about point noise, jitter, or waveform distortion. Still, it requires a plush and format-dependent clock recovery circuit. A powerful OPM technique using asynchronous amplitude samples has been demonstrated for OOK signals. ⁹¹ The asynchronous amplitude histogram (AAH) OPM technique takes advantage of the statistical properties of the amplitude histogram to excerpt information nigh signal OSNR, accumulated CD, and PMD besides in phase-modulated signals. ^92–94 The AAH technique satisfies many of the requirements put on the OPM system. Information technology is asynchronous and allows monitoring of multiple bespeak impairments.

The histogram is created past outset sorting the samples past their value. The values are mapped onto histogram bins that uniformly divide the dynamic range of sample values into northward levels. The histogram is formed by counting the number of amplitude samples falling into each of the bins and plotting the count as a function of the bin value. An example of an amplitude histogram for a 10-Gb/s RZ-DPSK point is shown in Figure 12.16. The horizontal axis of the histogram represents the number of sample points in respective bins, while the vertical axis corresponds to the sample value. The peaks of the histogram correspond to the valley and the tiptop of signal waveform, respectively. The samples in between the peaks correspond to the crossover points of the rising and falling edge. In order to analyze the shape of the histogram, a distribution is fitted to each of the peaks. Histogram parameters reflect the properties of signal waveform. Therefore, by tracking the statistical properties of the histogram, it is possible to evaluate the level of impairments affecting the optical signal.

Figure 12.16. Waveform of RZ-DPSK betoken and respective asynchronous amplitude histogram. Figure as well shows parameters used for monitoring bespeak impairments. ⁹²

OSNR monitoring using AAH relies on dependence of the variance of signal-spontaneous beat out racket on the ability of ASE racket. In the case of an NRZ-DPSK betoken generated by a stage modulator, the waveform is a constant-amplitude bespeak, as illustrated in Figure 12.17(a). The amplitude histogram acquired from the waveform consists of a single peak with mean value μ _avg . The fluctuation of waveform aamplitude due to signal-spontaneous trounce noise can be observed through the standard deviation σ _avg of histogram distribution, which is inversely proportional to the level of OSNR.

Figure 12.17. Waveform and the respective AAH for the NRZ-DPSK signals. (a) Parameter for monitoring OSNR. (b) Parameters for monitoring of CD. ⁹²

With this technique, the OSNR was evaluated from 22 to 38 dB in a 10-Gb/s NRZ-DPSK signal. ⁹² OSNR monitoring has besides been demonstrated for RZ-DPSK and RZ-DQPSK formats. In this case, the histogram differs from that of the NRZ-DPSK waveform. It is comprised of two distribution peaks corresponding to the RZ pulse meridian and the level of waveform valley betwixt the pulses. OSNR reduction due to the change in ASE noise level causes a waveform fluctuation due to signal chirapsia with ASE noise. Waveform fluctuation results in a reduced ratio betwixt the histogram elevation height and histogram summit at median level, every bit shown in Figure 12.16. This phenomenon enables evaluation of the OSNR from 17 to 27 dB for the RZ-DPSK signal and from 17 to xxx dB for the RZ-DQPSK indicate. ⁹² In this range the parameter scales linearly with the OSNR.

The AAH also serves to evaluate the level of residual CD affecting the betoken. In the example of the NRZ-DPSK betoken, the change of phase between $.25 introduces a frequency chirp proportional to the phase-modulator bandwidth. CD aggregating results in phase-to-amplitude conversion. This tin can be observed in the histogram that develops multiple peaks proportional to the conversion level. As CD does not influence the mean or standard departure of the central histogram peak, information technology is possible to evaluate the accumulated CD level independently from the OSNR level. For a x-Gb/south NRZ-DPSK, evaluation of CD is possible from −600 to +600 ps/nm. ⁹² The issue of CD aggregating in the RZ-DPSK or RZ-DQPSK signal can exist observed in the waveform as spreading of pulses in time. This is reflected in the amplitude histogram, equally shown in Effigy 12.16. By measuring the altitude between the peaks of the histogram, a parameter for evaluation of the accumulated CD tin be established. CD monitoring for a 10-Gb/s RZ-DPSK signal and twenty-Gb/s RZ-DQPSK point was demonstrated from −600 to +600 ps/nm. ⁹² Equally the AAH reflects the statistical backdrop of waveform amplitude, it is also sensitive to other transmission impairments, such as PMD. PMD monitoring has been demonstrated using the AAH for DGD up to 50 ps in the x-Gb/south RZ-DPSK ⁹⁴ and 20-Gb/s RZ-DQPSK ⁹² signals.

An alternative approach to monitoring the performance of phase-modulated signals using histogram analysis was presented in Reference 95. A self-clocked sampling module based on sum-frequency generation recovers a synchronized waveform of a demodulated channel. The samples forming the heart of the obtained middle diagram are used for construction of the histogram. A Q-parameter is calculated from the statistical properties of the histogram and related to transmission impairments in x-Gb/south RZ-DPSK and 40-Gb/s RZ-DQPSK signals.

Read total chapter

URL:

https://world wide web.sciencedirect.com/science/article/pii/B9780123749505000122

Thought of all-digital I/Q modulator*

Morteza Southward. Alavi , ... Robert Bogdan Staszewski , in Radio-Frequency Digital-to-Analog Converters, 2017

6.one Concept of Digital I/Q Transmitter

Fig. vi.1 illustrates the concept of the digital I/Q modulator. The desired IQ is constructed by vectorial summing of their composite I and Q digital vectors. Their code resolution (North _b ) must be high enough to encompass all I/Q points of the corresponding trajectory connecting the symbols [22].

Fig. 6.1. Digital I/Q modulation concept; Its related IQ constellation vectors.

This indicates that, for supporting only an m -symbol constellation diagram, the resolution of the digital I/Q modulator should be at least ⁱ

(six.ane) $\begin{array}{l}\hfill {\mathrm{Northward}}_b\ge {log}_2\left(\sqrt{\frac{\mathrm{thousand}}{4}}\right)\end{array}$

In improver, North _b as well affects the subsequent quantization noise, which is discussed in more detail in the following chapters. A significant result related to whatsoever transmit modulator is its agility to traverse from one I/Q point to another. Equally graphically depicted in Fig. 6.one by P ₁ and P ₂ paths, traversing along P _two trajectory instead of P ₁ makes the complex baseband modulation faster and, consequently, the modulator must manage a wider bandwidth as well as a higher sampling rate. To practise so, based on the arcadian cake diagram in Fig. vi.2, the I _BB and Q _BB digital baseband signals are upsampled as I _BB-upward and Q _BB-upwardly . This process ensures that the spectral images will exist attenuated and located far away from the carrier and thus can hands be filtered out. The I _BB-up and Q _BB-upwards are 2 × North _b -flake (N _b for in-phase equally well every bit N _b for quadrature component) upsampled digital signals, which should be direct upconverted to their continuous-time reconstructed RF output signal. As a outcome, these signals are applied to a pair of DRACs, comprising an array of 1-bit unit of measurement cell mixers and i-fleck unit of measurement jail cell DPAs.

Fig. vi.2. Digital I/Q modulation concept; Its related idealized block diagram.

The DRACs are clocked in tact of differential quadrature upconverting clocks I _P , I _N , Q _P , and Q _N . ² According to Fig. 6.1, the 4 quadrants of the constellation diagram must be addressed by the modulator. The switching between quadrants tin be achieved by swapping between I _P /I _North or/and between Q _P /Q _N according to the sign $.25 of I _BB-upwardly and Q _BB-upwards . The DRAC outputs are connected to a power combiner that facilitates the conversion of the upconverted digital signals into the reconstructed RF output. In fact, the digital I/Q modulator represents an RFDAC. In this approach, however, the primary challenge is related to the orthogonal summing of the I and Q DRAC outputs in order to reliably reconstruct the modulated RF betoken [26, 116–118].

Read total chapter

URL:

https://www.sciencedirect.com/science/article/pii/B9780128022634000064

Circuitous Electronic System Design Example

Peter Wilson , H. Alan Mantooth , in Model-Based Engineering science for Complex Electronic Systems, 2013

thirteen.3.3.ii.3 Quadrature Amplitude Modulation

While AM and FM are both fine for the transmission of music and speech – basically narrow band audio signals – in data-logging and sensor networks, it is of import to be able to send information reliably and securely. As we have seen with the AM OOK system, and the equivalent in an FM system frequency shift keying, the systems are fine for the manual of single bit streams; however, this does lead to fairly low data rates (unremarkably only a few kbits/due south). If we need to transmit multiple $.25 in one step, then a dissimilar approach is required and a common technique often used in cellular communications and information communications systems is quadrature amplitude modulation (QAM).

The bones arroyo with QAM is to utilize two AM or FM channels using oscillators shifted by 90° (i.e., one sin(ωt) and one cos(ωt)) combined to give a number of bits per transmission wheel. For instance, if nosotros say that in an AM system, where the digital inputs in each channel (incident and quadrature – otherwise known every bit I and Q) are implemented using OOK, then we tin transmit a "symbol" that consists of two bits at a fourth dimension instead of one. The effective bandwidth is double that of the single message being transmitted.

In that location are a few dissimilar options for demodulation, including zero-IF (direct conversion), low-IF, or loftier IF. Null-IF is the simplest scheme as information technology does not require an extra stage of demodulation at the intermediate frequency (IF), and zippo-IF will only work for an IQ system. The overview of a QAM system is given in Effigy xiii.thirteen.

Figure 13.thirteen. QAM system overview

Transmitting data using I and Q channels becomes a way of encapsulating magnitude and phase data if we consider the I and Q channels in the same way as complex data (real and imaginary). The data can then be described graphically using a "constellation" diagram, where the values of instantaneous data are plotted on an X–Y graph.

A simple scheme can exist to utilize a form of OOK in each channel to provide either positive or negative I or positive or negative Q signals, as shown in Figure 13.xiv.

Figure xiii.14. QPSK—default option

This appears simple; however, in that location are some issues. As can be seen from the rotation of the bits, the sequence is binary, not grey code, which makes the transitions uneven. Also, the coding depends on both bits in I and Q channels. An culling approach is to use the technique shown in Figure 13.15. This has the grey lawmaking sequence, and the advantage of this coding scheme is that the I channel depends on one bit and the Q aqueduct on the other.

Figure 13.xv. QPSK—modified pick

This QAM scheme can be modified to include multiple phases or multiple amplitude levels to give more symbols per transmission stage, for example 64-QAM has viii amplitude levels (2³) per channel – giving a total of 6 $.25 transmission per symbol. The pattern merchandise-off is to trade the ability to transmit multiple bits against indicate-to-racket tolerance.

This technique is too an example of a "coherent" system, where the assumption in all the constellation diagrams is that the oscillator frequencies are very accurately defined with very pocket-size errors. In practice, of form, there will be frequency and phase errors, and the result tin can be a rotation in the constellation diagram to reverberate that mistake. Therefore, practical systems do not measure the absolute phase differences simply rather the relative phase cycle-to-cycle.

Example thirteen.1

Modeling Modulators

We can use models to empathise the behavior of modulation schemes at a diversity of levels; all the same, a useful starting point is in the creation of a basic system level model of the QAM modulator described in Figure 13.13. If we start from the basis that this arrangement is operating on continuous real numbers (a typical system-level design), we can apply simple mathematical operators, such as multiply, addition, sine, and cosine functions to create the modulated signal. Equally we take seen already in this volume, we take options as to how we go about creating a model for system level elements such as this. If we have a general view of the modulator in terms of two inputs representing the I and Q channels and modulated output, we could apply an equation to implement the modulation:

(13.half dozen) $y(t)=i(t)*\cos (\omega t)+q(t)*\sin (\omega t)$

Using this arroyo in graphical modeling, we merely create a pinnacle-level symbol for the "modulator" with two inputs (i and q) and a unmarried output (y). For system-level modeling these will be defined equally "quantities" and have type "real" – in other words completely generic analog signals with no units defined at all. Using ModLyng, the resulting symbol and equation tin be seen in Figure 13.16.

Effigy 13.sixteen. System-level model of a modulator using equations

Clearly, this is a simple approach and very efficient; nevertheless, for more complex systems, sometimes the equations can become either very complicated, or difficult to manage. In such cases, as we have seen throughout this volume, it may be preferable to take a more than graphical approach to the process of creating models, and use primitive elements such as multipliers, adders, and trigonometric functions, such as SINE or COSINE. Taking this approach to model the same modulator would use a schematic description and leverage the existing simple system edifice blocks either available within ModLyng libraries or easy to make past the user.

Equally shown in Figure 13.17, the modulator is made up of several simple blocks and, although in some respects looks more complicated than the equation-level model, each of the building blocks is defined once and tin can so be reused any number of times in different models. Too, information technology is very piece of cake to and so employ the block diagram version of the model in reports or to explain concepts in design reviews to nonexperts, which can be very helpful. In both cases, the resulting model behaves in an identical fashion.

Effigy 13.17. Block-level modulator model

The same approach tin can exist taken when it comes to modeling the demodulator, and it makes a lot of sense to reuse many of the blocks used in the modulator. By adding a low laissez passer filter (LPF) we can construct a block-level demodulator as shown in Figure 13.18. The two new blocks in the demodulator are a simple gain block (y=x*a) and a LPF. The gain tin can be used to compensate for the inherent attenuation of 0.v in the demodulator, every bit shown in Figure 13.thirteen, by calculation a gain of ii.0 in each betoken path.

Figure 13.18. Demodulator block diagram

The LPF is a unproblematic model of a standard Laplace beginning lodge filter with the basic equation as shown in Eq. (13.7), where ω _c is the cutting-off frequency and g is the DC gain.

(13.7) $y=\frac{m}{s+{\omega }_c}x$

As we described in the affiliate on arrangement-level modeling, the way nosotros implement a Laplace office is to consider the operator south as a "d_by_dt" office. Therefore, if we rearrange Eq. (13.7) into the form shown in Eq. (xiii.viii) we can and so apply the Laplace operator as a derivative in Eq. (13.9).

(thirteen.8) $y(\mathrm{southward}+{\omega }_c)=\mathrm{yard}x\Rightarrow y\mathrm{south}+y{\omega }_c=gx\Rightarrow y{\omega }_c=\mathrm{grand}x-ys\Rightarrow y=\frac{\mathrm{i}}{{\omega }_c}\left(kx-y\mathrm{due\; south}\right)$

(13.9) $y=\frac{k}{{\omega }_c}x-\frac{1}{{\omega }_c}\frac{dy}{dt}$

This can then be modeled in the graphical modeling tool to provide a LPF role. The resulting modulator and demodulator can be tested past applying two different quadrature signals (a pulse down one channel and a sinusoid downwardly the other) and comparison the output waveforms with the input waveforms. The resulting test bench tin be seen in Effigy 13.nineteen.

Figure 13.nineteen. Quadrature amplitude modulation test bench

When the model was generated (in VHDL-AMS), the resulting simulations using an guess point frequency of i   kHz and carrier frequency of one   MHz (much lower than the design 433   MHz to reduce simulation time) were carried out. The results bear witness that the modulation scheme more often than not works; however, some refinement is clearly necessary.

One of the obvious issues is that the amplitude of the output is a little lower than it should exist. So why is this happening? As we have seen, the equations do not indicate any attenuation, and the system-level model is simply an encapsulation of those equations, and then we would non look a problem with the output. The solution is a result of the use of the simulation fourth dimension. Equally is often the case in a trigonometric or integration model, every bit the time (in this case) value increases, the potential for numerical noise becomes greater. This is coordinating to integration where the absolute value becomes greater, and the resulting numerical dissonance floor besides rises accordingly. A point occurs at which numerical racket begins to degrade the quality of the calculated trigonometric function (in this example a sine or cosine value) every bit the absolute value of time increases. In fact, in this example after about 200 cycles of the carrier fundamental, the numerical inaccuracy begins to have an effect. This can be seen clearly in Figure 13.twenty, where, after nigh 200 μs, the accurateness of the output degrades dramatically, whereas prior to this bespeak the output tracks the input very well.

Figure thirteen.20. Quadrature amplitude modulation simulation of arrangement level design

So how practise nosotros address this event? The solution is to modify the fourth dimension calculation function so that we summate the menses of the carrier and then calculate the relative time within a menstruation, so that the maximum absolute value of the time variable is always less than or equal to the catamenia of the carrier frequency. We tin can do this using a elementary digital model where the resolution (number of sample per menstruum) can be set up as a parameter and the new time calculated. The resulting simulation in Effigy 13.21 shows a much ameliorate correlation between the input and output, with a small filtering effect due to the LPF, apparent on the fast edges of the pulse input.

Figure 13.21. Modified system model with relative time rather than absolute time

So, where does this leave us in terms of a pattern? Nosotros now take a simulation model of the system that includes both the modulator and demodulator. We can change the modulation carrier frequency and can exam dissimilar waveforms to analyze the behavior of the system. Nosotros can as well change key parameters, such every bit the LPF cutting-off frequency, amplifier gain, and LPF order. For instance, what would be the consequence of a one   parts per one thousand thousand (ppm) frequency error between the modulator and demodulator in the system? We can but change the carrier frequency parameter in the demodulator to +one   ppm (1   Hz) of the modulator and the resulting effect tin be seen to exist pocket-size in Effigy xiii.22. What is, in fact, happening is that there is a minor frequency error in the demodulated bespeak as a outcome of the carrier error in the demodulator, and this gets progressively worse equally the difference increases.

Figure 13.22. Carrier frequency error of ane   part per million

And so, from a designer's perspective, what is the tolerance on frequency before we meet an effect on the output? Using the model we could simply increase the difference between the modulator and demodulator carrier frequencies until the divergence became intolerable; however, this is not a particularly useful way of designing the organization (although a trial and error approach like this is nevertheless often used in many examples). In practice, we would adopt to make the system tolerant of these potential differences and so an alternative technique could be used.

Instead of applying static signals with the IQ QAM modulator, if we utilize a sinusoidal waveform at a lower frequency (say i   kHz) to the I and Q inputs with a phase difference of π/two between them, and define a coding scheme where if nosotros wish to ship a "ane" then apply sine to the I input and cosine to the Q input, and vice versa for a "0", we can see how this affects the behavior of the modulator. As we saw previously, nosotros could ascertain a static value (DC) for I and Q signals and this would outcome in a static QAM every bit we saw in Figure 13.15. Past applying a sine or cosine indicate we produce a rotating vector equally shown in Figure 13.23.

Figure thirteen.23. Rotating vector of IQ signal

On first inspection we are no further forwards than we were earlier; however, this has a pregnant advantage in that we now do non intendance most how fast the vector rotates, only rather which direction the rotation is taking place. This gives much more tolerance to errors in the signal and very unproblematic signal processing can be used to estimate the rotating phase from sample to sample and therefore the value of the data being transmitted. The downside of this approach is that we merchandise off reliability for data rate, equally this approach is past its very nature going to exist relatively slow.

The other advantage of this type of approach is that the frequency mistake will manifest itself every bit a constant error in rotation (either too fast or as well tedious) and can therefore be filtered out using a simple moving boilerplate filter over a number of samples.

So far in this section we have explored diverse architectural options for the RF part of the pattern, which volition enable a system-level exploration to take identify. At this bespeak we tin investigate the individual building blocks in more item, at least to a first cut of the blueprint.

This procedure begins with capturing height-level specification evaluative criteria. For the chip-level test bench, operation measures at the output pins based on given input sequences can be designed for fleck-level pass/fail criteria. Fifty-fifty though the underlying models are nonexistent at this signal and the test would fail, spending time to define what "good" is at this point puts the focus squarely on capturing executable specifications. Then, as a subsequent step in the process. As the next section elaborates, ideal models can supervene upon the empty models to achieve the get-go passing executable specs in simulation. For example, for a given analog value on i of the input pins, the functioning criteria could be to compare the decoded digital equivalent and assess whether information technology is inside the tolerance expected. Being a largely ideal model at this stage, any fundamental problems with the compages tin be identified before any significant fourth dimension is spent on detailed circuit design.

At this phase of the blueprint process, therefore, it is fundamentally important to ensure that the overall design is consistent and coherent, without necessarily spending a long time establishing that the performance is entirely achieved; rather, the goal is to minimize the complexity of both the model and the resulting design assay. It could be said that at this indicate, the main aim of the projection managing director is to know that all the main blocks are in place, with resources to move into a more detailed pattern stage afterwards.

This department of modeling has brought us to a bespeak where we understand clearly not only the RF design options bachelor to united states, but also some potential pitfalls when we simulate our designs later. We can at present move onto the baseband department, in particular the data converter, and look at some of those issues.

Read total chapter

URL:

https://www.sciencedirect.com/science/commodity/pii/B9780123850850000130

A wideband two × 13-bit all-digital I/Q RFDAC*

Morteza South. Alavi , ... Robert Bogdan Staszewski , in Radio-Frequency Digital-to-Analog Converters, 2017

10.9.vi Verification of DPD I/Q Code Mapping

Examining this arroyo, a 256-symbol modulation is created. Based on the concept depicted in Fig. ten.18 A, the constellation diagram is continuously swept from the meridian-left to top-correct in a "snake"-similar manner and traversed back again to its original indicate in gild to preserve continuity. Note that, for simplicity, Fig 10.18A only illustrates a 16-symbol constellation diagram likewise as their time domain representations, which is exhibited in Fig 10.18B. These signals are and then upsampled and interpolated using an RRC interpolation filter to produce I _BB-up and Q _BB-up (see their I/Q trajectories in Fig. 10.18C). Next, the resultant signals are predistorted (I _DPD and Q _DPD ) using the lookup table of Fig. ten.eighteenB and loaded into 2 on-chip SRAMs. Fig. 10.18D shows the consequence of the I/Q DPD mapping on the I/Q trajectories of the original modulated signals. The RF output signal is downwards-converted, and its corresponding I/Q trajectories are exhibited in Fig. 10.18Due east, which demonstrates a good agreement with the original I/Q trajectories of Fig. 10.15C. I _syn and Q _syn are then downwardly-sampled and decimated to create the measured constellation diagram (Fig. 10.eighteenF). Note that its related EVM, RF power, and drain efficiency are −32   dB, 16.1   dBm, and 19%, respectively. It should be mentioned that, due to the limited information length of I _DPD /Q _DPD (i.e., 8192), which are repeatedly fed to the DRAC circuit from the commencement data signal to the terminal, any discontinuity between the offset data point and the last one creates an undesirable spectral bound. To alleviate this issue and to preserve the continuity, the information length of I _BB and Q _BB are doubled and applied to the RRC interpolation filter and then only the half of the data length of the subsequent I _BB-up and Q _BB-upwardly are exploited and practical to the DPD lookup table. This technique is referred to equally wrap-around process. As a upshot, the commencement points of the I/Q trajectories of Fig. 10.18C–Eastward, indicated with circles, have been shaped in such a style equally to ensure the continuity of the I/Q signals.

Fig. 10.eighteen. DPD measurements: (A) Simplified diagram of 16-symbol I _BB-up /Q _BB-up ; (B) Trajectories of their related time domain waveforms. DPD 256-symbol constellation mapping plots: (C) I _BB-upward /Q _BB-up trajectories; (D) predistorted I _DPD /Q _DPD trajectories; (Due east) measured I _syn /Q _syn trajectories; (F) measured 256-point constellation.

Read full chapter

URL:

https://www.sciencedirect.com/science/commodity/pii/B9780128022634000106

Ultralong-distance undersea transmission systems

Jin-Xing Cai , ... Neal South. Bergano , in Optical Fiber Telecommunications Vii, 2020

13.2.two.1 Probabilistic constellation shaping

Probabilistic shaping (PS) accomplishes Gaussian-like distribution by using a set of equidistant constellation points with a nonuniform probability distribution [36] , equally shown in the constellation diagram for PS-64QAM in Fig. xiii.19. PS requires a transformation of equally distributed input bits into constellation symbols with a nonuniform distribution. Such transformations accept been implemented using prefix codes, many-to-one mappings combined with a turbo code, distribution matching, or cutting-and-paste method. PS does crave redundancy only has attracted more attention due to its capability to approach Shannon faster than geometric shaping with an equivalent number of constellation points. Fig. 13.18 shows the AIR comparing PS-64QAM with 64QAM. At 8   b/southward/Hz, PS-64QAM is >1   dB better than 64QAM and is simply 0.i   dB abroad from the Shannon limit for AWGN channel. PS too provides a straightforward power to accomplish variable SE (Department 13.2.2.3).

Figure 13.nineteen. Probabilistically shaped-64QAM.

Experimentally, PS-64QAM has been used to demonstrate a SE of 7.iii   $.25/s/Hz over 6600   km over C+L bandwidth. In this sit-in, the SE is controlled to maximize chapters over the 8.95   THz optical bandwidth [37]. Moreover, PS-64QAM has been demonstrated in field trials offering half dozen   b/southward/Hz SE over 11,000   km [38].

In general, for limited number of constellation points, PS tin can outperform circular based geometric shaping past upwardly to few tenths of a dB in shaping gain for the same number of constellation points. From Fig. 13.18, PS-64QAM has ~0.five   dB higher shaping gain than 64APSK. Nevertheless both geometric and PS significantly outperform conventional QAM formats thanks to their superior power efficiency (PE) (i.e., maximum achievable rate per unit power).

Read full affiliate

URL:

https://www.sciencedirect.com/science/article/pii/B9780128165027000154

Optical performance monitoring based on linear optical sampling

Christophe Dorrer , in Optical Functioning Monitoring, 2010

viii.iv.2 Stage and aamplitude noise measurements

From LOS information, the amplitude and phase racket on a symbol can be obtained simply by because the amplitude and phase spread of the symbols in the complex airplane. Exemplary constellation diagrams of phase-encoded sources are shown in Figure eight.11. These phase-encoded signals were generated past directly driving a LiNbO₃ stage modulator with a ii-level PRBS drive. Tuning the voltage difference between the ii levels changes the relative optical phase of the two levels and increases the noise on each level, equally can exist observed in Figures 8.11(a) and (b). However, since just stage modulation is performed, the amplitude noise does not depend on the amount of stage modulation. Figure 8.11(c) presents a quantification of the phase and amplitude noise equally a function of the phase difference betwixt the two encoded levels: the stage noise increases linearly while the amplitude noise is constant, as expected. In exercise, this approach to the generation of phase-encoded signals is avoided considering of the inherent phase dissonance induced past the noise on the drive voltage. It is preferred to use a Mach-Zehnder modulator and benefit from the sign change of the transfer function when going through extinction. This provides a phase shift exactly equal to π regardless of the modulation amplitude, and aamplitude racket can exist minimized by proper tuning of the amplitude of the modulation. These properties are demonstrated in Figure 8.12. The two constellation diagrams plotted in Figures 8.12(a) and (b) were measured for Ac drive voltages with different amplitudes, the bias of the Mach-Zehnder modulator being set for extinction. A π phase shift between levels is obtained regardless of the bulldoze aamplitude, and the phase noise does not depend significantly on the drive amplitude. However, the amplitude noise decreases equally the drive aamplitude is increased, equally the sinusoidal transfer function of the modulator clamps the amplitude modulation when the voltage amplitude is sufficient (Figure 8.12(c)).

Figure 8.11. (a, b) Constellation diagrams of phase-modulated signals generated with a stage modulator for differential phase of π/two and π. (c) Standard deviation of the amplitude σ_ρ and phase σ_ψ of i of the symbols as a function of the differential phase (respectively round markers and foursquare markers). ¹⁹

Copyright © 2006 OSA.

Effigy 8.12. (a, b) Constellation diagrams of BPSK bespeak generated with Mach-Zehnder modulator for ii different amplitudes of the drive voltage. (c) Standard deviation of the amplitude and phase of one of the symbols as a function of the bulldoze voltage (respectively round markers and square markers). ¹⁹

Copyright © 2006 OSA.

The direct characterization of the amplitude and phase racket is important in many aspects of optical telecommunications. Taking equally an example the BPSK regeneration experiment described in Reference 28, the wavelength-conversion procedure modifies the amplitude and stage noise properties of the signal. A performance assessment of the channel in terms of scrap error charge per unit only leads to an indirect understanding of the properties of the converter. Effigy eight.xiii demonstrates the phase-regeneration capability of the proposed setup when operating on signals with stage noise. The constellation diagrams measured on the input indicate (Figures 8.13(a) and (c)) and on the converted point (Figures eight.13(b) and (d)) demonstrate a reduction of the stage dissonance, which could lead to an appreciable transmission functioning comeback. The standard divergence of the phase is reduced from 0.3 rad to 0.fourteen rad after conversion (Figures 8.13(c) and (d)).

Figure 8.13. Constellation diagrams measured at input and output of wavelength converter set to reduce the stage noise of a BPSK point. (a) and (c) are measured before the wavelength converter, while (b) and (d) are the corresponding converted signals. ²⁸

Copyright © 2008 IEEE.

Read full chapter

URL:

https://www.sciencedirect.com/science/commodity/pii/B9780123749505000080

Cell Phones

Louis Eastward. FrenzelJr., in Electronics Explained (Second Edition), 2018

Orthogonal Frequency Division Multiplexing

OFDM is a broadband modulation method like spread spectrum/CDMA. It takes a loftier-speed series binary betoken and spreads it over a wide bandwidth. The serial information is passed through a circuit that maps out the constellation diagram for the modulation to be used. It and so divides information technology into many slower-speed serial bitstreams. Each bitstream modulates a carrier on ane of many adjacent carriers in the available bandwidth. This technique effectively divides a broad bandwidth into many narrower subchannels or subcarriers as shown in Fig. 8.8. Sometimes dozens of channels are used, and in other cases hundreds or even thousands of carriers are used. The carrier frequencies are selected so they are orthogonal and as a result they volition not interfere with 1 some other fifty-fifty though they are directly adjacent. All the modulated channels are and then added together and the combination transmitted in the bachelor bandwidth. The type of modulation varies with the awarding but it is normally BPSK, QPSK, 16-QAM, 64-QAM, or 256-QAM.

Effigy 8.8. OFDM divides a channel into many subchannels or subcarriers, each of which is modulated by part of digital data to exist transmitted.

The basic technique is shown in Fig. 8.9. The only practical way to create an OFDM indicate is past using DSP. The DSP executes an algorithm called the changed fast Fourier transform (IFFT) that creates the multiple adjacent modulated carriers. The DSP data is applied to DACs to create I and Q analog signals that are sent to mixers for creation of the last point. That point is then amplified by a broadband linear amplifier before being transmitted. At the receiver, the information is recovered by another DSP executing the FFT. The output is the original fast serial data.

Figure 8.9. Generating OFDM is done by DSP and I/Q modulation.

OFDM seems impossibly complex while seemingly hogging spectrum space. Yet, it can transmit college speeds in smaller bandwidths than virtually other types of digital modulation. Information technology is very spectrally efficient. Furthermore, it is more resistant to multipath interference that tin cause microwave links to lose data due to a reflected signal interfering with another or a direct bespeak causing counterfoil and fading. About of the new wireless technologies today use OFDM, including the LTE cellular systems. Some examples of other OFDM uses include wireless local area networks (WLANs) such as Wi-Fi, wireless digital subscriber line (DSL) Internet access, some (European) digital Tv set, and Air conditioning power line networking.

Read total chapter

URL:

https://www.sciencedirect.com/science/article/pii/B9780128116418000084

Information Coding Techniques in Chipless RFID

Arnaud Vena , ... Smail Tedjini , in Chipless RFID based on RF Encoding Particle, 2016

3.half dozen.1 Constellation diagram and graphic representations

To go farther in the coding efficiency improvement, we can introduce hybrid coding techniques. "Hybrid" ways that coding will be generated past more than one parameter, for case the amplitude tin be combined with the phase. In the aforementioned mode as for the telecommunications systems using conventional modulation principles, a constellation diagram tin can exist established in social club to define the coding efficiency for a given frequency, or a given resonator instead of a given time. In the example in Figure 3.xiv(a), the two parameters used are the aamplitude and the stage for a given frequency. This arroyo is very similar to an IQ modulation scheme transposed to the frequency domain.

Figure 3.xiv. a) Frequency 2D constellation diagram and b) generalization, constellation diagram in three, or fifty-fifty N dimensions

In certain cases, more than two parameters can be modified for a given symbol (in the temporal or frequency domain), so that a constellation diagram with N dimensions can be adopted to represent N independent parameters, for a given frequency, as indicated in Figure three.xiv(b).

Previous constellations provide data on the coding efficiency for a given symbol in the temporal domain or more precisely here, in the frequency domain. A variant tin can exist used by integrating the frequency axis (or the time axis depending on the type of coding) to graphically display the generated ID. In this case, the constellation does not provide the total number of possible combinations for each frequency (or time) interval, but the whole of the lawmaking spread out over several frequencies (or in time). Such a representation may be useful to compare the unlike IDs generated by the tags, in club to make up one's mind the possible interference. We can imagine that a code recognition technique can exist implemented with the use of a graphical method past comparing the constellation modified with the expected responses. Still, in the case where increased detection robustness is necessary, a subset of codes can be selected to limit the possible recognition errors between ii similar codes. For this purpose, a minimum Hamming distance must be imposed.

For case, the most used conventional coding technique in the frequency domain is to attune the presence or the absence of a resonance peak in the spectrum. This is similar to an OOK coding and the modified constellation which represents the given code in Figure three.fifteen(a). Now, if the phase is also used, a modified 3-dimensional (3D) constellation must be used to stand for the code in Effigy 3.15(b). The obtained 3D curve thus represents a unique ID.

Effigy 3.fifteen. Graphical representation of an ID: a) in 2nd, with four resonators, of which only "A" amplitude varies; b) in 3D, with three resonators, of which both amplitude and phase vary

Earlier defining a constellation diagram, nosotros must take into account several practical parameters related to the detection system. The factors limiting the increase in the number of coding states are primarily related to the reading system resolution, and to the noise level generated by the electronics of the reception stage. In concrete terms, a constellation with very next points requires a reading arrangement with a very depression noise floor and a very proficient resolution of the analogue to digital conversion stage. Withal, a constellation with afar points is more robust in detection, simply the coding efficiency of the tag is lower.

Read total chapter

URL:

https://www.sciencedirect.com/scientific discipline/article/pii/B9781785481079500033

Fiber-optic communications

Govind P. Agrawal , in Applications of Nonlinear Cobweb Optics (3rd Edition), 2021

7.5.ane Symbols, baud, and modulation formats

To understand the terminology used for coherent lightwave systems, consider the electric field associated with an optical carrier at frequency ${\omega }_0$ :

(7.v.i) $\mathbf{E}(t)=\widehat{\mathbf{east}}a\cos ({\omega }_{0}t-\varphi )=\mathrm{Re}[\hat{\mathbf{e}}A\exp (-i{\omega }_{0}t)],$

where $\hat{\mathbf{e}}$ is the polarization unit vector, a is the aamplitude, and ϕ is the phase. Introducing the complex phasor as $A=a{\mathrm{east}}^{i\varphi }$ , ane tin construct a constellation diagram in which the existent and imaginary parts of A are plotted along the x and y axes, respectively. In the case of on–off keying, such a diagram has two points forth the existent centrality indicating that but the amplitude a changes from 0 to $a_1$ whenever a chip 1 is transmitted (with no change in the phase). In dissimilarity, if phase takes two values 0 and π while keeping the amplitude fixed, nosotros take a format known as binary stage-shift keying (BPSK). The constellation diagram then has two points along the real axis at positions $a_1$ and $-a_1$ .

The BPSK format does not amend spectral efficiency because it employs only two distinct values of the carrier stage. If the carrier phase is allowed to take four distinct values, say 0, $\pi /2$ , π, and $3\pi /2$ , one tin transmit 2 bits simultaneously in a unmarried symbol slot. Such a format is called quadrature PSK (QPSK), and its constellation diagram is shown in Fig. 7.twentyA. As seen there, one can assign four possible combinations of two bits (00, 01, 10, and 11) to 4 values of the carrier stage in a unique style. As a result, the fleck charge per unit is halved with the use of the QPSK format. This constructive scrap rate is called the symbol rate and is expressed in units of baud. In this terminology, borrowed from radio and microwave communications, stage values represent "symbols" and their number K represents the size of the alphabet. The symbol rate $B_{\mathrm{southward}}$ is related to the bit rate B by the simple relation $B={\log }_{\mathrm{two}}(M)B_{\mathrm{due\; south}}$ . Thus, if the QPSK format with $\mathrm{Thousand}=4$ is employed at $B_{\mathrm{southward}}=40$ Gbaud, information is transmitted at a bit rate of fourscore Gb/s. Of course, the fleck rate can be tripled if one employs viii singled-out values of the phase using the 8-PSK format. Fig. vii.20B shows the assignment of 3 $.25 to each symbol in this example.

Effigy vii.20. Constellation diagrams for (A) QPSK, (B) 8-PSK, and (C) 16-QAM modulation formats showing how multibit combinations are assigned to different symbols.

The fleck rate tin be enhanced further (at the same symbol rate) if the amplitude of the betoken is also varied from one symbol to side by side. An case is shown in Fig. 7.20C, where the aamplitude can take four possible values with four possible phases for each amplitude. This modulation format is known 16-QAM, where QAM stands for quadrature amplitude modulation. It should be emphasized that the assignment of bit combinations to various symbols in Fig. seven.20 is not arbitrary. The coding scheme, known as Grey coding, maps different fleck combinations to different symbols in such a way that merely a single fleck changes between two adjacent symbols separated by the shortest altitude in the constellation diagram. If Grey coding is not employed, a unmarried symbol fault can produce errors in multiple $.25, resulting in an increase in the organisation bit-error rate.

Spectral efficiency can be enhanced further by a factor of ii past exploiting the land of polarization of the optical carrier. This scheme is referred to every bit polarization-segmentation multiplexing (PDM). Information technology may appear surprising that such a scheme tin can work considering the state of polarization of a aqueduct does not remain fixed within an optical fiber but varies in a random manner considering of birefringence fluctuations. Nonetheless, it turns out that PDM tin exist employed successfully every bit long as the ii PDM channels at each wavelength remain orthogonally polarized over the unabridged link length. When coherent detection is employed at the receiver, it is possible to separate the both PDM channels with a suitable polarization-multifariousness scheme. The combination of QPSK and PDM reduces the symbol rate to one quarter of the actual chip rate and thus enhances the spectral efficiency by a gene of 4. Such a dual-polarization QPSK format is often employed in commercial systems because a 100-Gb/s signal can exist transmitted using a symbol rate of only 25 Gbaud.

Read full chapter

URL:

https://www.sciencedirect.com/science/commodity/pii/B9780128170403000134

gardnerorsespach.blogspot.com

Source: https://www.sciencedirect.com/topics/engineering/constellation-diagram