Exercise 3 Differential QAM (DQAM) EXERCISE OBJECTIVE When you have completed this exercise, you will be familiar with the use of differential encoding, using the ITU-T V.22 bis recommendation, to overcome phase ambiguity, and with its implementation in QAM modems. DISCUSSION OUTLINE The Discussion of this exercise covers the following points: Review of phase ambiguity Using differential encoding to overcome phase ambiguity DQAM using v.22 bis Advantages and disadvantages of differential encoding Differential encoding in the QAM/DQAM application DISCUSSION Review of phase ambiguity All modulation techniques that use phase to encode information, including QAM, suffer from phase ambiguity. With QAM, the data is conveyed by both the amplitude and the phase of the modulated signal. To detect the phase of the received signal, the demodulator compares its phase with a stable, recovered carrier that is generated by a carrier recovery circuit in the demodulator. Since all available phases are usually present in the received signal, the carrier recovery circuit can lock onto any one of these phases to regenerate the carrier, even though only one of them allows correct demodulation of the data. This phenomenon is known as phase ambiguity. Using differential encoding to overcome phase ambiguity Instead of using the data to set the phase at the modulator, the data can be used to change the phase by a specific amount. The demodulator then detects the changes in phase, rather than the absolute phase. Since this technique depends on the difference between successive phases, it is called differential QAM (DQAM). With DQAM, differential encoding is only used for the phase of the modulated signal; absolute (non-differential) encoding is used for the amplitude because there is no ambiguity in detecting the amplitude at the demodulator. Although the demodulator in a DQAM system arbitrarily locks onto one of the phases in the received signal, this phase ambiguity is unimportant since only the phase differences are important. Festo Didactic 39866-00 57
Exercise 3 Differential QAM (DQAM) Discussion DQAM using v.22 bis The ITU (International Telecommunication Union) is an international organization established to standardize and regulate international radio and telecommunications. Founded in Paris in 1865 as the International Telegraph Union, it is the world's oldest international organization. ITU-T refers to the telecommunication standardization sector of the organization. Quadrants are numbered as follows: 2 1 3 4 With 16-DQAM, there are sixteen symbols, or quadbits. The constellation has sixteen points (four in each quadrant). According to the ITU-T recommendation V.22 bis, the first dibit (two bits) of each quadbit is encoded as a phase quadrant change relative to the quadrant occupied by the preceding signal element (constellation point). The second dibit defines one of 4 signaling elements associated with the new quadrant. Since the first dibit in each quadbit changes the phase quadrant of the QAM signal, we will refer to this dibit as the phase dibit. Bits one and two of the second dibit determine the relative amplitude of the I-channel and Q-channel contribution to the QAM signal. The second dibit will therefore be referred to as the amplitude dibit. The colors red-orange and blue-green are used in this exercise to differentiate the two dibits. Table 7 shows the V.22 bis mapping of phase dibit to phase change. This table also shows also shows all possible changes in quadrant for each phase change. Table 7. Differential phase mapping. Phase dibit Phase Change Quadrant Changes 00 90 01 0 11 270 10 180 1 2 2 3 3 4 4 1 1 1 2 2 3 3 4 4 1 4 2 1 3 2 4 3 1 3 2 4 3 1 4 2 Figure 44 shows the mapping of the amplitude dibits to the four constellation points in the first quadrant. Since the amplitude of the mixer output signal in each channel of the modulator is proportional to 1 or 3, these two bits determine the relative amplitude contribution of the I and Q channel carriers to the QAM signal. 58 Festo Didactic 39866-00
Exercise 3 Differential QAM (DQAM) Discussion Q + 3 10 11 I+3Q 3I+3Q + 1 00 01 I+Q 3I+Q I + 1 + 3 Figure 44. Absolute (non-differential) mapping of dibits to amplitudes in the first quadrant. Figure 45 shows the complete signal constellation defined by the V.22 bis recommendation. The phase dibit determines the changes in phase quadrant. The amplitude dibit determines which of the four constellation points of the new quadrant is used. Note that in each quadrant, 00 is mapped to the point nearest the origin and 11 is mapped to the point furthest from the origin. The pattern in each of quadrants 2, 3, and 4 is obtained by rotating the pattern in quadrant 1 about the origin by 90, 180, and 270, respectively. Festo Didactic 39866-00 59
Exercise 3 Differential QAM (DQAM) Discussion The first dibit of each quadbit is encoded as a phase quadrant change relative to the quadrant occupied by the preceding signal element (constellation point). Phase Quadrant 2 Q Phase Quadrant 1 The second dibit of each quadbit defines one of 4 signaling elements associated with the new quadrant. 11 01 +3 10 11 10 00 +1 00 01 I -3-1 +1 +3 01 00-1 00 10 11-3 10 01 11 Phase Quadrant 3 Phase Quadrant 4 Figure 45. V.22 bis signal constellation. To illustrate the advantage of using differential encoding for the phase, consider the series of data bits 0000 1100 0100. In this example, only the phase dibits (in red) change from one quadbit to the next. Since the amplitude dibits (in aqua) are all 00, the constellation point nearest the origin will be used in each quadrant. When absolute (non-differential) coding is used, if the demodulated phase is incorrect due to phase ambiguity, the demodulated constellation is always rotated with respect to the transmitted constellation. In this case, all decoded quadbits are erroneous. Table 8 shows the effect of differential encoding (using the encoding of Table 7). In this table, the initial phase is assumed to be in the first quadrant. The phase dibit 00 in the first quadbit of our sequence produces a phase change of 90, shifting the transmitted constellation point into the second quadrant. The next phase dibit 11 produces a phase change of 270, shifting the constellation point back to the first quadrant. Because of phase ambiguity, the demodulated constellation is always rotated by the same amount with respect to the transmitted phase. However, since the information is contained in the phase changes, rather than in the absolute phase, this does not cause errors in the decoded data. 60 Festo Didactic 39866-00
Exercise 3 Differential QAM (DQAM) Discussion Table 8. Differential encoding example with 90 demodulation phase error. Transmitted phase dibit Phase Change Transmitted Phase Demodulated Phase Detected Phase change Decoded phase dibit (initial phase) 00 90 90 00 11 270 270 11 01 0 0 01 Advantages and disadvantages of differential encoding With differential encoding, the data is encoded using changes in phase, rather than using absolute phases. Phase ambiguity therefore does not affect the decoded data. Immunity to phase ambiguity is the advantage offered by differential encoding. Differential encoding, however, requires slightly more complex circuitry in the modulator and demodulator. There is one other disadvantage in using differential encoding. Since the phase changes convey the data, an error in detecting the phase of one modulation symbol produces errors in two successive data symbols. This increase in the number of errors can often be compensated for, however, by slightly increasing the transmitted power in order to reduce the probability of error due to channel noise. Differential encoding in the QAM/DQAM application In the LVCT QAM/DQAM application, the I- and Q-channels of the modulator (the D/A Converters and mixers) use a fixed mapping between the quadbits they receive and the resulting constellation points. In Step 26 of Exercise 1, you observed this mapping. In order to implement differential encoding, this mapping is not changed. Instead, the Differential Encoder in the modulator alters the quadbits it receives from the Serial to Parallel Converter in order to generate the required phase changes. It does this by comparing each input quadbit it receives with the previous quadbit, which was stored in a register, and calculates a new quadbit that will produce the required phase change, using the mapping shown in Table 7. Festo Didactic 39866-00 61
Outline The Differential Decoder in the demodulator reverses the process in order to recover the original data. PROCEDURE OUTLINE This Procedure is divided into the following sections: Set-up and connections Differential (V.22 bis) encoding in the modulator The effect of differential encoding on demodulation PROCEDURE Set-up and connections 1. Turn on the RTM Power Supply and the RTM and make sure the RTM power LED is lit. File Restore Default Settings returns all settings to their default values, but does not deactivate activated faults. Double-click to select SWapp 2. Start the LVCT software. In the Application Selection box, choose QAM/DQAM and click OK. This begins a new session with all settings set to their default values and with all faults deactivated. b If the software is already running, choose Exit in the File menu and restart LVCT to begin a new session with all faults deactivated. 3. Make the Default external connections shown on the System Diagram tab of the software. For details of connections to the Reconfigurable Training Module, refer to the RTM Connections tab of the software. b Click the Default button to show the required external connections. 4. As an option, connect a conventional oscilloscope to the QAM Demodulator I-CHANNEL OUTPUT and QAM Demodulator Q-CHANNEL OUTPUT (refer to the RTM Connections tab of the software). Use the conventional oscilloscope in the X-Y mode to observe the constellation. (The Low-Pass filters in the QAM Modulator must be set to On.) Differential (V.22 bis) encoding in the modulator 5. Connect the Oscilloscope probes to the QAM Modulator as follows: Oscilloscope Probe Connect to Signal 1 TP14 I-channel D/A Converter output 2 TP15 Q-channel D/A Converter output 62 Festo Didactic 39866-00
Connect the Logic Analyzer probes to the QAM Modulator as follows: Logic Analyzer Probe Connect to Signal C TP2 CLOCK INPUT E TP5 Frequency Divider output 1 TP6 Serial to Parallel Converter output (MSB) 2 TP7 Serial to Parallel Converter output 3 TP8 Serial to Parallel Converter output 4 TP9 Serial to Parallel Converter output (LSB) 5 TP10 I-channel D/A Converter input (MSB) 6 TP11 I-channel D/A Converter input (LSB) 7 TP13 Q-channel D/A Converter input (MSB) 8 TP12 Q-channel D/A Converter input (LSB) a Make sure you connect probe 7 to TP13 and probe 8 to TP12. 6. Make the following settings: Generator Settings: Generation Mode... User Entry Binary Sequence... 0000 QAM Settings: Differential Encoding... Off (QAM) Scrambler... Off Make sure to set the Input of both Channel 1 (X) and Channel 2 (Y) to On. Oscilloscope: X-Y... On Display Mode... Dots Sampling Window... 200 ms Channel 1 (X)... 1 V/div Channel 2 (Y)... 1 V/div Logic Analyzer: S1 Data... [ch1, ch2, ch3, ch4] S2 Data... [ch5, ch7] Observe the signals on the Logic Analyzer and on the Oscilloscope. The Logic Analyzer should show that all signals are in a zero state. Which constellation points(s) are displayed on the Oscilloscope? Festo Didactic 39866-00 63
7. Turn the V.22 bis Encoder On, or set Differential Encoding to On (DQAM). Which constellation points are now displayed on the Oscilloscope (see Figure 46)? Explain why. Figure 46. Constellation points for the Binary Sequence 0000. Generator Settings: Generation ModeUser Entry... Binary Binary Sequence... 0000 QAM Settings: Differential Encoding... On (DQAM) Logic Analyzer Settings: Display Width... 20 ms Clock Grid... Rising Edge Source... Ext Source Edge... Rising Clock Edge... Rising S1 Data... [ch1, ch2, ch3, ch4] S2 Data... [ch5, ch7] 8. Record data on the Logic Analyzer. Figure 47 shows an example of what you may observe (your results may be slightly different). Figure 47. V.22 bis Encoder inputs (Ch 1-4) and outputs (Ch 5-8). The four Serial to Parallel outputs are displayed in Ch 1 (MSB) to Ch 4 (LSB) of the Logic Analyzer. These four channels are combined into Symbol 1. These are all in the zero state because each quadbit is 0000. 64 Festo Didactic 39866-00
The MSB input of each D/A Converter controls the phase of the modulated signal in the same channel, and the LSB input controls the amplitude. The two phase inputs are shown in Ch 5 and Ch 7 of the Logic Analyzer and the two amplitude inputs are shown in Ch 6 and Ch 8. Since Symbol 2 on the Logic Analyzer is configured as a combination of Ch 5 and Ch 7, this Symbol displays useful phase information, as shown in Figure 48. Phase Quadrant 2 1 0 [2] Ch 7 (TP13) 0 0 0 [0] Phase Quadrant 1 1 0 Ch 5 (TP10) 1 1 [3] 1 0 1 [1] Phase Quadrant 3 Phase Quadrant 4 Figure 48. Ch5 and Ch7 values, the corresponding Logic Analyzer Symbols (in square brackets), and phase quadrants. V.22 bis encodes the phase dibits of each quadbit as a change in phase quadrant. The pattern in Symbol 2 of the Logic Analyzer shows exactly how the phase changes from one symbol interval to the next. Referring to Figure 48 and symbol channel S2 on the Logic Analyzer, describe how the phase is changing and explain why. Festo Didactic 39866-00 65
9. In Figure 49, identify the 00 amplitude dibit in each quadrant. Phase Quadrant 2 Phase Quadrant 1 + 3 + 1-3 - 1 + 1 + 3-1 - 3 Phase Quadrant 3 Phase Quadrant 4 Figure 49. V.22 bis signal constellation. 10. Set the Binary Sequence to 0001. Record data on the Logic Analyzer and make sure the quadbits at the output of the Serial to Parallel Converter are all 0001. (Channels Ch1 to Ch4 and symbol S1 should show 0001 and [1] respectively. If this is not the case, click the Drop 1 Bit button and record data on the Logic Analyzer again until they do. Figure 50 shows an example.) Figure 50. V.22 bis Encoder inputs (Ch 1-4) and outputs (Ch 5-8). 66 Festo Didactic 39866-00
Which constellation points are now displayed on the Oscilloscope (see Figure 51)? Explain why. Figure 51. Constellation points for the Binary Sequence 0001. In Figure 49, identify the 01 amplitude dibit in each quadrant. 11. Use the Binary Sequences 0010 and 0011 to identify all remaining dibits in Figure 49 (verify the quadbits with the Logic Analyzer and use the Drop 1 Bit button as necessary). Do your results agree with Figure 45? 12. Change the Binary Sequence to any four-bit sequence beginning with 01. Then use the Logic Analyzer and the Drop 1 Bit button to ensure that the phase dibit of each quadbit is 01? How many constellation points are displayed on the Oscilloscope? Explain. Festo Didactic 39866-00 67
13. Set the Binary Sequence to different sequences beginning with 10 and with 11. Use the same method to analyze the results. What is required to produce a 270 phase change in each symbol interval? 14. Experiment with V.22 bis encoding using longer binary sequences. Find sequences of four quadbits (at the output of the Serial to Parallel Converter) that: a. display four constellation points in one quadrant only b. display four constellation points in two quadrants only (as in Figure 52) c. display all 16 constellation points Figure 52. Four constellation points in two quadrants. The effect of differential encoding on demodulation 15. Make the following settings: QAM Settings: Differential Encoding... Off (QAM) 68 Festo Didactic 39866-00
Generator Settings: Binary Sequence... 0011 0101 1110 0111 Delay... Use the value recorded in Step 17 of Exercise 2. Logic Analyzer: S1 Data... [ch1, ch2, ch3, ch4] S2 Data... [ch1, ch3] 16. Disconnect all Logic analyzer probes in the QAM Modulator. Then connect the Logic Analyzer probes to the QAM Demodulator as follows: b To rapidly disconnect all Logic Analyzer probes, right-click in the diagram and select Disconnect All Logic Analyzer Probes in the context-sensitive menu. Logic Analyzer Probe Connect to Signal C TP22 CLOCK OUTPUT E TP24 Frequency Divider output 1 TP16 V.22 bis Decoder output (MSB) 2 TP17 V.22 bis Decoder output 3 TP18 V.22 bis Decoder output 4 TP19 V.22 bis Decoder output (LSB) 5 TP21 DATA OUTPUT 6 TP25 BSG DELAYED DATA OUTPUT Connect the Oscilloscope probes to the QAM Demodulator as follows: Oscilloscope Probe Connect to Signal 1 TP14 I-channel Decision Circuit output 2 TP15 Q-channel Decision Circuit output 17. Record data on the Logic Analyzer and compare the traces in Ch 5 and Ch 6. If these are not identical, click the Shift 90 button in the demodulator, and record data, until they are. In the modulator, click the Drop 1 Bit button, if necessary, until the displayed quadbits correspond to the data sequence (see Figure 53). S1 should display [ 3 ] [ 5 ] [ E ] [ 7 ] (the hex values of the quadbits 0011 0101 1110 0111). Festo Didactic 39866-00 69
Logic Analyzer Settings: Display Width... 20 ms Clock Grid... Rising Edge Source... Ext Source Edge... Rising Clock Edge... Rising S1 Data... [ch1, ch2, ch3, ch4] S2 Data... [ch1, ch3] Figure 53. No phase error. Use the Oscilloscope in the X-Y mode to observe the constellation (see Figure 54). Oscilloscope Settings: Channel 1... 1 V/div Channel 2... 1 V/div Display Mode... Dots X-Y... On Sampling Window... 50 ms Figure 54. Four-point constellation with absolute encoding. In the demodulator, click the Shift 90 button. What happens to the constellation? Observe the data on the logic Analyzer. Is the data is correctly decoded? 70 Festo Didactic 39866-00
Click the Shift 90 button data several times, observing the constellation and the data on the Logic Analyzer. With absolute encoding, what is required for the data to be correctly decoded? Oscilloscope Settings: Channel 1 (X)... 1 V/div Channel 2 (Y)... 1 V/div Display Mode... Dots X-Y... On Sampling Window... 200 ms 18. Turn Differential Encoding on and observe the constellation and the data on the Logic Analyzer. If necessary, click the Drop 1 Bit button in the modulator until you obtain a constellation with four points in a straight line, as shown in Figure 55. (The line may be oriented in any direction.) Figure 55. Four-point constellation. 19. Click the Shift 90 button in the demodulator Phase Tracker several times and describe what happens to the constellation in the receiver. For each different constellation, record data on the Logic Analyzer and compare the demodulated data (Ch 4) with the original data (Ch 5). What is the effect of using differential encoding? 20. When you have finished using the system, exit the LVCT software and turn off the equipment. Festo Didactic 39866-00 71
Exercise 3 Differential QAM (DQAM) Conclusion CONCLUSION In this exercise, you observed how differential encoding is implemented in a QAM system. You saw that the quadbits are modified by the Differential Encoder so that part of the data is conveyed by changes in phase, rather than by absolute phases. You observed that, once the coded quadbits are decoded in the demodulator, phase ambiguity of the recovered carriers does not affect the recovered data. REVIEW QUESTIONS 1. How is data represented in the modulated waveform using V.22 bis encoding? 2. Does phase ambiguity exist in a DQAM system? If so, what effect does it have on the modulation process? 3. Why is absolute (non-differential) encoding used to set the amplitude of the QAM signal? 4. What is the main disadvantage of using differential encoding and how can this be overcome? 72 Festo Didactic 39866-00