Modem Design, Implementation, and Testing Using NI’s LabVIEW Prof. Brian L. Evans Embedded Signal Processing Laboratory The University of Texas at Austin Contributions by Vishal Monga, Zukang Shen, Ahmet Toker, and Ian Wong, UT Austin

2 Outline  Real-Time DSP Course  Single Carrier Transceiver  Sinusoidal Generation  Digital Filters  Data Scramblers  Pulse Amplitude Modulation  Quadrature Amplitude Modulation  Multicarrier Transceiver  Conclusion

3 Real-Time DSP Course: Overview  Objectives of junior-level class  Build intuition for signal processing concepts  Translate signal processing concepts into real-time digital communications software  Lecture: breadth (three hours/week)  Digital signal processing algorithms  Digital communication systems  Digital signal processor architectures  Laboratory: depth (three hours/week)  Deliver voiceband modem  “Design is the science of tradeoffs” (Yale Patt)  Test/validate implementation Over 500 served since

4 Real-Time DSP Course: Which DSP?  Students are third-year undergraduates  Fixed-point DSPs for high-volume products  Battery-powered: cell phones, digital still cameras …  Wall-powered: ADSL modems, cellular basestations …  Fixed-point issues  Using non-standard C extensions for fractional data  Converting floating-point programs to fixed-point  Manual tracking of binary point prone to error  Floating-point DSPs  Feasibility for fixed-point DSP realization  Shorter prototyping time  Program TI TMS320C67x DSP in C  Code Composer Studio 2.2

5 Real-Time DSP Course: Textbooks  C. R. Johnson, Jr., and W. A. Sethares, Telecommunication Breakdown, Prentice Hall,  Intro to digital communications and transceiver design  Matlab examples  S. A. Tretter, Comm. System Design using DSP Algorithms with Lab Experiments for the TMS320C6701 & TMS320C6711,  Assumes DSP theory and algorithms  Assumes access to C6000 reference manuals  Errata/code: Bill Sethares (Wisconsin) Rick Johnson (Cornell) Steven Tretter (Maryland)

6 Lab 1. QAM Transmitter Diagram Lab 4 Rate Control Lab 6 QAM Encoder Lab 3 Tx Filters Lab 2 Passband Signal LabVIEW demo by Zukang Shen (UT Austin)

7 Lab 1. QAM Transmitter Diagram LabVIEW Control Panel QAM Passband Signal Eye Diagram LabVIEW demo by Zukang Shen (UT Austin)

8 square root raise cosine, roll-off = 0.75, SNR =  raise cosine, roll-off = 1, SNR = 30 dB passband signal for 1200 bps mode passband signal for 2400 bps mode Lab 1. QAM Transmitter Diagram

9 Lab 2. Sine Wave Generation  Aim: Evaluate three ways to generate sine waves  Function call  Lookup table  Difference equation  Three output methods  Polling data transmit register  Software interrupts  Direct memory access (DMA) transfers  Expected outcomes are to understand  Signal quality vs. implementation complexity tradeoff  C6701 EVM board’s stereo codec operation  Interrupt mechanisms and DMA transfers

10 Lab 2. Sine Wave Generation  Evaluation procedure  Validate sine wave frequency on scope, and test for various sampling rates (14 sampling rates on board)  Method 1 with interrupt priorities  Method 1 with different DMA initialization(s) LabVIEW DSP Test Integration Toolkit 2.0 Code Composer Studio 2.2 C6701 Fall 2003 HP 60 MHz Digital Storage Oscilloscope Spring 2004

11 Lab 3. Digital Filters  Aim: Evaluate four ways to implement discrete-time linear time-invariant filters  FIR filter: convolution in C and assembly  IIR Filter: direct form and cascade of biquads, both in C  IIR filter design gotchas: oscillation & instability  In classical designs, poles sensitive to perturbation  Quality factor measures sensitivity of pole pair: Q  [ ½,  ) where Q = ½ dampens and Q =  oscillates  Elliptic analog lowpass IIR filter  p = 0.21 at  p = 20 rad/s and  s = 0.31 at  s = 30 rad/s [Evans 1999] Qpoleszeros ±j ±j ±j ±j classical Qpoleszeros ±j ±j ±j ±j optimized

12 Lab 3. Digital Filters  IIR filter design for implementation  Butterworth/Chebyshev filters special cases of elliptic filters  Minimum order not always most efficient  Filter design gotcha: polynomial inflation  Polynomial deflation (rooting) reliable in floating-point  Polynomial inflation (expansion) may degrade roots  Keep native form computed by filter design algorithm  Expected outcomes are to understand  Speedups from convolution assembly routine vs. C  Quantization effects on filter stability (IIR)  FIR vs. IIR: how to decide which one to use

13 Lab 3. Digital Filters  Test Equipment  Agilent Function Generator  HP 60 MHz Digital Storage Oscilloscope  Spectrum Analyzer  Evaluation Procedure  Sweep filters with sinusoids to construct magnitude and phase responses Manually using test equipment, or Automatically by LabVIEW DSP Test Integration Toolkit  Check filter output for cut-off frequency, roll-off factor…  FIR: Compare execution times (in Code Composer) of C without compiler optimizations C with compiler optimizations C callable assembly language routine  IIR: Compute execution times (in Code Composer)

14 Lab 4. Data Scramblers  Aim: Generate pseudo-random bit sequences  Build data scrambler for given connection polynomial  Descramble data via descrambler  Obtain statistics of scrambled binary sequence  Expected outcomes are to understand  Principles of pseudo-noise (PN) sequence generation  Identify applications in communication systems

15 Lab 4. Data Scramblers  Evaluation procedure  Check scrambler output for various deterministic sequences as input(s)  Descrambler must recover input sequence from scrambled one  Test for sequence period, autocorrelation and other significant statistical properties  Using DSP Test & Measurement Toolkit instead  Compute autocorrelation of PN sequence  Compare this autocorrelation with that of white noise generated by LabVIEW to measue PN sequence quality

16 Lab 5. Digital PAM Transceiver  Aim: Develop PAM transceiver blocks in C  Amplitude mapping to PAM levels  Interpolation filter bank for pulse shaping filter  Clock recovery via phase locked loops g T,0 [n] g T,1 [n] anan D/A Transmit Filter Filter bank implementation 4-PAM d -d -3 d 3 d D/A Transmit Filter anan gT[n]gT[n] L L samples per symbol

17 Lab 5. Digital PAM Transceiver  Expected Outcomes are to understand  Basics of PAM modulation  Zero inter-symbol interference condition  Clock synchronization issues  Test Equipment: Same as Lab 3  Evaluation Procedure  Generate eye diagram to visualize PAM signal quality  Observe spectrum of modulated signal  Prepare DSP modules to test symbol clock frequency recovery subsystem

18 Lab 6. Digital QAM Transmitter  Aim: Develop QAM transmitter blocks in C  Differential encoding of digital data  Constellation mapping to QAM levels  Interpolation filter bank for pulse shaping filter D/A anan gT[m]gT[m] L + cos(w 0 n) bnbn gT[m]gT[m] L sin(w 0 n) Serial/ parallel converter 1 Bit stream Map to 2-D constellation J L samples per symbol

19 Lab 6. Digital QAM Transmitter  Expected outcomes are to understand  In-phase and quadrature modulation principles  Bandwidth efficiency issues  Test equipment: same as Lab 5  Evaluation procedure  Verify differential encoding and QAM mapping  Generate eye diagram to visualize QAM signal quality  Observe spectrum of modulated signal

20 Lab 7. Digital QAM Receiver – Part 1  Aim: Develop QAM receiver blocks in C  Carrier recovery  Coherent demodulation  Decoding of QAM levels to digital data  Expected outcomes are to understand  Carrier detection and phase adjustment  Design of receive filter  Probability of error analysis to evaluate decoder  Test equipment: Same as Lab 6  Evaluation procedure  Recover and display carrier on scope  Regenerate eye diagram and QAM constellation  Observe signal spectra at each decoding stage

21  Multicarrier modulation divides broadband channel into narrowband subchannels  No inter-symbol interference if constant subchannel gain and ideal sampling  Based on fast Fourier transform (FFT)  ADSL/VDSL and IEEE a/g & a subchannel frequency magnitude carrier channel Each ADSL/VDSL subchannel is 4.3 kHz wide (about a voice channel) and carries QAM encoded subsymbol Developed Voiceband Transceiver: Now What? Got Anything Faster?

22 Multicarrier Modulation by IFFT x x x + g(t)g(t) g(t)g(t) g(t)g(t) x x x + Discrete time g(t) : pulse shaping filter X i : i th symbol from encoder I Q

23 Multicarrier Modulation (ADSL) N-point Inverse Fast Fourier Transform (IFFT) X1X1 X2X2 X1*X1* x0x0 x1x1 x2x2 x N-1 X2*X2* X N/2 X N/2-1 * X0X0 N real- valued time samples forms ADSL symbol N/2 subchannels (carriers) QAM Mapping I Q Mirror complex data (in blue) and take conjugates:

24 Multicarrier Modulation (ADSL) CP: Cyclic Prefix N samplesv samples CP s y m b o l ( i ) s y m b o l ( i+1) copy D/A + transmit filter Inverse FFT ADSL frame is an ADSL symbol plus cyclic prefix

25 Multicarrier Demodulation (ADSL) N-point Fast Fourier Transform (FFT) N time samples N/2 subchannels (carriers) S/P

26 P/S QAM decoder invert channel = frequency domain equalizer S/P quadrature amplitude modulation (QAM) encoder mirror data and N -IFFT add cyclic prefix P/S D/A + transmit filter N -FFT and remove mirrored data S/P remove cyclic prefix TRANSMITTER RECEIVER N/2 subchannelsN real samples N/2 subchannels time domain equalizer (FIR filter) receive filter + A/D channel Bits each block programmed in lab and covered in one full lecture each block covered in one full lecture P/S parallel-to-serial S/P serial-to-parallel FFT fast Fourier transform ADSL Transceiver: Data Xmission MHz

27 Modem Design, Implementation, and Testing Using NI’s LabVIEW Dr. Brian L. Evans Associate Professor The University of Texas at Austin Telecom and University Tracks