A Design-to-Test Methodology for SDR and Cognitive Radio Authors: Greg Jue & Bob Cutler, Agilent Technologies
Agenda SDR Waveform Challenges SDR Waveform Design SDR Hardware Testing Cognitive Radio Algorithm Development Challenges Cognitive Radio Testbed 2
Software Defined Radios Flexibility Radio can support multiple waveforms: Different formats, different revisions of a format, backwards compatibility, future-proofing Combination of DSP/FPGA/GPP C++/HDL Flexibility increases demands on RF HW performance HW may be flexible or reconfigurable to more efficiently support waveforms with significantly different characteristics (e.g. OFDM vs MSK) Portability Across single vendors platforms (usually proprietary) Across multiple vendors platforms (based on standards such as SCA) Portability of waveform components (e.g. Viterbi decoder) 3
Portability and Flexibility Challenges and opportunities RF performance determined by both hardware and software. Performance could change with bug fix. Hardware platforms may come from different vendors and have different capabilities. Not quite write-once, run anywhere. Probe points in the signal path are now digital, as well as analog. Need a consistent way to measure. Component implementations in C++, HDL, possibly also from different vendors. Need to design and test hardware to support waveforms that have yet to be invented. Can use test waveforms for development, diagnostics and manufacturing test. 4
Impacts of SDR Technology on Test RF performance is a combination of baseband processing, radio configuration, and RF hardware performance Hard to isolate root causes of performance problems. Finger pointing between hardware and software teams and/or suppliers Need consistent way to quantify RF Performance in the both the hardware and in the soft-bits. Abstraction layers help with portability, but can introduce performance and optimization issues (e.g. timing, or optimal hardware configuration for a waveform) as well as organizational issues (hardware and software from different vendors) Baseband Processing RF 5 5
Example: High TX ACP Several Potential Sources of Error FPGA/DSP D/A PA Insufficient stop band rejection on baseband DSP filter Poor design Truncated filter coefficients Insufficient numeric resolution (adds digital noise) Insufficient numeric range (non-linear over/under flow) Improper timing (PA enable, frequency settling, glitches, etc) Improperly programmed hardware (excess gain, misconfigured parts) Distortion in D/A Poor analog baseband/if filtering IF Distortion due to improper levels Spurious signals Excessive phase noise PA Distortion A 2 db problem is usually a combination of several sub-1db problems. Need to accurately characterize performance in both the digital and analog signal paths 6 6
A Consistent Way To Measure Simulation Design Signal analysis Spectrum analysis Bit-accurate Algorithm Code generation Code Generation FPGA/DSP D/A Digital (logic analyzer) Signal analysis Spectrum analysis Hardware config Timing Packet inspection RTOS scheduling Baseband/IF/RF (scope) Signal analysis Spectrum analysis Hardware config Timing RF/IF (VSA) Signal analysis Spectrum analysis Phase noise Levels Distortion 7 7
Impacts of SDR Technology on Test Future Proofing Need a different test strategy when goal is to support future waveforms Can t just test supported waveforms, need to test radio configurations that might be used with future waveforms When deployed, will Waveform C have adequate RF performance? (ACP, BER, EVM, spurious, noise figure, etc) Waveform A 2009 Waveform B 2009 Waveform C 2013 Baseband Processing FPGA/DSP/GPP Configure Converters and Flexible RF How do we ensure that Waveform C will have adequate performance when installed in radios deployed in 2009? 8 8
SDR Technology Usually Implies Flexible Hardware Signal bandwidth dynamic range Signal BW Signal power stats (Distortion req s) Tunable or broadband antenna for frequency coverage Low phase noise (NB, OFDM) Fast tuning (FHSS, duplex) Phase coherent (MIMO) High efficiency (FM) High linearity(ofdm) Pulsed, continuous TEST APPROACHES 1. Test with real waveforms and their configurations 2. Test broad selection of representative waveforms 3. Test with custom waveforms to exercise different hardware configurations 9 9
Use Simulation to Design Test Waveforms, Introduce Defects, and Verify Test Coverage Noise only DC offset Quadrature error Delay mismatch Distortion 10 10
Agenda SDR Waveform Challenges SDR Waveform Design SDR Hardware Testing Cognitive Radio Algorithm Development Challenges Cognitive Radio Testbed 11
Design SDR RF Using Various Types of Waveform Formats Use waveform sources to design SDR RF Waveform Sources HDL code FPGA hardware Simulation models Algorithm code Waveform signal source Simulated RF transmitter design 12
Example 1: Use HDL-Based WiMAX Waveform to Design SDR RF Transmitter Simulated SDR transmitter output Waveform sources HDL code FPGA hardware Simulation models Algorithm code EVM = 8.4% Mobile WiMAXt m is a registered trademark of the WiMAX Forum. Simulated RF transmitter design VSA measurement 13
Example 2a: Use FPGA-Based Legacy Waveform to Design SDR RF Transmitter Simulated SDR transmitter output Waveform sources HDL code FPGA hardware Simulation models Algorithm code EVM = 9.1% Simulated RF transmitter design VSA measurement 14
Example 2b: Re-Configure FPGA-Based Waveform to Evaluate SDR RF Transmitter Design Interoperability Waveform sources HDL code FPGA hardware Simulation models Algorithm code Reconfigure legacy FPGA waveform for a new waveform (LTE) Simulated SDR transmitter output EVM=10.5% Simulated RF transmitter design VSA measurement 15
Example 2c: Probing an FPGA Waveform with Dynamic Probe Waveform sources HDL code FPGA hardware Simulation models Algorithm code Simulated RF transmitter design Preliminary work-in-progress 16
Example 3a: Use Simulation-Based WiMAX Waveform to Design SDR RF Receiver Waveform sources HDL code FPGA hardware Simulation models Algorithm code Waveform simulation source Waveform simulation receiver Simulated RF receiver design Pre-configured algorithm models (customizable) Select ADC model 17
Example 3a Results: WiMAX BER vs. ADC Jitter QPSK BER vs. ADC Jitter vs. EbNo 16 QAM BER vs. ADC Jitter vs. EbNo 64 QAM BER vs. ADC Jitter vs. EbNo Red: 4% ADC Jitter Blue: 6% ADC Jitter Green: 8% ADC Jitter 18
Example 3b: Replace Waveform to Evaluate SDR Receiver Design Interoperability New BER results New waveform simulation source Replace WiMAX waveform source & receiver with LTE New waveform simulation receiver Simulated RF receiver design 19
Example 4: Use Algorithm Code Waveforms Waveform sources HDL code FPGA hardware Simulation models Algorithm code Customize OFDMA algorithms 20
*Please note: this next section of the presentation contains Preliminary product information that is part of new product development (next 3 slides) 21
SCA Waveform Rapid Prototyping Concept Physical-layer environment for waveform development/verification Waveform Components/ Blocks RF Tx RF Channel/ RF Interferers/Jammers RF Rx Waveform Components/ Blocks Component Model Export Import OE in the loop Functional Component wrapper Deployable component SCA compliant environment For component design, implementation, and deployment Preliminary 22
PHY Link Example 5: SCA Waveform Design Export C++ and XML Preliminary 23
Example 5: Verify End-to-End System with OE In the Simulation Loop PA Output Receiver Digital Output RF Transmitter RF Receiver ADI ADC RF Channel BER Sink QPSK Transmitter (OE-in-the-loop) Preliminary QPSK Receiver (OE-in-the-loop) 24
Agenda SDR Waveform Challenges SDR Waveform Design SDR Hardware Testing Cognitive Radio Algorithm Development Challenges Cognitive Radio Testbed 25
SDR Hardware Testing SDR testing challenges: Custom/proprietary waveforms not supported by COTS test equipment Flexible SDR test platforms are needed for today s and tomorrow s waveforms Different tools used between design and test- makes it difficult to debug issues Solution- Combine the flexibility of simulation with test equipment for flexible SDR testing 26
Adding Flexibility to SDR Testing with Simulation Test waveform coding/decoding SW-defined Customizable algorithms Customizable test waveforms 14 Bit A/D board DUT 16822A Logic Analyzer with Agilent SystemVue* * Note: SystemVue does not ship with logic analyzer 27
OFDMA BER Hardware Test Results 28
Simulate an SDR Receiver with a Hardware Front End (N6841 RF sensor) Wideband RF sensor Simulated RF receiver design Simulated SDR receiver output VSA measurement HW DUT test signal 29
Agenda SDR Waveform Challenges SDR Waveform Design SDR Hardware Testing Cognitive Radio Algorithm Development Challenges Cognitive Radio Testbed 30
Cognitive Radio Many definitions of CR. A radio that is aware of its environment and adjusts its behavior accordingly. Key application for CR is Dynamic Spectrum Access (DSA) Radio adjusts frequency, power, modulation based on sensed spectrum, location, policy and databases Complimentary to SDR in this application 31
Filling the Whitespace Goal: Increase spectrum utilization without causing interference 32
CR Design and Measurement Considerations Interference (actual, or potential for) Radio system performance (capacity, link establishment and reliability) Radio physical layer performance (e.g. adjacent channel power) Environment sensing performance (spectrum sensing, location sensing) Policy performance (does the policy over, or under protect) Radio environment (channel, noise, occupancy) 33
Challenges of Spectrum Sensing Performance of various spectrum sensing algorithms False positives, false negatives Response to real-world signal environment (dynamic, many signals) Radio design Spurious Amplitude accuracy Intermod distortion Sensitivity Selectivity Frequency accuracy Speed/complexity/cost tradeoffs 34
CR Development Needs Need to characterize, capture, and replicate real-world spectral environments. Needs to be done over time, frequency and location. Need to use captured environments to evaluate CR algorithms and radio link performance. Need to evaluate performance using non-ideal radios. Need a flexible and comprehensive CR R&D Testbed! 35
Agenda SDR Waveform Challenges SDR Waveform Design SDR Hardware Testing Cognitive Radio Algorithm Development Challenges Cognitive Radio Testbed 36
Cognitive Radio R&D Testbed 37
CR Algorithm Development and Testing Environment 38
Mobile WiMAX Case Study 39
Step 1: Capture Signal and Bring into SystemVue Captured CR environment 40
Step 2: Whitespace Math Algorithms Determine Valid Whitespace Frequency Rules Policy Valid whitespace determined within the policy Rising/falling edges detected to determine whitespace 41
Debugging Whitespace Algorithms Single-step through code Add/remove breakpoint Code variable values are displayed as code is single-stepped 42
Step 3: Whitespace Math Algorithms Determine Valid Whitespace WiMAX spectrum (scaled and centered in the valid whitespace) 43
Analyze Detect-and-Avoid Interferer Scenarios Narrowband interferer Sweep narrowband interferer vs. frequency to evaluate impact on OFDMA BER 44
Step 4: Identify Detected Signals in Simulation or with Test Equipment Sensed spectrum 45
Video Demo with SystemVue + N6841A N6841A is Remotely Located Across Washington State Remotely located N6841A RF sensor www.agilent.com/find/eesof-cognitive-whitepaper 46
New Whitepaper Available: www.agilent.com/find/eesof-cognitive-whitepaper 47
Summary Use waveforms sources in various formats (HDL, FPGA hardware, simulation models, math algorithms) to design SDR transmitters and receiver and evaluate interoperability Use improved SCA waveform flow for SDR waveform design and test Seamless integration between design and test capability creates flexible SDR testing platform enables R&D engineers to develop and test algorithms and hardware with real field signals Evaluate Cognitive Radio link performance, perform what-if detect-andavoid interference scenarios Explore a Cognitive Radio simulation example in the SystemVue 2009.08 example set request a free evaluation at: www.agilent.com/find/eesof-systemvue-latest-downloads Or, contact your local Agilent representative 48
Additional Resources Product Web sites: SystemVue http://www.agilent.com/find/systemvue RF sensors http://www.agilent.com/find/rfsensor Whitepapers and application notes: Videos: Cognitive Radio Algorithm Development and Testing: http://www.agilent.com/find/eesof-cognitive-whitepaper Software Defined Radio Measurement Solutions: http://cp.literature.agilent.com/litweb/pdf/5990-4146en.pdf Solutions for Addressing SDR Design and Measurement Challenges http://www.agilent.com/find/sdr http://www.agilent.com/find/powerofx Web video of CR Testbed discussed in this paper: http://www.agilent.com/find/eesof-cognitive-whitepaper 49
Thank You! 50
www.agilent.com www.agilent.com/find/eesof-systemvue For more information about Agilent EEsof EDA, visit: www.agilent.com/find/eesof For more information on Agilent Technologies products, applications or services, please contact your local Agilent office. The complete list is available at: www.agilent.com/find/contactus Contact Agilent at: Americas Canada (877) 894-4414 Brazil (11) 4197 3500 Mexico 01800 5064 800 United States (800) 829-4444 Asia Pacific Australia 1 800 629 485 China 800 810 0189 Hong Kong 800 938 693 India 1 800 112 929 Japan 0120 (421) 345 Korea 080 769 0800 Malaysia 1 800 888 848 Singapore 1 800 375 8100 Taiwan 0800 047 866 Thailand 1 800 226 008 Europe & Middle East Austria 01 36027 71571 Belgium 32 (0) 2 404 93 40 Denmark 45 70 13 1515 Finland 358 (0) 10 855 2100 France 0825 010 700* *0.125 /minute Germany 07031 464 6333 Ireland 1890 924 204 Israel 972-3-9288-504/544 Italy 39 02 92 60 8484 Netherlands 31 (0) 20 547 2111 Spain 34 (91) 631 3300 Sweden 0200-88 22 55 Switzerland 0800 80 53 53 United Kingdom 44 (0) 118 9276201 Other European Countries: www.agilent.com/find/contactus Product specifications and descriptions in this document subject to change without notice. Agilent Technologies, Inc. 2010 Printed in USA, October 29, 2010 5990-6694EN