Real-Time Spectrum Analysis (RTSA) -Triggering, and Signal Capture/Playback for Agile and Elusive Signals. Keysight Technologies

Transcription

1 Real-Time Spectrum Analysis (RTSA) -Triggering, and Signal Capture/Playback for Agile and Elusive Signals Keysight Technologies

2 A brief history of Keysight Technologies : Hewlett-Packard years A company founded on electronic measurement innovation : Agilent Technologies years Spun off from HP, Agilent became the World s Premier Measurement Company. In September 2013, it announced the spinoff of its electronic measurement business 2014: Keysight begins operations November 1, Keysight is an independent company focused 100% on the electronic measurement industry We believe in Firsts Bill Hewlett and Dave Packard s original vision, which launched Silicon Valley, shaped our passion for firsts 75 years ago. Today we are committed to provide a new generation of firsts software-oriented solutions that create value for Keysight investors and valued insights for our customers. Page Keysight Technologies

3 What is Real-Time Spectrum Analysis? General Definition of Real-Time Measurement Operations where all signal samples are used in calculating measurement results of some kind (usually spectrum) Real-Time Bandwidth (RTBW) The widest analysis bandwidth where an analyzer can maintain real time operation Duration of maintaining real time operation is not specified; it may be assumed to be short term or long term or unlimited Current usage for Signal Analyzers A spectrum or FFT analyzer having a signal processing path where most or all of the samples, even at wide bandwidths, are used to create a spectral display or to trigger signal measurement or acquisition (sometimes both) Page 3

4 Capture and Analyze Highly Elusive Signals Real-time Spectrum Analysis From this...to this (demo) Know you ve got it Page 4

5 Basic Real-Time Displays Density Spectrum Spectrogram Power vs Time Also know as Histogram Persistence Color indicates number of hits Screen typically updates every 30ms Persistence can be manual or infinite Accumulate all FFT s to a single trace Apply detector Superimposed on the density display Used for marker operations Real Time Spectrum slices no gaps 10,000 spectrogram traces available Scroll through stored traces Use markers on and between traces PvT over configurable range Gapless time data transformed to frequency domain Different displays available Level based trigger available Page 5

6 Frequency Mask Trigger (FMT) Build Mask from trace and add offsets if required Edit table or use mouse to drag the mask points to the desired location Various criteria for Trigger: Enter, Leave, Inside, Outside, Enter Leave, Leave Enter FMT (demo) Upper, Lower or Both masks available Import or Export masks as required FMT Combined with 89600B VSA software for further analysis Page

7 Analyze Complex Signals with VSA Software Page 7

8 Dynamic Range Lowest noise floor -157 dbm at 10 GHz (radar & EW) Best SFDR across an entire 160 MHz (-75 dbc) PXA s dynamic range with low-noise path (LNP) provides the lowest noise floor, delivering significantly better performance than other RTSAs Page

9 Case Study #1: 2.45 GHz ISM Band 100 MHz Wide Spread Spectrum Techniques Lightly Regulated, Signals Not Explicitly Coordinated WLAN (typically dominant occupant), Bluetooth, Cordless Phones And 1 kw+ microwave ovens! And almost anything else! Very Dynamic Content not known at any specific point in time Often very low occupancy Lots of Collision Tolerance Needed Potential Cliff Effect at Some Degree of Crowding Page 9

10 ISM Band View: Swept Spectrum Analysis Single Sweep (fast but retrace not as fast) Peak Hold (short) Band: 100 MHz Span: 160 MHz Peak Hold (long) Dynamics X2 (swept RBW & changing signals) Confuse the View Peak Hold Detects Signals but Frequency Overlap Obscures Signals Typically Short Duty Cycle, Total Low Occupancy POI is Approximately Low (useful concept even if not strictly applicable) Page 10

11 ISM Band View: Real-Time Analyzer Density or Histogram Display ISM (demo) Gap-Free Analysis No Signals Missed Density Display Shows Signals Inside Signals Fast, Full Display Update Rate Shows Signal Dynamics See Unexpected Behavior, Even If Infrequent Individual Display Updates Combine Thousands of Spectra Page 11

12 ISM Band View: Real-Time Analyzer Spectrogram Display Top Trace is Latest or Selected Spectrum Same Spectral Data as Density/Histogram Individual Spectrogram Line Represents Many Spectra (typically thousands) View into Dynamics, Timing Acquisition Time Setting Affects Way Spectra are Combined to Form Display Update WLAN Signals Active Continuously? What are Vertical Bars? A Multitone Signal? Page 12

13 Spectrogram Display Adjusting Acquisition Time Shorter Acquisition Time Determines Amount of Data Processed for Spectrum or Spectrogram Trace Update Each Trace Update and Spectrogram Line Represents Fewer Spectra More Trace Updates/Second, Finer Time Resolution Spectrogram Covers Shorter Interval Determines Time Scale of Spectrogram Use Spectrogram Trace Buffer to Examine Signal Dynamics Page 13

14 The Swept Analysis Mode A swept LO w/ an assigned RBW. Covers much wider span. Good for events that are stable in the freq domain. Magnitude ONLY, no phase information (scalar info). Captures only events that occur at right time and right frequency point. Data (info) loss when LO is not there. Lost Information Lost Information Lost Information Swept LO Freq Time Page 14

15 IQ Analyzer (Basic) Mode Complex Spectrum and Waveform Measurements A parked LO w/ a given IF BW Parked LO Collects IQ data over an interval of time. Performs FFT for timefreq-domain conversion Captures both magnitude and phase information (vector info). Data is collected in bursts with data loss between acquisitions. Lost Information Analysis BW Freq Meas Time or FFT Window Length Meas Time or FFT Window Length Time Page 15

16 Real Time Spectrum Analysis A parked LO w/ a given IF BW Collects IQ data over an interval of time. Data is corrected and FFT d in parallel Vector information is lost Advanced displays for large amounts of FFT s Parked LO Freq Acquisition or slice time Acquisition or slice time Real-time BW Time Page 16

17 Real-Time Operation In Real-Time Operation the Analyzer s Processing (CALC) is Fast Enough to Keep Up with All Data Samples However some data may still be lost CALC Time Includes FFT or Power Spectrum, Averaging, Display Updates, etc. T =time record for FFT processing CALC = time for measurement calculations Page 17

18 The FFT At first glance Window Window Samples Page 18

19 Overlap Processing If Processing is Faster than Sampling, Perform Additional FFTs With Partially-New Time Records as Samples Come In Overlap 0% Overlap 50% Processor Idle Window Function Avoid Loss of Data Due to Windowing Accurate Amplitude Measurements of Short Duration Signals Page 19

20 What is a Real-Time Analyzer? Specific Analyzer Type vs. Analysis Characteristic In this context it means several or all of the following things: Gap free: no dead time between acquisitions; all sampled data is processed; process is continuous Consistent measurement & display update speed: all hardware implementation of FFT spectrum generation and not subject to Windows task interruptions High speed measurements: many thousands of FFT generated spectra per second, vastly exceeding software FFT speed High speed and/or data-dense displays: combine large numbers of measurements to form responsive, insight-producing displays Spectral trigger or frequency mask trigger (FMT) Definitions vary, what is real-time analysis for you? Page 20

21 Probability of Intercept (POI) Treated Informally Here Finding signals important to you, expected or unexpected Reasonable effort and time to discovery Signals found can be measured as necessary Knowledge of signals may be perfect or simply adequate Ability to find elusive signals and signal behavior quickly and reliably, whether the goal is to discover or characterize the signals, or to understand detail as an element of troubleshooting or optimization Evaluate These Techniques Against Your Requirements Page 21

22 Frequency Mask Trigger (FMT) Trigger Based on In-Band Power (scalar) Spectrum Spectral Mask Evaluation, High/Low Limits, Logic Trigger Generated from High Speed FFT Engine Trigger Timing Resolution: 1/FFT Period Trigger Not Time-Aligned Trigger a Single Spectrum Measurement or Time Capture (VSA), etc. Page 22

23 Vector Signal Analysis: The Logical Extension and Complement to a Real-Time Analyzer Superheterodyne Downconversion to All-Digital IF Adaptable to swept spectrum analysis and real-time analyzer Full vector data preserved Measurements Short/Long Data Block Vector, Demodulation Capture Memory Page 23

24 VSA Spectrum of ISM Band Signals Initial Measurement Not Much More Informative FFT Avoids Problem of Moving Analysis Aperture (RBW filter) However the default data block or time record for most analyzers is fairly short Page 24

25 VSA Spectrum of ISM Band Signals Long Time Record (100,000 points) Spectrum Much More Informative Higher POI Magnitude or RF Envelope Color Persistence Spectrum Page 25

26 VSA Spectrum of ISM Band Signals Long Time Record & Enhanced Displays Not Real Time with Live Measurements Cumulative History Is Real Time Plus Overlap with Time Capture and Playback Separating History (frequency of occurrence) and Recency Digital Persistence Select Segment/Region to Analyze from Time Capture buffer Many Spectra Combined Page 26

27 Improving POI with FFT Analyzers Several Steps for Dramatic Improvement Use Much Longer Time Records (25,000 points or more) Time record length goes up faster than FFT speed goes down Narrow Frequency Span to the Minimum Needed Time record is longer (in time) for given number of points Use IF Magnitude Trigger, Pre/Post Delays, Holdoff Avoid measuring when no signal present in span Use Persistence or Density/History Displays Typically persistence, to make peaks last long enough to be noticed Peak Hold Result: POI Dramatically Improved, Though may not be Near 100% Page 27

28 Triggering in VSAs and Real-Time Analyzers IF Magnitude & Frequency Mask Triggers Triggered Individual Measurements or Signal Captures Sqrt (I 2 + Q 2 ) of each sample, is easier & faster (VSAs) FFT power Spectrum is slower but more selective (RTSAs) VSA now does both IF Mag or Freq Mask Trig Measurements Short/Long Data Block Vector, Demodulation Capture Memory Page 28

29 VSA Attributes for Detailed, Flexible Analysis The Next Steps After Finding a Signal Detailed Signal Analysis, Including Vector, Demodulation Time Capture, Playback Connection to Many Signal Analzyer Front-Ends Signal analyzers Oscilloscopes Digitizers Multi-Channel Analysis, Including Channel Time Alignment Including multi-channel capture & analysis Beam-forming, MIMO, space-time coding Multi-Band Analysis, Multi-Measurements Excellent Complement to Single-Channel RTSA Page 29

30 Keysight 89600B VSA Signal Capture: Just Press the Red Button Use Existing Center, Span (Wider is ok) Press Record Then Press Play Adjust Capture if Desired Length (seconds, samples, time records) or use default Input Range Frequency Center, Span Better, Faster Insights from Measuring Same Signal with Different Settings Page 30

31 Post-Processing: Play, Pause Restart Simultaneous Spectrum and Time (RF Envelope, IF Time) is a Good Start Add Traces Correlate Time Behavior Change Time Scale Spectrum Spectrum Change Anything Else Pause, Loop or Restart RF Envelope IF Time IF Time Page 31

32 Navigating Time Capture in the VSA The Window Controls You Already Know Start Time Analysis Position Drag Analysis Position & See Results Instantly Stop Time Sweeps Continuous Single-Step Click, Drag Numeric, Graphic Entry Page 32

33 Gap-Free, Overlapped Spectrogram: Analysis of Amplitude, Frequency, Time, See Everything At Once See Entire Event at a Glance Find Unexpected Behavior Select Any Trace From Deep Buffer and Measure Spectrogram produced from capture/playback Page 33

34 Capture and Spectrogram Show Environment Understand Frequency, Time, Power, Modulation Relationships Page 34

35 Another ISM Band Capture WLAN, Bluetooth, Microwave Ovens 16.7 ms Page 35

36 Flexible Triggering to Initiate Capture IF Magnitude Trigger Frequency Mask Trigger Pre/Post-Trigger Delay Adjustable Power Level ± Delay Holdoff time, type Other Types Channel External/TTL Periodic Magnitude Freq. Mask Page 36

37 RTSA Frequency Mask Trigger in 89601B VSA Setup window (free run / RT) Setup window (Similar to RTSA) Control free/trig Acq IQ time RT/FMT engine mask data RT spect. RT histo. RT spectro. Time, spectrum demod, etc. RT display (triggered) VSA Time Inp. capt. PXA VSA Page 37

38 Capture Wide, Analyze Any Freq. Segment Post-Capture Span, Center Frequency Changes Apply to All Modes including Analog & Digital Demodulation Page 38

39 Focus on a Time and Frequency Region Marker CF, reduce span Play back until signal appears Click/drag analysis region, position Page 39

40 Span, Center Freq Change to Analyze, Demodulate 2.5 MHz span, Bluetooth demod preset, short result length Page 40

41 Save a Time/Frequency Portion of Capture Use Capture, Triggering to find What You Want Select the Signal or Time Region Desired Change Center Frequency and/or Span Save Segment for Later Analysis, Playback, Re-Use Page 41

42 Pass Captured Data to Other Tools, Processes Captured Data is Fully Time and Frequency Corrected Data is Fully Alias Protected Save as MATLAB (several formats), CSV, Text Send as ARB Register Data RF Out Page 42

43 Start With Problem/Need, Choose Best Tool Swept Spectrum Analyzer (especially with digital IF) Familiar meas, user interface; compatible with established std s, practices Frequency flexibility including span, RBW, VBW Maximum accuracy, dynamic range, sensitivity Measurement flexibility: phase noise, noise figure, ACPR, EMI, apps, etc. Real-Time Analyzer Find elusive signals, some ability to characterize (spectrogram, PVT) Spot unknown signals, signal behavior Monitor spectrum, trigger on spectral behavior Vector Signal Analyzer (VSA software on signal analyzer platform) Detailed vector analysis and demodulation Signal capture, flexible playback Post-processing for signal selection, re-use Single user interface for multiple hardware platforms, multi-channel analysis Modern Signal Analyzer Hardware May Support All Three Page 43

44 Case Study #2: S-Band Acquisition Radar Raster Scanning Pulse Width 6 µs PRI 600 µs 7 Pulses, 10 MHz Spacing, Stepping Low - High Frequency Pulses -30 MHz to +30 MHz Page 44

45 Case Study #2: S-Band Acquisition Radar Fast, Clear Signal View Blue Indicates Very Low Duty Cycle Amplitude Varies over Seconds Set Persistence Long Peaks Consistent Page 45

46 S-Band Acquisition Radar Spectrum Analyzer View Peak Hold Many Measurements Required Signal Still Not Clear Dynamics Not Shown Page 46

47 S-Band Acquisition Radar Real-Time Spectrogram View Long Acquisition Time, Long Persistence Excellent for Long- Term Signal View (seconds) Spot a Pattern Spot Big Pulse Set Many Spectra (default 10,000) Combined, Pulses Still Shown Together Page 47

48 Frequency Mask Trigger, Radar Signal Spectrum & Level Trigger Trigger Acquisition Only on High Peak Can Trigger on Defective Pulses, etc. Peak (Blue) vs. Combined (white) Single Display Update Still Made From Many Spectra But Trigger From Single One Page 48

49 How to Identify the Signal Immediately? Real-time Trigger Real-time trigger used for pulse signal identification Time Qualified Triggering(TQT) FMT Amplitude Amplitude Amplitude Frequency 5 GHz band FMT does not work if equal amplitude signals overlap in the frequency domain Frequency Time BUT, overlapping signals in the frequency domain can be resolved by time domain trigger. Use case: To trigger on a pulsed signal in presence of other similar signals that lasts for either longer or shorter durations Page 49

50 VSA Post-Capture Tune/Zoom Digital Resampling, Digital Local Oscillator Focus on Signal of Interest, Filter Out Other Signals Select Analysis Time/Interval Use Gating, Windows (uniform/rect. here, pulse is self-windowing) Any Analysis Type, Including Vector, Demodulation Aerospace & Defense Symposium Page 50

51 Triggering: Find & Capture Signal of Interest Powerful Measurement Leverage Easy, Effective Ways to Monitor Environment for Signal Look for Expected and Unexpected Signals Avoid Measuring When No Signal Present Monitor Other Frequency Bands? Trigger on Other Activities External trigger from your circuit Oscilloscope or logic analyzer Consider all you know about signals, systems, transitions Take advantage of repeating signals, inter-signal timing, pos/neg delays Triggering Can Enhance Measurement Performance Time or synchronous averaging Periodic trigger Trigger a Time Capture Page 51

52 Special Triggers: IF Mag. & Frequency Mask IF Magnitude Real-time calculation of magnitude in selected span Precise, repeatable time alignment Negative & positive trigger delays Selectable level & polarity Selectable holdoff, holdoff type Playback IF magnitude trigger Frequency Mask Real-time calculation of spectrum & test against spectral mask Upper, lower limits Build from trace & adjust or manual parameter entry Trigger timing ambiguity ± 1 FFT Logic: Enter/leave, in/out, enter leave, leave enter Negative & positive trigger delays Time qualified triggering (TQT) Page 52

53 More Information 2013 Keysight A/D Symposium paper: Understanding Probability of Intercept for Intermittent Signals by Richard Overdorf Real-Time Spectrum Analyzer (RTSA) PXA X-Series Signal Analyzer, literature number EN Vector Signal Analysis Basics, literature number EN Webcast: Use capture, playback & triggering to completely analyze a signal spx?cc=ca&lc=eng&ckey= &nid= &id= &pselect=SR.GENERAL Page 53

54 Questions? Thank you for listening! Page