Microwave Doppler Radar Sensor for Detection of Human Vital Signs and Mechanical Vibrations Prof. Jenshan Lin University of Florida Gainesville, FL http://www.lin.ece.ufl.edu/ Video 2 A Simple Theory How does it work? Doppler Effect. x(t) sin(t) x(t) v0t T (t ) R(t ) d0 x(t ) Constant Velocity Frequency Shift Periodic Chest Wall Movement Phase Shift T (t) cos 2 ft (t) 4 d0 4 x(t) 2d0 R(t) cos 2 ft (t ) c 3 A Simple Theory • Use transmitted signal as reference to mix with received signal by using a mixer or multiplier (similar to direct down-conversion), the output baseband signal contains the vital sign signal in its phase. T (t) cos 2 ft (t) 4 d0 4 x(t) 2d0 R(t) cos 2 ft (t ) c T(t) x R(t) Baseband Signal B(t): 4 d0 4 x(t) B(t) cos 0 4 d0 Antenna-to-target round-trip delay 4 x(t) Phase modulation due to chest-wall movement 0 Phase delay in receiver circuit (t) (t 2d0 ) c 4 Small Angle Approximation (Linear) 4x(t ) 4d 0 B (t ) cos 0 90, 270 , … When small negligible If x(t) << and [(4d0)/+0] = 90°, 270°, … 4x(t ) 4x(t ) B (t ) sin This is what we want! It is similar to measuring very small phase noise using FM discriminator technique. 5 Range Correlation Effect • If the distance between radar sensor and target is small enough, close-in phase noise of R(t) and T(t) are correlated. Close-in phase noise correlated T (t) cos 2 ft (t) 4 d0 4 x(t) 2d R(t) cos 2 ft (t 0 ) c 4 d0 4 x(t) B(t) cos 0 ~ 0 @ short-range A. D. Droitcour, O. Boric-Lubecke, V. Lubecke, J. Lin, G. Kovacs, “Range Correlation and I/Q performance benefits in single chip silicon Doppler radars for non-contact cardiopulmonary signs sensing,” IEEE Trans. Microwave Theory Tech., Vol. 52, No. 3, pp. 838-848, March 2004 6 Early Research Effort Rabbit Human • J. C. Lin, "Noninvasive Microwave Measurement of Respiration," Proceedings of the IEEE, vol. 63, no. 10, p. 1530, Oct. 1975. 7 Early Research Effort 10 GHz • • K.-M. Chen and H.-R. Chuang, “Measurement of Heart and Breathing Signals of Human Subjects Through Barriers with Microwave Life-Detection Systems,” IEEE EMBC 1988. – 10GHz: 1.5 ft of dry bricks – 2GHz: 3 ft of dry bricks H.-R. Chuang, Y.-F. Chen, and K.-M. Chen, “Automatic ClutterCanceler for Microwave Life-Detection System,” IEEE Trans. Instrumentation and Measurement, Vol. 40, No. 4, August 1991. 8 First Non-Contact Vital Sign Sensor Chip • • • • 0.25 µm BiCMOS 1.6 GHz transmission frequency 3.75mmx3.75mm Output power = 6.5dBm Baseband out Balun RF in (from Rx antenna) RF out (to Tx antenna) Balun RF Out to Tx Antenna Buffer Free-running VCO. No PLL. VCO Mixer Balun LO Buffer VCO Mixer Test Structures Baseband + Buffer Baseband - Mixer RF In from Rx Antenna VCO RF Buffer LO Buffer RF Buffer A. D. Droitcour, O. Boric-Lubecke, V. Lubecke, J. Lin, “0.25m CMOS and BiCMOS Single Chip Direct Conversion Doppler Radars for Remote Sensing of Vital Signs,” IEEE International Solid State Circuits Conference Digest of Technical Papers, pp. 348-349, 2002. 9 Ka-band Bench-Top System Baseband LabVIEW DAQ Ka radar Antenna Subject under test 10 User Interface Labview – data acquisition and signal processing 11 Typical Test Result Time-Domain Signal Detected Signal (V) 3 2 1 0 -1 -2 -3 0 5 10 15 Time (Sec) 20 25 Calculated Heart Rate Beats/Min 95 90 2 % Higher than Reference 85 80 0 Detected heart beat Reference heart beat 2 % Lower than Reference 5 10 15 Time (Sec) 20 25 12 Overnight Measurement Beats/Min Measured from Back 80 Reference (Fingertip sensor) 60 100 Results of Ka - band radar 60 40 40 20 20 Breaths/Min Beats/Min 80 Carrier freq: 27.1 GHz Duration: 6 hours Accuracy: 92.96% (Accuracy defined as within 2% of reference) 0 0 1 2 3 4 Time (Hour) 5 0 6 C. Li, J. Lin, Y. Xiao, "Robust Overnight Monitoring of Human Vital Signs by a Non-contact Respiration and Heartbeat Detector," Proceedings of the 28th IEEE Engineering in Medicine and Biology Society Annual International Conference, pp. 2235-2238, September 2006 13 Nonlinear Doppler Phase Modulation Small angle approximation B(t ) cos( 4 xh (t ) 4 xh (t ) 4 xr (t ) 4 xr (t ) ) when = 90 and xh(t), xr(t) << However, at mm-wave, the respiration movement is no longer very small. xr(t) << 14 Nonlinear Doppler Phase Modulation Periodic body movements due to heartbeat and respiration xh(t) = mh·sinωht, xr(t) = mr·sinωrt B(t ) cos( 4 xh (t ) Jl ( k l k=0 l=0 4 xr (t ) 4 mh )J k ( k=1 l=0 ) 4 mr )cos(kr t lht ) k=2 k=0 l=0 k=-1 k=3 l=1 l=1 l=0 k=1 l=1 f 0 R1 R2 C1a R3 H1 C1b C. Li, Y. Xiao, J. Lin, "Experiment and Spectral Analysis of a Low-Power Ka-Band Heartbeat Detector Measuring from Four Sides of a Human Body," IEEE Transactions on Microwave Theory and Techniques, IMS2006 Special Issue, Vol. 54, No. 12, pp. 4464-4471, December 2006. 15 Accurate Measurement of Periodic Motion • Determining factors for harmonics due to phase modulation – Wavelength, or carrier frequency – Residual phase – Amplitude of periodic movement H1 : H 2 : H 3 : H 4 | J1 ( 4 m ) cos |:| J 2 ( 4 m )sin |: 4 m 4 m | J3 ( ) cos |:| J 4 ( )sin | . H1 H2 H3 H4 The amplitude of movement can be accurately determined from the ratio of harmonics! C. Li, J. Lin, "Non-Contact Measurement of Periodic Movements by a 22-40GHz Radar Sensor Using Nonlinear Phase Modulation," IEEE MTT-S International Microwave Symposium Digest, pp. 579-582, June 2007 16 Measurement Example 1 5 0.5 4 0 -0.5 0 5 10 Time (Second) 15 1 Harmonic Ratio Normalized Spectrum Baseband Signal (V) • Movement period T = 3 sec amplitude = 2 mm • fRF: 40 GHz, • Transmission power: 50 µW • Distance: 1.65m H3/H1 ratio H4/H2 ratio 2.056 mm 3 Meaured H3/H1 2 2.045 mm 1 0.5 Meaured H4/H2 0 0 0 1 1/3 2/3 1 4/3 Frequency (Hz) 5/3 2 (a) Baseband signal, spectrum 1.5 2 2.5 Displacement (mm) 3 (b) Displacement extraction, self-verification 17 Random Body Movement Cancellation – Concept Heart 90 0 0 Body I f Qf 90 Qb I b DAQ Demodulation DAQ • Physiological movements on both sides of the body move in the same direction relative to their respective detecting radar • Random body drift move in the opposite direction relative to their respective detecting radar 18 Random Body Movement Cancellation – Theory 4 xh1 (t ) 4 xr1 (t ) 4 y (t ) S ( t ) exp j Front: f 1 4 xh 2 (t ) 4 xr 2 (t ) 4 y(t ) 2 Back: Sb (t ) exp j xh1, xh2, xr1, xr2: physiological movements xh1 xh 2 xr1 xr 2 y y(t): random body movement Combine signals from both sides: 4 xh1 (t ) xh 2 (t ) 4 xr1 (t ) xr 2 (t ) S fb (t ) S f (t ) Sb (t ) exp j 1 2 y(t) (random body movement) disappeared! 19 Random Body Movement Cancellation – Experiment Heart Res 90 0 Res 0 Body I f Qf Qb DAQ Demodulation 90 Ib DAQ Isolation between two radars: antenna polarizations & slight frequency offset between TX and RX 20 Random Body Movement Cancellation – Result 1 0 -0.5 0.5 I Channel Q Channel 2 4 6 8 Time (Second) Back (b) 10 -0.5 -1 0 2 4 6 8 Time (Second) 10 12 (a) Front Back 0.5 0 0 12 I Channel Q Channel 0 Spectrum Front 0.5 -1 0 Amplitude (a) Combined Spectrum Amplitude 1 1 20 40 60 Beats/Min 80 100 120 (b) Recovered heartbeat 0.5 0 0 20 40 60 Beats/Min 80 100 C. Li, J. Lin, “Random Body Movement Cancellation in Doppler Radar Vital Sign Detection,” IEEE Transactions on Microwave Theory and Techniques, vol. 56, issue 12, pp. 3143-3152, December 2008. 120 21 Portable Module Can be attached to laptop and powered through USB cable or powered by battery. 22 Search and Rescue Robot 23 Software Configurable Radar Sensor Chip LO I + LO I 5.8 GHz Radar Receiver LNA Mixer VGA I MixerI PreAmp LNA RFin Q VGA Pre-Amp RFin Gm Boosted Bias DAQ Bandgap 3 Wire Control Unit Clock Data Constant Gm CPLD Iout + I Iout - VGA Q Qout + MixerQ Gm-boosted Bandgap Const-Gm bias 3-wire control Qout - Load Load Clk Data LOQ + LOQ UMC 0.13 µm CMOS 1.2×1.2 mm2 C. Li, X. Yu, D. Li, L. Ran, J. Lin, "Software Configurable 5.8 GHz Radar Sensor Receiver Chip in 0.13 μm CMOS for Non-contact Vital Sign Detection," IEEE RFIC Symposium Digest of Papers, June 2009 24 Test Result 0 -20 0 CSD Spectrum I/Q Signal [mV] 20 5 1 10 15 Time (Second) Respiration 20 25 CSD Spectrum I/Q Signal [mV] Power: 1.5 V battery 3-wire program: Xilinx XC9536 CPLD Antenna: 2-by-2 patch array, 9 dB gain DAQ: NI USB-6008, 12 bit, 0-5V input Heartbeat 0.5 0 0 20 40 60 Beats/Min 80 100 120 Detect from the back @ 0.5m away 20 0 -20 -40 0 5 1 10 15 Time (Second) 20 25 Respiration 0.5 0 0 Heartbeat 20 40 60 Beats/Min 80 100 120 Detect from the front @ 1.5m away 25 Applications Security Rescue Baby Monitor See-thru-wall Radar Battlefield Triage Telemedicine Sleep Apnea Veterinary Medicine 26 Applications • • • • • • Lie Detector – covert detection Emotion Detector – vital sensor in a phone Personal Radar Tricorder Sports – fitness center/gym RF Vibrometer Gaming/Entertainment – quiz contests, video games – virtual reality 27 Summary • Prototypes have been demonstrated at different integration levels – bench-top, modules, and IC chips. • Basically, the technology can be used to detect any motion of an object reflecting radio waves • With proper signal processing, useful and interesting information can be extracted for various applications. – Biometric Radar possible 28