By Associate Professor Electrical Engineering Department King Fahd University of Petroleum & Minerals Dhahran, Saudi Arabia Wednesday, December 1, 14 1 st Saudi Symposium for RADAR Technology 9 1 December 14 Riyadh, Saudi Arabia 1
Outline Introduction to TWRI Applications High Resolution(UWB, SAR) Image Formation Challenges Multipath propagation and ghosts Front wall effects Antenna directivity Image formation Wall characterization and compensation Measurement Results Compensation Results Conclusion Recommendations
Applications & Prototypes
High Resolution Imaging To get higher resolution Image: 1) Use of Wide bandwidth (UWB) ) Use Wide Aperture, Antenna Array or Synthetic Aperture (SAR) 4
What is UWB? UWB signals have BW > 5 MHz or B f. = BW f c = f H f L (f H +f L )/ Advantages Trade bandwidth for a reduced transmission power Immunity to multipath Low cost and low complexity (promissed) Disadvantages Coexistence with other systems Fast ADC requirement Application High data rate/ High positioning resolution Short distance Low power Though-wall imaging 5
Synthetic Aperture Radar 6
Beamforming & Image Formation Target τ mp τ mq M M II I xx q q = w w m m s t τ mp + ττ mq m=1 m=1 7 h tt t=
Image Formation (SAR) 8
Challenges in TWRI 1. Multipath Propagation & Ghosts. Front Wall: 1. Attenuation. Reflection 3. Reverberation (ringing) 3. Thermal noise & Interference 4. Limited Measurement Capability 9
Multipath 1
Ghost Formation Number of Targets= 1 1 1 z-axis "pixel" 3 4 5 6 7 Ghosts 9 8 7 6 5 4 3 8 9 Target 1 1 4 6 8 1 1 14 x-axis "pixel" 11
Front Wall 1
Front Wall Compensation Objective: Compensate for the UWB through-wall propagation Why? How? EM waves through a wall gets attenuated slows down gets dispersed Attributed to dispersive and attenuative properties of the wall In localization and Imaging: Defocuses target image Displaces the target from its true position. Investigating the interaction of the electromagnetic wave incident on the wall Characterizing the wall in terms of electrical properties (insertion transfer function, delay, dielectric constant, insertion loss etc) Using the wall model to compensate for the wall effect 13
Literature in Wall Characterization Literature Building Material Methodology Frequency Reported Parameters Application Type Condition Structure Simulation Theoretical Experimental Oblique incidence Direct incidence UWB X Band S Band 16 GHz 1 1 GHz Dielectric constant Insertion loss Return loss Transmission coefficient Reflection coefficient Communication Imaging Detection & Localization RF/Shielding Delay Loss tangent 14
Short Background on Dielectric Properties of Walls Every material has a unique set of electrical characteristics dependent on its dielectric properties Dielectric stores energy when an external electric field is applied Electric flux density D is: r = absolute permittivity, ε r = relative permittivity, ε = permittivity of free space E = the electric field. Permittivity complex quantity. Dielectric constant = relative permittivity (ε r ) (ε r ) energy storage (ε r ) - loss factor 15 r D E ' r j '' r Parameter r tan d Τ loss tangent = energy lost energy stored '' r = tan ' r Meaning dielectric Constant conductivity delay loss tangent slab thickness propagation constant Attenuation coefficient Phase coefficient Reflection coefficient Transmission coefficient
Measurement Setup Measuring magnitude and phase components are important in that They are needed to fully characterize the obstruction IFFT - impulse response characterization Calibration Measurements can be performed in both time domain and frequency domain techniques using coaxial or transmission line methods, or free-space radiated methods Frequency domain Higher dynamic range No synchronization Noise reduction techniques Free-space radiated setup Contactless, non-destructive Proper machining with cavity/waveguide Matches the final application (radar, communication) 16
Frequency Domain Setup Amplifier Network Analyzer Antenna Antenna air air 1.5 m cable Wall N-type to SMA connector Low noise amplifier Port 1 Vector Network Analyzer Port 4.5 m cable 1.5 m cable S Parameter Test Set Data Processing Wideband cables 17
Transmission Measurement Free Space E i E t fs Region II (Material) Region I (Air) Region III (Air) E i E r d E t H ( j) Et ( j) Ei ( j) fs Et ( j) E ( j) i E E t fs t ( j) ( j) X X t fs t ( j) ( j) 18 z S 1 ( j) H( j) e j d c
Reflection Measurements Wall Material Metallic Reflector 19
Analysis Method Dielectric constant is related to insertion transfer function through a complex equation Free Space Incident wave establishes E i E t fs Reflected wave in region I (air) Forward-backward traveling waves in region II(wall) Transmitted wave in region III(air) Region II (Material) Imposing boundary conditions for E and H fields, Region I (Air) Region III (Air) E i E t E r d z
Wall Parameter Calculation Summary S 1 frequency Domain Transmission & Reflection measurements -1 - Freespace Through Wall S1( free space) H( j) S ( through wall ) 1 Filter IFFT on S 1 data, with zero padding to obtain impulse response Peak-to-peak time domain differential delay using a sliding correlator. First estimate of delay and loss Time gating to remove unwanted reflections FFT H( j) (gated) Use low loss equations r ' '', r H ( j) e ' Magnitude, Magnitude, db Amplitude in (V) -1-3 - -4-3 -5 x 1-4 -4-6 8-5 -7 6-6 -8 4-7 -9 Insertion Transfer Function Filter Filtered Insertion transfer function -1-8 4 4 6 6 8 8 1 1 Frequency, GHz 1 1 14 14 16 16 18 18 Amplitude in (V) - -4-6 1 x 1-4 Freespace Impulse Response 8 4 6 Through 8 wall Impulse 1 Response 1 14 time, ns 6 1 x 1-4 4 8 ' ' ( j ) d 6 r 1 ( j ) d r 1 e ' ' 4 r r - c Amplitude (V) ' ' r r - -6-4 -6 4e jd Freespace Impulse Response Through wall Impulse Response.4*Window(Free space).4*window(through wall) Free-Space Impulse Resopnse Through Impulse Response 4 5 6 7 8 9 1-4 time in secs x 1-9 1 Dielectric constant, Loss tangent, Attenuation constant -8 4 6 8 1 1 14 time, ns
Measurement Results Transmission and Reflection results Comparing Transmission & Reflection Repeatability & Variability Compare with Literature Multiple Walls Accuracy Related Issues
Insertion Transfer Function (Magnitude), db db Insertion Transfer Function (Magnitude), db db Transmission Transmission - - -4-4 -6-6 Wood Wood -8 Gypsum -8 Glass Gypsum Glass -1 4 6 8 1 1 14 16 18 4 6 Frequency, 8 GHz 1 1 14 16 18 Frequency, GHz Transmission Transmission - - -4-4 -6-6 Wood - fit Wood fit -8 Gypsum - fit -8 Gypsum fit Glass - fit Glass fit -1-1 4 6 8 1 1 14 16 18 1 1 14 16 18 Frequency, GHz Frequency, GHz 3 Insertion transfer function Coefficients of linear or quadratic fits to extracted the parameters af + b or af + bf + c, f(ghz) Wood Glass Gypsum a b c a b c a b c -.45 -.785 -- -.44-1.687 -- -.175 -.4615 --
Diecletric Constant Diecletric Constant 1 1 1 8 8 6 6 4 4 Reflection Reflection Reflection false solutions Wood Wood Gypsum Gypsum Glass Glass 4 6 8 1 1 14 16 18 4 6 1 1 14 16 18 Frequency, GHz Frequency, GHz Diecletric Constant Diecletric Constant 1 1 8 8 6 6 4 4 Reflection Reflection Wood - fit Wood - fit Gypsum - fit Gypsum - fit Glass - fit Glass - fit 4 6 8 1 1 14 16 18 4 6 Frequency, 8 1 GHz 1 14 16 18 Frequency, GHz Coefficients of linear or quadratic fits to extracted the parameters af + b or af + bf + c, f(ghz) 4 Wood Glass Gypsum a b c a b c a b c Dielectric constant -.945 -.91 4.1995 -.66 -.8 7.419 -.13 -.9 3.4
Multiple walls Single wood wall Double glass wall air gap 5 Spaced wall: wood and gypsum Three layer: glass-woodglass
Double layer walls -1 x 1-4 - Free-space single wood double wood 8 6 Free-space single wood double wood -3 4 Magnitude, db -4-5 Amplitude (V) -6 - -7-4 -8 4 6 8 1 1 14 16 18 Frequency, GHz -6 3.5 4 4.5 5 5.5 6 6.5 7 7.5 8 8.5 time, secs x 1-9 single wood double wood Insertion Transfer Function, db - -4-6 -8-1 -1-14 6 4 6 8 1 1 14 16 18 Frequency, GHz
Spaced double walls (Air gap) -1 - -3 Free-space free space with cm with 5cm with 1cm Magnitude, db -4-5 -6-7 various air gap sizes -8-9 -1 4 6 8 1 1 14 16 18 Frequency, GHz 1 x 1-3.8.6 free space various air gap sizes Free-space cm 5cm 1cm.4 Amplitude (V). -. -.4 -.6 -.8 7-1 4 4.5 5 5.5 6 6.5 7 7.5 8 8.5 time, secs x 1-9
Sources of Error Antenna Antenna Characteristics Alignment Antenna-Wall separation Cables & Connectors Cable length Connector mismatch Accuracy Related Issues Measurement setup & Lab environment Air-gaps, edge effects Scatterers Repeatability and Variability Magnitude, db -1 - -3-4 15 GHz Cable A Cable B Cable C Cable ABC -5 4 6 8 1 1 14 16 18 Frequency, GHz 8
Insertion Transfer Function, db Insertion Transfer Function, db -1-1 - - -3-3 -4-4 -5-5 -6-6 Repeatability and Variability Wood, Meas 1 Wood, Meas Wood 1-7 -7 44 6 8 1 1 14 16 18 Frequency, GHz Repeatability of measurements allows us obtain the same results for measurements taken at different instances of time for the same wall sample Measurements are said to have low variability if they yield approximately the same results for different samples of the same material Dielectric constant Dielectric constant 44 3.5 3.5 3 3.5.5 Wood,, Meas 1 Wood Wood,, Meas Meas Wood Wood 1 1.5 1.5 4 6 8 1 1 14 16 18 4 6 Frequency, 8 GHz 1 1 14 16 18 9 Frequency, GHz
5 x 1-4 Wood Wall Compensation Amplitude, V -5 Target Only (Everything except wall) Target + Wall (Everything + Wall) Compensated 7 7.5 8 8.5 9 9.5 Time, sec x 1-9 x 1-4 Glass Amplitude, V 4 - Target Only (Everything except wall) Target + Wall (Everything + Wall) Compensated -4 Amplitude, V x 1-4 4 - Gypsum Target Only (Everything except wall) Target + Wall (Everything + Wall) Compensated 7 7.5 8 8.5 9 9.5 Time, sec x 1-9 -4 3 7 7.5 8 8.5 9 9.5 Time, sec x 1-9
Wall Compensation Correct the adverse effect of the wall On the outcome of the detection process so that the true target position can be obtained On the pulse shape for imaging communication & receiver design purposes Using 3 methods Constant Amplitude and Delay (CAD) Frequency Dependent Data (FDD) Data Fitting (FIT) 31
Target Measurement Target Only Target + Wall d1 x 1-4 Target Object range Tx Rx Measurement System Amplitude, V - Target Only Target + Wall d antenna -4 3 Wall Material -4 4 6 8 1 1 Time, sec x 1-9
Constant Amplitude & Delay Method (CAD) S 1 frequency Domain Transmission & Reflection measurements S1( free space) H( j) S ( through wall ) 1 Filter Assumption of Constant amplitude attenuation Constant delay due to wall S 1 frequency Domain Target measurements Target + Wall IFFT Target Only IFFT IFFT on S 1 data, with zero padding to obtain impulse response Target + Wall Target Only Peak-to-peak time domain differential delay using a sliding correlator. First estimate of delay and loss CAD a Target +Wall t_shift Compensated version of Target Only Time gating to remove unwanted reflections FFT H( j) (gated) 1 x 1-4 8 6 4 Freespace Impulse Response Through wall Impulse Response a thruwall ( t) freespace ( t) Use low loss equations Amplitude in (V) - t _ shift ( transmissi on) r ' '', r -4-6 33 Dielectric constant, Loss tangent, Attenuation constant 4 5 6 7 8 9 1 time in secs x 1-9
Comparing the three Methods x 1-4 Wood 5 4 3 Target Only Target + Wall CAD-Compensated FDD-Compensated FIT-Compensated Amplitude, V 1-1 - -3-4 -5 7 7. 7.4 7.6 7.8 8 8. 8.4 8.6 8.8 9 Time, sec x 1-9 34
Summary Conclusion Wideband EM characterization of walls (unique band) Wood, glass, gypsum Transmission and Reflection Insertion loss, dielectric constant 1-18 GHz Results compare good with literature Multiple walls Double layer Three layer Wall compensation Constant Amplitude & Delay (CAD) Frequency Dependent Data (FDD) Data Fitting (FIT) 35
Current Research: Direction with Single Antenna 36
THANK YOU Book Chapter A. H. Muqaibel, M. Al-Sunaidi, N. M. Iya and A. Safaai-Jazi, Chapter 1: Wall Attenuation and Dispersion, Book title: "Through Wall Radar Imaging". Publisher: CRC Press. Publication Date: Dec 13 1 Acknowledgment Mr. Mohammad Tamim Al-Khdary for Helping in preparing this presentation