Calibration Concepts of Multi-Channel Spaceborne SAR

Transcription

1 DLR.de Chart 1 > CEOS Workshop 2016 > Tobias Rommel > September 7 th, 2016 Calibration Concepts of Multi-Channel Spaceborne SAR T. Rommel, F. Queiroz de Almeida, S. Huber, M. Jäger, G. Krieger, C. Laux, M. Martone, M. Villano, S. Wollstadt, M. Younis German Aerospace Centre (DLR)

2 DLR.de Chart 2 > CEOS Workshop 2016 > Tobias Rommel > September 7 th, 2016 Outline Internal System Calibration via Calibration Loops Raw-Data Based Calibration Approach Experimental Verification Practical Implementation and Advantages Conclusions

3 DLR.de Chart 3 > CEOS Workshop 2016 > Tobias Rommel > September 7 th, 2016 Antenna Concept for Exemplary System e.g. Sentinel 1a NG (Next Generation) Dual Polarized On-Board Channel Calibration Hybrid Beam-Forming (Analog + Digital Beam-Forming) Digital Beam-Forming in Azimuth and Elevation Elevation Azimuth Analog Beam-Forming in Elevation

4 DLR.de Chart 4 > CEOS Workshop 2016 > Tobias Rommel > September 7 th, 2016 TR Module Simplified Block Diagram RF Tx Signal HPA LNA f 0 ADC ASIC (DBF) Down-Converter, Analog Beam-Forming and Digital Backend

5 DLR.de Chart 5 > CEOS Workshop 2016 > Tobias Rommel > September 7 th, 2016 Calibration Loops (1) RF Tx Signal Tx Cal During Transmission Rx Path is Switched off to Avoid Interferences Due to Antenna Coupling RF Cal Signal Rx Cal After Each Rx Window Sequentially

6 DLR.de Chart 6 > CEOS Workshop 2016 > Tobias Rommel > September 7 th, 2016 Calibration Loops (2) RF Cal Signal Short Cal After Each Rx Window Sequentially Sequential Calibration of all TR Modules after Each Receive Window Tx Rx Cal Tx Rx Cal Tx Rx Cal 0 First TRM Second TRM Third TRM t

7 DLR.de Chart 7 > CEOS Workshop 2016 > Tobias Rommel > September 7 th, 2016 Disadvantages of This Approach TR Module Characterization by Three Transfer Functions (Tx + Rx + Short) This Approach Needs too Long for a Large Number of TR Modules Very Complex Calibration Network is Required -> Heavy + Additional Errors Individual Calibration Routines for Down-Converter, Analog Beam-Forming Network and ADC s are Required

8 DLR.de Chart 8 > CEOS Workshop 2016 > Tobias Rommel > September 7 th, 2016 Raw-Data Based Calibration

9 DLR.de Chart 9 > CEOS Workshop 2016 > Tobias Rommel > September 7 th, 2016 Raw-Data Based Calibration using Spatial Correlation s 2 (t) s 1 (t) s 0 (t) Errors d d Assumptions: Statistically Uniform and Clutterlike Scene Equal Element Spacing d Each Antenna Element Illuminates Approximately the Same Area Properties of the Spatial Correlation Function: Depends Only on the Element Spacing and not on the Absolute Position Independent of the Scene for Homogeneous Scenes

10 DLR.de Chart 10 > CEOS Workshop 2016 > Tobias Rommel > September 7 th, 2016 Raw-Data Based Calibration using Spatial Correlation Amplitude A corr,i = E s 0(t) s 0 (t) E s i (t) s i (t) = a 0 a i Ratio of Auto-Correlation Functions to Normalize Signal Intensities between Channels Phase Spatial Covariance Function between Elements i and j: σ i,j = E s i (t) s j (t) Differential Phase Shift: ψ i,j = arg σ i,j = φ i + (δ i δ j ) - Linear in a Uniform Array along Azimuth - Identical for all Adjacent Channel Pairs Phase Errors δ i, δ j

11 DLR.de Chart 11 > CEOS Workshop 2016 > Tobias Rommel > September 7 th, 2016 Simulation Results Target Simulation System Parameter: Element Spacing: 0.5 λ Center Frequency: 6.0 GHz No. Elements in Elevation: 12 No. Targets: 2000 Maximum Phase Error: ±15

12 DLR.de Chart 12 > CEOS Workshop 2016 > Tobias Rommel > September 7 th, 2016 Simulation Results System Parameter: Element Spacing: 0.5 λ Center Frequency: 6.0 GHz No. Elements in Elevation: 12 No. Targets: 2000 Maximum Phase Error: ±15 Differential Phase ψ i,j between Antenna Elements Residual Phase Error After Calibration: < 1.1

13 DLR.de Chart 13 > CEOS Workshop 2016 > Tobias Rommel > September 7 th, 2016 Simulation Results Array Factor System Parameter: Element Spacing: 0.5 λ Center Frequency: 6.0 GHz No. Elements in Elevation: 12 No. Targets: 2000 Maximum Phase Error: ±15

14 DLR.de Chart 14 > CEOS Workshop 2016 > Tobias Rommel > September 7 th, 2016 Experimental Verification of the Raw-Data Based Calibration Approach

15 DLR.de Chart 15 > CEOS Workshop 2016 > Tobias Rommel > September 7 th, 2016 MIMO-SAR Demonstrator Transmit Power: 21.5 m Carrier Frequency: 9.58 GHz Sampling Frequency: 1.0 GHz Signal Bandwidth: 300 MHz Transmit Channels: 4 (1 used) Receive Channels: 8 (4 in Elevation used) Velocity: 10 cm/s Polarization: VV Fully Coherent

16 DLR.de Chart 16 > CEOS Workshop 2016 > Tobias Rommel > September 7 th, 2016 Measurement Setup Fence Rails 4.0 m 10.0 m

17 DLR.de Chart 17 > CEOS Workshop 2016 > Tobias Rommel > September 7 th, 2016 Measurement Setup

18 DLR.de Chart 18 > CEOS Workshop 2016 > Tobias Rommel > September 7 th, 2016 Experimental Results Impulse Response of a Single Corner Reflector after DBF

19 DLR.de Chart 19 > CEOS Workshop 2016 > Tobias Rommel > September 7 th, 2016 Experimental Results Processed SAR Image After Elevation DBF Uncalibrated Calibrated Significantly Improved Contrast Between Meadow and Fence

20 DLR.de Chart 20 > CEOS Workshop 2016 > Tobias Rommel > September 7 th, 2016 Advantages of the New Approach

21 DLR.de Chart 21 > CEOS Workshop 2016 > Tobias Rommel > September 7 th, 2016 Advantages of the New Approach Simplified Architecture of TR Module New: Before: Path for Tx Cal Allows for Simultaneous Error Correction of Multiple TR Modules No Dedicated Calibration Window Required -> Increased Swath Width

22 DLR.de Chart 22 > CEOS Workshop 2016 > Tobias Rommel > September 7 th, 2016 Conclusions The Conventional Calibration Methods Currently used for Spaceborne SAR Systems cannot be Applied to Multi-Channel Systems New Data-Driven Approaches have been Investigated and are Promising It is Anticipated that Data-Driven Calibration will Supplement the Current Methods => Combined Approach is Suggested This Work is Part of the ESA Study Calibration And Data Reduction for Digital Beamforming Instruments

23 DLR.de Chart 23 > CEOS Workshop 2016 > Tobias Rommel > September 7 th, 2016 References [1] K. Kikuchi, M. Nishihara, H. Yamamoto, S. Mizuno, F. Yamaki and T. Yamamoto, An X-Band 300-Watt Class High Power GaN HEMT Amplifier for Radar Applications, SEI TECHNICAL REVIEW, No. 81, Oct [2] ESA study Single GaN Chip HPA/LNA for Radar Applications Consortium: TNO+Astrium+UMS [3] E. Hesham Attia and Bernard D. Steinberg, Self-cohering Large Antenna Arrays Using the Spatial Correlation Properties of Radar Clutter, IEEE Trans. on Antennas and Propagation, vol. 37, no. 1, January 1989 [4] G. Farquharson, P. Lopez-Dekker S.J. Frasier, Contrast-Based Phase Calibration for Remote Sensing Systems With Digital Beamforming Antennas, IEEE Trans. on Geoscience and Remote Sensing, vol. 51, no.3, pp , March 2013 [5] T. Rommel, M. Younis and G. Krieger, Demonstration of simultaneous quad-polarization SAR imaging for extended targets in MIMO-SAR, 2016 German Microwave Conference (GeMiC), Bochum, 2016, pp