www.dlr.de Chart 1 > Antenna Arrays for Robust GNSS > A. Konovaltsev > 17.11.2014 Antenna Arrays for Robust GNSS in Challenging Environments Presented by Andriy Konovaltsev Institute of Communications and Navigation German Aerospace Center (DLR) International Technical Symposium on Navigation and Timing 17-Nov-2014, ENAC, Toulouse, France
www.dlr.de Chart 2 > Antenna Arrays for Robust GNSS > A. Konovaltsev > 17.11.2014 Radio Frequency Interference in GNSS Weak GNSS signals compete with high power ground emissions and accumulated interference sources. Fake GNSS Signal Weak GNSS Signal Accumulated Noise (UWB etc.) GNSS Receivers High Power Pulsed (DME,TACAN etc.) Jamming (e.g. PPD) High Power Continuous Wave (Harmonics from TV stations etc.)
www.dlr.de Chart 3 > Antenna Arrays for Robust GNSS > A. Konovaltsev > 17.11.2014 Countermeasures against Interference Antenna Array Hardware Array Signal Processing Appropriate antenna reception pattern Out-of-band filtering High compression point of LNA High dynamic range Multi-bit ADC, Interferer detection Adaptive ADC levels Short relaxation time of AGC Digital pulse blanking Adaptive notch filter Despread processing gain Robust design of tracking loops Deeply coupling with INS Coupling with INS Sensor fusion
www.dlr.de Chart 4 > Antenna Arrays for Robust GNSS > A. Konovaltsev > 17.11.2014 Added Value of Antenna Array Processing Signal of Interest Radio Frequency Interference Mitigation Unwanted signal? Directions of arrival Spoofing Detection N Multipath Mitigation W E S Attitude Determination Improved Availability Better Accuracy
www.dlr.de Chart 5 > Antenna Arrays for Robust GNSS > A. Konovaltsev > 17.11.2014 Outline Algorithms used with antenna arrays Pre- and post-correlation processing Practical realization of a GNSS receiver with an antenna array Antenna and RF frontend Digital signal processing Exemplary test results Summary
www.dlr.de Chart 6 > Antenna Arrays for Robust GNSS > A. Konovaltsev > 17.11.2014 Pre-Correlation Techniques (1) GNSS signals buried into noise Radio frequency interference is well observable, which can be helpful e.g. for estimation of jammer direction of arrival (DOA) Often used in such architecture: 1 2 Front-end A/D Front-end A/D 3 L Front-end A/D w 1 w 2 w 3 Σ y(t) Correlation & Tracking Desired reception pattern: Spatial zero to mitigate jammer Front-end A/D w L w x(t) Array Weights Control
www.dlr.de Chart 7 > Antenna Arrays for Robust GNSS > A. Konovaltsev > 17.11.2014 Pre-Correlation Techniques (2) Power minimisation algorithm for jammer mitigation: w ooo = arg min w E y t 2 = arg min w wh R xx w subject to w 1 = 1 Space-time adaptive processing: 1 Front-end A/D Z -1 Z -1 Z -1 Z -1 w 1,center tap = 1 w 11 w 12 w 13 w 1(K-1) w 1K 2 Front-end A/D Z -1 Z -1 Z -1 Z -1 3 Front-end A/D w 21 w 22 w 23 Z -1 Z -1 w 3(K-1) w 2K Z -1 Z -1 Σ Correlation & Tracking w 31 w 32 w 33 w 3(K-1) w 3K L Front-end A/D Z -1 Z -1 Z -1 Z -1 w 41 w 42 w 43 w 4(K-1) w 4K
www.dlr.de Chart 8 > Antenna Arrays for Robust GNSS > A. Konovaltsev > 17.11.2014 Pre-Correlation Techniques (2) Power minimisation algorithm for jammer mitigation: w ooo = arg min w E y t 2 = arg min w wh R xx w subject to w 1 = 1 Space-time adaptive processing: 1 Front-end A/D Z -1 Z -1 Z -1 Z -1 w 1,center tap = 1 w 11 w 12 w 13 w 1(K-1) w 1K 2 Front-end A/D Z -1 Z -1 Z -1 Z -1 3 Front-end A/D w 21 w 22 w 23 Z -1 Z -1 w 3(K-1) w 2K Z -1 Z -1 Σ Correlation & Tracking w 31 w 32 w 33 w 3(K-1) w 3K L Front-end A/D Z -1 Z -1 Z -1 Z -1 w 41 w 42 w 43 w 4(K-1) w 4K
www.dlr.de Chart 9 > Antenna Arrays for Robust GNSS > A. Konovaltsev > 17.11.2014 Post-Correlation Techniques (1) Due despread processing gain, GNSS signals have positive SNR Beamforming goal is to optimize Signal-to-Noise-plus-Interference Ratio (SNIR) Parameter, e.g. DOA, estimation of GNSS signals and multipath echoes, spoofing detection. 1 2 Front-end A/D Correlation w 1 Beamforming Tracking 1 3 L Front-end Front-end A/D A/D Correlation w 2 Beamforming Tracking 2 Front-end A/D Correlation w N Beamforming Tracking N
www.dlr.de Chart 10 > Antenna Arrays for Robust GNSS > A. Konovaltsev > 17.11.2014 Post-Correlation Techniques (2) Minimum Mean Squared Error (MMSE) or reference signal beamforming: w ooo = arg min w E r t y t 2 = arg min w E r t wh x 2 Eigenbeamforming [1]: reference signal, e.g. data modulation w ooo = u d where u d is eigenvector corresponding to the highest eigenvalue of post correlation signal covariance matrix R xx Typical array beam pattern with post-correlation beamforming: [1] M. Sgammini et al, Blind Adaptive Beamformer Based on Orthogonal Projections for GNSS, ION GNSS, Sept. 2012, Nashville, TN, USA
www.dlr.de Chart 11 > Antenna Arrays for Robust GNSS > A. Konovaltsev > 17.11.2014 Practical Realization: Antenna Arrays Developed by DLR Galileo E1/E6 standard and miniaturized Galileo E1/E5 standard and miniaturized GPS miniaturized GPS conformal
www.dlr.de Chart 12 > Antenna Arrays for Robust GNSS > A. Konovaltsev > 17.11.2014 Practical Realization: Complete System active antenna front end FPGA embed. PC directional coupler RF filter LNA RF filter RFamp I Mixer LP filter IFamp I IFamp I BP filter VGA A/D- Wandler 14 Bit Virtex-4 cpci LO cpci Calibration Signal Array high-rate processing (DDC, correlation) cpci Chassis low-rate processing (tracking loops, array processing, PVT)
www.dlr.de Chart 13 > Antenna Arrays for Robust GNSS > A. Konovaltsev > 17.11.2014 Practical Realization: Signal Processing (1) Hardware used since 2007: 8 ADC channels with 14 bit @104 MSPS Virtex-4 SX55 Embedded Intel Dual Core 2 GHz CPU cpci Bus (Transferrate: 20 MByte/s DMA) Windows XP operation system Up to 14 satellite channel with 4- element array, MMSE beamforming and DOA estimation using 2-D ESPRIT algorithm
www.dlr.de Chart 14 > Antenna Arrays for Robust GNSS > A. Konovaltsev > 17.11.2014 Practical Realization: Signal Processing (2) Hardware used now: 16 ADC channels with 14 Bit @ 125 MHz Xilinx Virtex 6 SX475T Intel I7 Quad-Core Host CPU PCIe Interface with up to 700 Mbyte/s Linux operation system About 50 satellite channels with 4-element array 9 channel E1/L1 front-end
www.dlr.de Chart 15 > Antenna Arrays for Robust GNSS > A. Konovaltsev > 17.11.2014 Testing Array System: Lab Simulations Use of theoretical array model Controlled environment Large variety of possible scenarios 4 x phase-aligned GNSS simulators GALANT Receiver
www.dlr.de Chart 16 > Antenna Arrays for Robust GNSS > A. Konovaltsev > 17.11.2014 Field Test Results from Galileo Test Environment (GATE) Gate TMS GPS constellation Gate TMS RFI Gate TMS Receiver under Test - 8 Galileo Pseudolites, Virtual Satellite Mode, - Static and dynamic interference scenarios - Repeater and spoofing tests
www.dlr.de Chart 17 > Antenna Arrays for Robust GNSS > A. Konovaltsev > 17.11.2014 Measurement Set-Up -Interference -Adaptive Antenna -GATE Tx -Kehlsteinhaus
www.dlr.de Chart 18 > Antenna Arrays for Robust GNSS > A. Konovaltsev > 17.11.2014 Example of Array Beam Pattern (1) No RFI RFI RFI 21 21 DOA of SV MMSE beamforming is used DOA of Jammer
www.dlr.de Chart 19 > Antenna Arrays for Robust GNSS > A. Konovaltsev > 17.11.2014 Example of Array Beam Pattern (2) No RFI RFI 15 15 21 RFI RFI 21 DOA of SV MMSE beamforming is used DOA of Jammer
www.dlr.de Chart 20 > Antenna Arrays for Robust GNSS > A. Konovaltsev > 17.11.2014 Repeater Set-Up to Test Spoofing Detection Repeater Tx antenna Repeater Rx antenna Repeater Tx GALANT antenna Antenna LNA
www.dlr.de Chart 21 > Antenna Arrays for Robust GNSS > A. Konovaltsev > 17.11.2014 DOA Estimation for Repeater Signals 90 90 0 180 0 270 270 Static scenario Dynamic scenario
www.dlr.de Chart 22 > Antenna Arrays for Robust GNSS > A. Konovaltsev > 17.11.2014 Availability of PVT, Effect of Signal Blockage and Multipath (1) Positioning using single element
www.dlr.de Chart 23 > Antenna Arrays for Robust GNSS > A. Konovaltsev > 17.11.2014 Availability of PVT, Effect of Signal Blockage and Multipath (2) Positioning using the entire array
www.dlr.de Chart 24 > Antenna Arrays for Robust GNSS > A. Konovaltsev > 17.11.2014 Combining Beamforming and Vector Tracking Fusion of vector tracking and digital beamforming Ideally suited for aiding satellites with blocked LOS or small spatial separation to RFI sources (no re-acquisition in interfered scenarios!) Significantly improved performance in terms of interference robustness and positioning availability Correlator Correlator Correlator Correlator Digital Beamforming Code Discr. Freq. Discr. Code Gen. Code NCO VDFLL Kalman Filter PVT Carr. Gen. Carr. NCO Sat 1 Sat N GALANT Presentation > A.Konovaltsev > 9 March 2012
www.dlr.de Chart 25 > Antenna Arrays for Robust GNSS > A. Konovaltsev > 17.11.2014 Test Results in PPD-Jammer Scenario Reference Receiver 1 Reference Receiver 2 DLR Multi Antenna Rx RFI M. Cuntz, A. Konovaltsev, C. Hättich, G. Kappen, and M. Meurer, Vector Tracking with a Multi Antenna GNSS Receiver, ION GNSS 2012, Sept. 2012, Nashville, TN, USA.
www.dlr.de Chart 26 > Antenna Arrays for Robust GNSS > A. Konovaltsev > 17.11.2014 Test Results in Urban Environment Array Antenna COTS Antenna - Multi Antenna Vector Tracking - Single Antenna LS PVT
www.dlr.de Chart 27 > Antenna Arrays for Robust GNSS > A. Konovaltsev > 17.11.2014 Benefit of Beamforming in Rural Scenario (1) Single antenna Beamforming* * Eigenbeamforming algorithm
www.dlr.de Chart 28 > Antenna Arrays for Robust GNSS > A. Konovaltsev > 17.11.2014 Benefit of Beamforming in Rural Scenario (2) Az: 314 El: 75 Az: 311 El: 7 Standard deviation of code-minus-carrier observables: PRN 5 8 15 21 26 27 28 Elevation, 44 52 30 7 75 23 60 σ_psr, Single Antenna, m σ_psr, Multi Antenna, m 6.03 33.12 1.20 2.44 6.15 1.55 20.35 1.67 1.54 0.98 1.41 0.62 0.84 3.39
www.dlr.de Chart 29 > Antenna Arrays for Robust GNSS > A. Konovaltsev > 17.11.2014 Summary Antenna arrays can be very helpful in challenging environments such as radio frequency interference, multipath, signal obstructions etc. Solutions for practical realization are already available today. From application point of view, introduction of this technology is a tradeoff between expected benefits and increased complexity/costs. Trends in the field of GNSS antenna arrays: Array miniaturization Use of COTS antenna elements for building low-cost arrays Sophisticated calibration methods to keep induced code and carrier range biases very low
www.dlr.de Chart 30 > Antenna Arrays for Robust GNSS > A. Konovaltsev > 17.11.2014 Back-up Slides
www.dlr.de Chart 31 > Antenna Arrays for Robust GNSS > A. Konovaltsev > 17.11.2014 Practical Realization: Further Examples of Antennas Arrays 7-Element Array by SATIMO (France) Frederic Leveau et al, Anti-Jam Protection by Antenna: Conception, Realization, Evaluation of a Seven-Element GNSS CRPA, GPS World, Feb. 2013 4-Element Array built with COTS Elements Yu-Hsuan Chen et al., Off-the-Shelf Antennas for Controlled-Reception-Pattern Antenna Arrays, GPS World, Feb. 2013
www.dlr.de Chart 32 > Antenna Arrays for Robust GNSS > A. Konovaltsev > 17.11.2014 Practical Realization: RF Front End Single PLL-Synthesizer per band Up to 9 channels Common 10 MHz reference OCXO Crosstalk Isolation > 50 db 30 MHz Bandwidth 9 ch E1 front end 4 channel E1 front end PCB Crosstalk simulations
www.dlr.de Chart 33 > Antenna Arrays for Robust GNSS > A. Konovaltsev > 17.11.2014 Characteristics of PPD-Jammer - Approximately 9 dbm output power - 13 MHz bandwidth - Sweep time 40 us - Center frequency 1.575 GHz Power/frequency (db/hz) Power/frequency (db/hz) 100 0-100 -200 0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2 Frequency (GHz) Periodogram Power Spectral Density Estimate -50-100 -150-200 1.53 1.54 1.55 1.56 1.57 1.58 1.59 1.6 1.61 1.62 Frequency (GHz)