www.dlr.de Chart 1 Navigation für herausfordernde Anwendungen Robuste Satellitennavigation für sicherheitskritische Anwendungen PD Dr.-Ing. habil. Michael Meurer German Aerospace Centre (DLR), Oberpfaffenhofen
www.dlr.de Chart 2 Example Landing of Aircraft without Visual Sight Good times Not so good times CAT III- Landing with Pilot s Eyes CAT III Landing with Pilot s eyes. 100 50 30 ft Reliability of Navigation is essential for Safety of Lifes! We need: Trust in navigation solution even in case of system failures and in case of disturbances from outer sources -
www.dlr.de Chart 3 Sources of Interference in GNSS Weak GNSS signals compete with high power ground emissions weak GNSS signal GNSS Receiver Out-of-band Emissions (Harmonics of TV stations etc.)
www.dlr.de Chart 4 Interference Unintentional Out-of-Band-Emission Major technical reasons: Harmonics of powerful transmitted signals Intermodulation products Insufficient band limitation in transmitter front-ends Characteristics exempl. interference signal: Band width (bursts): 7 MHz Average signal power: -87.4 dbm Interference-to-Signal-Ratio: 42.6 db GPS C/Acarrier Galileo OSsub carrier L1 Band Numerous examples for further potential interference in L1: 2. Harm. DVB-T TV channel 60 & 61 (ca. 782-798 MHz, typ. Tx power: 10-100 kw) 3. Harm. analog UHF TV channel 27 & 28 (ca. 518-534 MHz, typ. Tx power: 100-3000 kw) 15./16./17./18. Harm. analog FM radio (87.3-105.2 MHz, typ. Tx power: 1-200 kw) 10. Harm. maritime VHF communication system (ca. 156-174 MHz, typ. Tx power: 25 W)
www.dlr.de Chart 5 Sources of Interference in GNSS Weak GNSS signals compete with high power ground emissions weak GNSS signal Accumulated Noise GNSS Receiver Inband Emissions (DME,TACAN etc.) Out-of-band Emissions (Harmonics of TV stations etc.)
www.dlr.de Chart 6 Example Inband interference due to DME (1) Distance Measurement Equipment (DME): Terrestrial aeronautical navigation aid Signal structure: impulse pair Gauß pulse with half-intensity width of 3,5 μs Separation of pulses in pair 12 μs DME signal is sequence of pulse pairs Typical Tx power of DME station: 1 kw Time Domain Frequency Domain Source: Thales ATM norm. amplitude time [μs] norm. power density [db] frequency [MHz]
www.dlr.de Chart 7 Exempl. Scenario Measured DME interference Hotspot Frankfurt: DME-Interference in E5 band in various flight levels E5b
www.dlr.de Chart 8 Sources of Interference in GNSS Weak GNSS signals compete with high power ground emissions Intentional Interference (Jamming) Intentional disturbance weak GNSS signal Accumulated Noise GNSS Receiver Inband Emissions (DME,TACAN etc.) Out-of-band Emissions (Harmonics of TV stations etc.)
www.dlr.de Chart 9 Interference by Personal Privacy Devices Intentional interference is reality! Personal Privacy Devices (jammers) heavily disturb GPS and GBAS Usage illegal Costs: $ 30 - $ 200 in Internet Mitigation techniques necessary GBAS Reference Antenna Jammer Source: R.H. Mitch et al., Signal Characteristics of Civil GPS Jammers, ION GNSS 2011 Airport Newark Liberty International, motorway close to airport
www.dlr.de Chart 10 Achilles heels of a GNSS Receiver analog Impact of high-power interference signals: digital Active components of analog part driven into compression, Erroneous adjustment of automatic gain control (AGC), Overdrive of A/D converter, Cross correlation products (CCP) due to interference more powerful than CCP of desired GNSS signal in correlation process Acquisition and tracking of GNSS-signals not possible Impact of interference has to be mitigated at all stages of processing within the GNSS receiver
www.dlr.de Chart 11 Analog Part Interference Mitigation (1) analog digital Active Antenna: Strong band limitation Optimized antenna diagram (Suppression of low elevations) Frontend: Optimized band limitation (minimum band with) and amplification Active components with largest possible linear region of operation
www.dlr.de Chart 12 Analog Part Interference Mitigation (2) analog digital Automatic Gain Control (AGC) and A/D converter: High dynamic of A/D converter (high word width) Optimized gain factors for AGC
www.dlr.de Chart 13 Digital Part Frequency Domain Adaptive Filtering Frequency Domain Adaptive Filtering (FDAF) frequency Pulseblanking Notchfiltering Hybrid (FDAF) Time
www.dlr.de Chart 14 Digital Part Frequency Domain Adaptive Filtering Frequency Domain Adaptive Filtering (FDAF) frequency FDAF: Detection and Suppression of Interference in Frequency and Time Domain Efficient Suppression for narrow band and pulsed interference Pulseblanking Notchfiltering Hybrid (FDAF) Time
www.dlr.de Chart 15 Exempl. Scenario Measured DME interference Hotspot Frankfurt: DME-Interference in E5 band in various flight levels E5b
www.dlr.de Chart 16 Frequency Domain Adaptive Filtering Performance C/N 0 degradation of Galileo E5b signal due to DME interference no FDAF FDAF
www.dlr.de Chart 17 Analog and Digital Part Digital Beamforming (BF) Adaptation of Antenna Gain (AG) of a 4 x 4 antenna array Basic principle of Adaptation: High antenna gain in direction of GNSS signal Low antenna gain ( Null ) in direction of interference.. w* 1 w* 2 w* KRx Further signal processing for PVT estimation
www.dlr.de Chart 18 Digital Beamforming BF after Correlation Beamforming (Weighting, adaptation of weights) after correlation for each GNSS signal (tracking channel) individually: K Rx K Rx N Sat * K Rx N Sat low rate processing, @ <1000 sps 1 Antenna K Rx RF RF Front End + Digitalization K Rx (I+Q) Dig. IF K Rx Code Correlation/ Tracking N Sat* K Rx Digital Beamformer N Sat Receiver Proc. Unit Nav. Solution Unit Carrier Tracking N Sat* K Rx Integrated GNSS Receiver K Rx : Number of Ant. Elements N Sat : Number of GNSS Satellites.. w* 1 w* 2 w* N
www.dlr.de Chart 19 Digital Beamforming Example for Performance Scenario: 2x2 Antenna Array, GPS L1 signal, 2 interferer signals Occurance of interference GNSS satellite Beamforming gain Receiver tracks GPS L1-signal with acceptable C/N0 Receiver looses tracking C/N0 tracking threshold interferer 2 broad band 1MHz Interferer 1 CW ISR: J/S=45 db
www.dlr.de Chart 20 DLR Robust Adaptive Multi-Antenna GNSS Receiver cpci Gehäuse Arrayantenna Frontend FPGA Board PC
www.dlr.de Chart 21 Measurement Campaign in GATE Test Area Interference GPS Constellation Adaptive Antenna GATE Galileo TX
www.dlr.de Chart 22 Exemplary Interference scenario I DME signal DME signal: EUROCAE62 MOPS-Scenario Galileo E5a band Tx power: -60 dbm 0 dbm Jammer-to-Signal-Radio (JSR) : 20 db 80 db Single Antenna Reception (without beamforming)
www.dlr.de Chart 23 Interference Suppression FDAF Achievable C/N 0 vs. jammer power with FDAF without FDAF
www.dlr.de Chart 24 Exemplary Interference scenario II broad band noise Broad band noise: Typ: AGWN Band width: 4 MHz Spectrum at Analog Front End Output with AWGN RFI Tx power: -60 dbm 0 dbm Jammer-to-Signal-Radio (JSR) : 20 db 80 db Broad band noise (Tx power = -15 dbm) JSR ~ 65 db
www.dlr.de Chart 25 Interference Suppression by Spatial Processing Without Interference With Interference (RFI) approx sat position approx interferer position
www.dlr.de Chart 26 Interference Suppression by Spatial Processing Without Interference With Interference (RFI) RFI on RFI off 60 C/N 0 (db-hz) 55 50 45 40 35 spatial proc. conventional 30 14:31:20 14:32:00 14:32:40 14:33:20 14:34:00 14:34:40 approx sat position approx interferer position
www.dlr.de Chart 27 DBF Interference Suppression Galileo PRN 19 No RFI RFI Approx PRN Position Approx RFI Position
www.dlr.de Chart 28 DBF Interference Suppression Galileo PRN 19 No RFI RFI 60 55 RFI on RFI off C/N 0 (db-hz) 50 45 40 35 30 14:31:20 14:32:00 14:32:40 14:33:20 14:34:00 14:34:40 Approx PRN Position Approx RFI Position
www.dlr.de Chart 29 DBF Interference Suppression Galileo PRN 19 No RFI RFI 60 RFI on RFI off 55 C/N 0 (db-hz) 50 45 40 Performance of digital beamforming depends on spatial separation between line of sight direction of interferer and GNSS signal 35 30 14:31:20 14:32:00 14:32:40 14:33:20 14:34:00 14:34:40 Approx PRN Position Approx RFI Position
www.dlr.de Chart 30 Spatially Selective Signal Blockage N 11 3 W 13 1 7 E 31 S 17 Jammer Spatial signal loss due to interference
www.dlr.de Chart 31 Multi-Antenna Receiver with Vector Loops Interferer is dominant signal after ADC Signal 20 db below noise Signal subspace with suppressed RFI used for beamforming Aiding weak signals Pre- Multi- Eigen- Code Discr. whiten Antenna- beam- -ing Correlator forming Code Gen. Freq. Discr. Code NCO VDFLL Kalman Filter PVT Carr. Gen. Carr. NCO Sat 1 Sat N
www.dlr.de Chart 32 Measurement Campaign in Berchtesgaden 2012 Static RFI Source and Dynamic Receiver Jammer: Gate TMS Personal Privacy Devices (comparable to Newark scenario) Data recording: 2 bit, 200 s for post processing with multi antenna vector tracking receiver Eigenbeamforming, VDFLL Jammer Gate TMS Jammer Vehicle Rx Vehicle DBF Receiver
www.dlr.de Chart 33 Scenario without Interference Reference track RFI
www.dlr.de Chart 34 Scenario with Personal Privacy Device Jammer (1/3) Reference Receiver 1 Reference Receiver 2 DLR Multi Antenna Rx RFI
www.dlr.de Chart 35 Scenario with Personal Privacy Device Jammer (2/3) Reference Receiver 1 Reference Receiver 2 DLR Multi Antenna Rx RFI
www.dlr.de Chart 36 Scenario with Personal Privacy Device Jammer (3/3) Reference Receiver 1 Reference Receiver 2 DLR Multi Antenna Rx RFI
www.dlr.de Chart 37 Sources of Interference in GNSS Weak GNSS signals compete with high power ground emissions Intentional Interference (Jamming) Intentional disturbance Spoofing/ Meaconing (fake GNSS signal) weak GNSS signal Accumulated Noise GNSS Receiver Inband Emissions (DME,TACAN etc.) Out-of-band Emissions (Harmonics of TV stations etc.)
www.dlr.de Chart 38 GPS-Repeater Problem and Threat
www.dlr.de Chart 39 GPS-Repeater Problem and Threat Jumps especially in vertical position Only observed on the ground in vicinity of hangar Most likely due to RFI from GPS repeater Take-off Landing
www.dlr.de Chart 40 Robustness against fake GNSS signals Joint attitude estimation and exploitation of directional information to identify erronous signals (multipath / spoofing) Expected DoA? Estimated DoAs Attitude Estimation
www.dlr.de Chart 41 Summary and Conclusions Safety-of-life (SoL) applications require reliable navigation Driving factor up-to-now : Unintentional Interference Important for future: Intentional jamming in civil applications is highly relevant (see Newark Incident) as well as Intentional or Unintentional transmission of fake signals (spoofing/meaconing/gps repeater) Robust multi-antenna GNSS receivers have been developed and have the potential to mitigate threats
www.dlr.de Chart 42 Thank you for your attention!
www.dlr.de Chart 43