Riding out the Rough Spots:Scintillation-Robust GNSS Carrier Tracking

Dr. Todd E. HumphreysRadionavigation LaboratoryUniversity of Texas at Austin

UT Radionavigation Laboratory

UT Radionavigation Lab Research Agenda GNSS Spoofing

Characterize spoofing signatures Develop receiver-autonomous defenses Develop augmentation-based defenses

(GPS + eLORAN + Iridium + …)

GPS Jamming Develop augmentation-based defenses Locate jamming sources by combining

data from a network of receivers

Indoor Navigation Pioneer collaborative navigation Develop augmentation-based indoor nav

techniques (GPS + eLORAN + Iridium + …)

Natural GNSS Interference Improve tracking loop robustness to

scintillation

UT Radionavigation Laboratory

Ionospheric Diagnosis via Arrays of GPS Receivers

Ionospheric Monitoring (sparse array) Ionospheric Tomography (dense array)

Incident plane wave

Disturbedionosphere

Diffractedwavefront

Linear array ofGRID receivers

Nominal magneticfield direction

CASESConnected Autonomous Space Environment Sensors

Cornell University, UT Austin, ASTRA LLC

AFOSR STTR Proposal, 2008

CASESConnected Autonomous Space Environment Sensors

Cornell University, UT Austin, ASTRA LLC

AFOSR STTR Proposal, 2008

UT Radionavigation Laboratory

CASES Sensor Evolution

V0V1

V2

UT Radionavigation Laboratory

Carrier Tracking GoalsReceiver noise and scintillation-induced

phase errors Cycle slips (phase unlock)

Total loss of carrier lock (frequency unlock)

Analyze scintillation effects on GPS receivers; isolate cause of phase unlock

Model scintillation well enough to generate realistic synthetic scintillation

Synthesize scintillation to test tracking loop strategies

Design phase tracking loops for operation in scintillation

Strategy

Long-term Goals Eliminate frequency unlock Minimize cycle slips and

generally reduce phase errors

UT Radionavigation Laboratory

Carrier Tracking Goals

Analyze scintillation effects on GPS receivers; isolate cause of phase unlock

Model scintillation well enough to generate realistic synthetic scintillation

Synthesize scintillation to test tracking loop strategies

Design phase tracking loops for operation in scintillation

Long-term Goals Eliminate frequency unlock Minimize cycle slips and

generally reduce phase errorsStrategy

UT Radionavigation Laboratory

Analyze: The Empirical Scintillation LibraryCanonical fades

UT Radionavigation Laboratory

Fading Interpreted on the Complex Plane

UT Radionavigation Laboratory

Model: Distill Scintillation Down to Essential Characteristics for Carrier Tracking

Standard statistical

analysis techniques

Standard statistical

analysis techniquesDPSK bit error

prediction with Rice and 2nd-order

Butterworth models

DPSK bit error prediction with Rice

and 2nd-order Butterworth models

UT Radionavigation Laboratory

Synthesize: Turn the Model Around

UT Radionavigation Laboratory

Scintillation Simulator Implementation

UT Radionavigation Laboratory

Hardware-in-the-loop Scintillation Robustness Evaluation

Scintillation Simulator Simulated time history

GNSS Signal Simulator

GNSS Receiver Phase difference time history

UT Radionavigation Laboratory

Design: Scintillation-hardened Tracking Loops

• Straightforward approach: navigation data bit prediction

• Incorporate the observed second-order dynamics into a Kalman filter whose state includes the complex components of z(t)

• Combine this with a Bayesian multiple-model filter that spawns a new tracking loop whenever a data bit is uncertain. Prune loops at parity check.

GOAL: Ts > 240 seconds for {S4 = 0.8, 0 = 0.8 sec., C/N0 = 43 dB-Hz}

(a factor of 10 longer than current best)

GOAL: Ts > 240 seconds for {S4 = 0.8, 0 = 0.8 sec., C/N0 = 43 dB-Hz}

(a factor of 10 longer than current best)

UT Radionavigation Laboratory

Traditional Approach to Carrier Modeling

UT Radionavigation Laboratory

A New Approach to Carrier Modeling

UT Radionavigation Laboratory

A Multiple-Model Approach to Data Bit Estimation

UT Radionavigation Laboratory

The GPS Assimilator

The GPS Assimilator modernizes and makes existing GPS equipment resistant to jamming, spoofing, and scintillation without requiring

hardware or software changes to the equipment

The GPS Assimilator modernizes and makes existing GPS equipment resistant to jamming, spoofing, and scintillation without requiring

hardware or software changes to the equipment

A Backward-Compatible Way to Harden Existing UE Against Scintillation

UT Radionavigation Laboratory

All digital signal processing implemented in C++ on a high-end DSP Marginal computational demands:

Tracking: ~1.2% of DSP per channel Simulation: ~4% of DSP per channel

Full capability: 12 L1 C/A & 10 L2C tracking channels 8 L1 C/A simulation channels 1 Hz navigation solution Acquisition in background

GPS Assimilator Prototype

UT Radionavigation Laboratory

Summary

Models of scintillation effects on phase tracking loops must faithfully capture deep fades

The mean time between differentially-detected navigation bit errors is a good lumped

indicator of scintillation severity The triple accurately

predicts For carrier tracking, scintillation

modeling & simulation can be boiled down to two parameters: S4 & τ0

A hardware-in-the-loop scintillation testbed has been built and validated

Carrier tracking techniques inspired by the proposed model promises to extend

UT Radionavigation Laboratory

Acknowledgements

CASES sensor development funded by STTR grant through AFOSR via ASTRA LLC

Adaptation of CASES sensor for Antarctic deployment funded by ASTRA LLC

UT Radionavigation Laboratory

Model: Link Cycle Slips to Differentially-Detected Bit Errors

UT Radionavigation Laboratory

Amplitude Distribution:

Rice distribution applies

p(|z(t)|) can be summarized by the S4 index

Rice distribution applies

p(|z(t)|) can be summarized by the S4 index

UT Radionavigation Laboratory

Autocorrelation Function:Empirical Spectrum vs. Models

2nd-order Butterworth autocorrelation model applies

R() can be summarized by 0

2nd-order Butterworth autocorrelation model applies

R() can be summarized by 0