TightlyCoupled Opportunistic Navigation for Deep Urban and Indoor
Tightly-Coupled Opportunistic Navigation for Deep Urban and Indoor Positioning Ken Pesyna, Zak Kassas, Jahshan Bhatti, and Todd Humphreys Presentation at ION 2011|September 23, 2011
Outline I. Motivate & define Tightly-Coupled Opportunistic Navigation (TCON) II. Explore “signals of opportunity” for TCON III. Discuss the central estimator to optimally fuse signals together IV. Present experimental results
Goal Optimally extract navigation and timing information from ambient radio signals Tightly-Coupled Opportunistic Navigation is a framework to achieve this goal
Tightly-Coupled Opportunistic Navigation § TCON is an optimal generalization of specific hybrid navigation technologies TCON § Generalize § Optimize Cellular [1] GNSS Iridium [3] [1] [2] [3] HDTV [2] R. Rowe, P. Duffett-Smith, et. al. , “Enhanced GPS: The tight integration of received cellular timing signals and GNSS receivers for ubiquitous positioning, ” in Position, Location, and Navigation Symposium, IEEE/ION, 2008. J. Do, M. Rabinowitz, and P. Enge, “Performance of hybrid positioning system combining GPS and television signals, ” in Position, Location, And Navigation Symposium, IEEE/ION, 2006. M. Joerger, et. al. , “Iridium/gps carrier phase positioning and fault detection over wide areas, ” ION GNSS 2009.
TCON: Breaking it Down Tightly-Coupled Ø Signals downmixed and sampled with the same clock Ø Absolute time correspondence at the nanosecond level Opportunistic Navigation Ø Receiver continuously searches for signals from which to extract navigation and timing information Ø Receiver employs on-the-fly signal characterization: Ø Clock stability Ø Clock offset Ø Carrier-to-noise ratio Ø Transmitter location
Tightly-Coupled Opportunistic Navigation
Signals of Opportunity § TCON treats all RF signals as potential signals of opportunity Ø GNSS signals: GPS, Galileo, Glonass Ø Cellular signals: CDMA, GSM, 4 G LTE, & Wi. MAX Ø Other satellite signals: Iridium Ø Other ground-based signals: HDTV, Wi-Fi
What SOP characteristics are desirable? 1. High received carrier-to- Unfortunately, we almost never get all three noise ratio properties if we are not working with dedicated Good frequencynavigation stability signals 2. 3. Known location/timing offset
Freestyle Navigation § CDMA Cellular Ø Carrier-to-noise is high, penetrates well Ø Towers do not move Ø Only roughly synchronized to GPS Ø Carrier stability varies from provider to provider § Iridium Ø Carrier-to-noise ratio is higher than GNSS Ø Global coverage Ø Not continuous – “Bursty” TDMA structure Ø Ambiguous in carrier phase from burst to burst
Centralized Estimator
Centralized Estimator § Optimally combines observables from all § § SOPs Our preferred implementation is an extended Kalman filter: Base State: SOP state: § Full State:
Carrier Phase Measurement Models § GPS carrier phase measurement model: § CDMA carrier phase measurement model: § Iridium carrier phase measurement model:
Dynamics Model § Dynamics dependent on current state and process noise: § Process noise covariance formed by models of clock dynamics:
Adaptive Dynamics Model Output * Update SOP Clock Model Run Kalman Filter old Iterate * Can only be done when is observable (to within a constant offset) new
Simple TCON Demo: Experiment Setup National Instruments RFSA Storage GRID Software Receiver MATLAB EKF
Simple TCON Demo: SOP Wardriving
A Simple Demonstration of TCON (2) Characterizing CDMA SOP (1) Obtaining Truth Update SOP Clock Model Run Kalman Filter 6 x old new (3) Demonstrate postcharacterization CDMA SOP only estimate of 2 x Iterate Truth Estimate CDMA-based estimate vs.
So What? § Achieve our estimate error of § § § by differencing the CDMA-only estimate from the truth variations: The stability of will affect the length of achievable GNSS coherent integration in weak signal conditions In this experiment, CDMA-only TCON (after-characterization) could supply coherent integration times beyond 100 seconds 100+ sec. coherent integration allows GNSS acquisition of signals below 0 d. B-Hz assuming all else ideal Mean squared coherence of
Conclusions § TCON is a framework to optimally extract § § navigation and timing information from ambient radio signals Tight-coupling at the carrier-phase level and SOP characterization are essential to TCON A simple TCON demonstration on timing showed a TCON-enabled receiver can coherently integrate beyond 100 seconds using characterized CDMA signals
Questions?
- Slides: 21