Open Loop Software receiver applied to GPS signals under strong fading and multipath conditions

1
Open Loop Software receiver applied to GPS
signals under strong fading and multipath
conditions

• Josep M. Aparicio
• IEEC

2
Project Goals

• To develop a GPS passive radar
• Particularly important in oceanography
• Sea level altimetry
• Sea surface roughness
• Some concepts applicable to limb sounding

3
Description

• A clean propagation environment leads to one
  single image of the signal at the receiver
  (delayed and Doppler shifted).
• In some cases propagation is multiple Several
  images received, with different delays and
  shifts
• Reflection off the sea.
• Propagation through turbulent/moist troposphere.
• Images may interfere fading

4
Open Loop mode

• Closed Loops Measurement matched to predictive
  model
• The model helps interpret the observations
• Open Loops Measurement is unpredictable
• Standard Closed Loop receiver Searches and
  follows the correlation peak, without paying
  attention to peaks substructure.

5
Our Open Loop Receiver

• GPS antenna
• L1-tuned radio receiver
• High-speed digitizerrecorder
• (Std receiver This step is a PLL tracker)
• We usually use more than one system
• Antennae pointed upwards/downwards
• RHCP, LHCP

6
Processing

• Input EM field at the antenna
• Search for C/A contents (i.e. Correlation)
• Satellite separation
• For each satellite
• Delay-Doppler analysis
• Amplitude phase in the delay-Doppler space
• Identification of main components
• Substructure of components

7
Multipath

• Reflection
• In our case we used signal polarization to
  distinguish both, and proceeded to separate
  substructure in the reflected signal.

8
Example Distance between both peaks

• Reflection seen from an aircraft
• D2h sin(elev) ?D

9
Time-frequency analysis (I)

• Do I see what I expect?
• I expect a direct signal with certain predictable
  Doppler and delay
• If the sea was very calm I could also expect a
  reflection with a slightly different Doppler and
  a larger, but still predictable delay
• Is that what I see?

10
Time-frequency analysis (II)

• X Time (1 sec)
• Y Frequency (-200 to 200 Hz)

11
Time-frequency analysis (III)

• Another reflected sample (2 sec).
• PLL impossible
• Statistical products
• Doppler spread
• Main Doppler rate
• Granularity
• Crossings

12
Conclusion (I)

• Open Loop procedures are able to extract data
  when PLL fails
• Target Statistical info rather than phase
  reconnection
• There is few statistical info when PLL works
• If PLL works no need for open loop
• If PLL doesnt, there is no single phase to
  recover-gt focus on stats

13
Conclusion (II)

• Open Loop procedures are expensive in data volume
  and CPU.
• But in limb sounding, only few sec.
• Products are only statistical.
• Phenomenologically all multipath problems are
  similar
• Each one (sea surface reflection, limb sounding,
  etc) requires a map between stats and geophysical
  info.