configs

Example Configuration Files

This directory contains example GNSS degradation configuration files for use with strapdown-sim. These scenarios demonstrate the various GNSS denial and degradation modes supported by the simulation.

Configuration File Format

All configurations use YAML format and contain three main sections:

• scheduler: Controls when GNSS measurements are available
• fault: Controls how GNSS measurements are corrupted
• seed: Random seed for reproducibility

See the User Guide for detailed documentation.

---

Available Scenarios

Baseline (No Degradation)

File: baseline.yaml

No GNSS degradation - all measurements pass through unchanged. Use this as a reference for comparing degraded scenarios.

strapdown-sim -i input.csv -o baseline.csv closed-loop --config baseline.yaml

---

Degraded Accuracy Scenarios

Full Rate with Degraded Accuracy

File: degraded_fullrate.yaml

GNSS measurements at full rate but with AR(1)-correlated position and velocity errors. Models low-SNR or multipath conditions.

Reduced Rate with Degraded Accuracy

File: degraded_5s.yaml

GNSS measurements every 5 seconds with degraded accuracy. Combines scheduling and fault injection.

---

Scheduling Scenarios

Fixed Interval

File: sched_10s.yaml

GNSS updates every 10 seconds with no measurement corruption. Models reduced update rate.

Duty Cycle

File: duty_10on_2off.yaml

Alternates between 10 seconds of GNSS availability and 2 seconds of denial. Models periodic outages.

---

Spoofing Scenarios

Slow Bias (Soft Spoofing)

File: slowbias.yaml

Slowly drifting position bias that appears plausible to the filter. Simulates gradual trajectory manipulation.

File: slowbias_rot.yaml

Same as above but with rotating drift direction.

Hijack (Hard Spoofing)

File: hijack.yaml

Abrupt position offset during a fixed time window. Simulates hard spoofing attack.

---

Combined Scenarios

Combo: Reduced Rate + Degraded

File: combo.yaml

Combines reduced update rate (5s) with degraded measurement accuracy.

Combo: Duty Cycle + Hijack

File: combo_duty_hijack.yaml

Combines duty-cycled availability with hard spoofing.

---

Usage Examples

Command Line

# Using a config file
strapdown-sim -i data/input.csv -o results/output.csv closed-loop \
  --config examples/configs/degraded_5s.yaml

# Override seed for different realization
strapdown-sim -i data/input.csv -o results/output.csv closed-loop \
  --config examples/configs/degraded_5s.yaml \
  --seed 123

Batch Processing

#!/bin/bash
# Run all scenarios
for config in examples/configs/*.yaml; do
  name=$(basename "$config" .yaml)
  strapdown-sim -i data/input.csv -o "results/${name}.csv" closed-loop \
    --config "$config"
done

---

Alternative Formats

The same configurations are also available in JSON and TOML formats:

• gnss_degradation.json - JSON format example
• gnss_degradation.toml - TOML format example
• json/ subdirectory - Additional JSON examples

---

Creating Custom Scenarios

To create a custom scenario:

1. Copy an existing configuration as a starting point
2. Modify the scheduler and/or fault sections
3. Save with a .yaml, .json, or .toml extension

Scheduler Options

# All measurements pass through
scheduler:
  kind: passthrough

# Fixed interval updates
scheduler:
  kind: fixed_interval
  interval_s: 10.0
  phase_s: 0.0

# Duty cycle (ON/OFF periods)
scheduler:
  kind: duty
  on_s: 30.0
  off_s: 60.0
  phase_s: 0.0

Fault Model Options

# No corruption
fault:
  kind: none

# AR(1) correlated noise
fault:
  kind: degraded
  rho_pos: 0.99
  sigma_pos_m: 3.0
  rho_vel: 0.95
  sigma_vel_mps: 0.3
  r_scale: 5.0

# Slow drifting bias
fault:
  kind: slowbias
  drift_n_mps: 0.02
  drift_e_mps: 0.0
  q_bias: 1e-6
  rotate_omega_rps: 0.0

# Hard spoofing window
fault:
  kind: hijack
  offset_n_m: 50.0
  offset_e_m: 0.0
  start_s: 120.0
  duration_s: 60.0

---

Reproducibility Notes

• Always specify the seed value for reproducible results
• The same seed with the same configuration will produce identical results
• Use different seeds to generate multiple realizations for Monte Carlo analysis