Analyze Rf Signal

Top-Level RF Signal Analysis & Orchestration Skill 🛰️

Use this skill as the primary entry point whenever the user requests an end-to-end analysis of an unknown complex IQ signal dataset (.sigmf-data, .sigmf-meta, or .cu8 file) or asks to "analyse this RF signal".

Triggering Context

Activate this skill whenever the user asks to:

• "analyse this RF signal"
• "analyze this RF signal"
• "analyze RF capture <file>"
• "full pipeline signal analysis"
• "end-to-end signal analysis"

Interactive 4-Stage Workflow

 ┌─────────────────────────────────────────────────────────────────────────────┐
 │ 1. Spectrum Scanning & Channel Discovery                                   │
 │    scan_wideband_channels() -> List candidate channels & local SNR         │
 └──────────────────────┬──────────────────────────────────────────────────────┘
                        │
 ┌──────────────────────▼──────────────────────────────────────────────────────┐
 │ 2. Interactive User Channel Selection                                      │
 │    Present candidate table -> Prompt user for Channel Index / Offset        │
 └──────────────────────┬──────────────────────────────────────────────────────┘
                        │
 ┌──────────────────────▼──────────────────────────────────────────────────────┐
 │ 3. Automated End-to-End DSP Pipeline                                        │
 │    DDC Extraction -> signal-analysis -> signal-cleanup                     │
 │    -> modulation-recognition -> signal-synchronization -> demodulation       │
 └──────────────────────┬──────────────────────────────────────────────────────┘
                        │
 ┌──────────────────────▼──────────────────────────────────────────────────────┐
 │ 4. Standalone Flowgraph Receiver Generation                                │
 │    build-gnuradio-flowgraph -> Generate executable Python top_block       │
 └─────────────────────────────────────────────────────────────────────────────┘

Stage 1: Wideband Spectrum Scanning & Channel Discovery

Scan the wideband spectrum to detect and list active signal channel peaks:

PYTHONPATH=/usr/lib/python3/dist-packages:./ .agents/skills/analyze-rf-signal/scripts/analyze_rf_pipeline.py \
  --input <path_to_sigmf_file> \
  --scan_only

Stage 2: Interactive User Channel Selection

Display candidate channel table (Rank, Frequency Offset, Absolute RF Frequency, Peak Power, Local SNR, Bandwidth, Channel Type) and ask the user to select which candidate channel index (--channel_index N) or manual offset (--manual_offset OFFSET_HZ) to process.

Stage 3: Target Channel Extraction & Automated DSP Pipeline

Execute the full DSP analysis chain on the chosen channel:

PYTHONPATH=/usr/lib/python3/dist-packages:./ .agents/skills/analyze-rf-signal/scripts/analyze_rf_pipeline.py \
  --input <path_to_sigmf_file> \
  --channel_index <selected_index> \
  --output_dir /tmp

Under the Hood Skill Execution:

1. Target Channel Extraction (rtlsdr-hardware-capture / DDC): Digital downconversion to baseband.
2. Spectral Inspection (signal-analysis): Welch PSD, SNR estimation, occupied bandwidth.
3. Signal Cleanup (signal-cleanup): DC offset suppression, Gram-Schmidt I/Q balancing, bandpass filtering, AGC normalization.
4. Modulation Recognition (modulation-recognition): Automatic classification across AM, FM, ASK, BPSK, QPSK, 8PSK, 16QAM, 64QAM, GFSK using higher-order cumulants and constellation variance.
5. Carrier & Clock Synchronization (signal-synchronization): CFO estimation, Costas Loop phase lock, Symbol Sync clock recovery (for digital schemes).
6. Payload Demodulation (signal-demodulation): Audio WAV extraction (analog AM/FM) or bitstream / PWM-OOK byte payload decoding (digital schemes).

Stage 4: Executable Top Block Receiver Generation

Generate standalone GNU Radio Python receiver script (build-gnuradio-flowgraph skill):

PYTHONPATH=/usr/lib/python3/dist-packages:./ .agents/skills/build-gnuradio-flowgraph/scripts/build_gnuradio_flowgraph.py \
  --input /tmp/extracted_channel.sigmf-data \
  --output_script /tmp/receiver_top_block.py