A fast and simple CSI decoder written in Python.

You can:

• Plot CSI samples from .pcap files generated
• Read CSI samples to use in your Python programs

1. Install dependencies: pip install numpy matplotlib
2. Open the config.py file and select your WiFi chip. Default is bcm43455c0 (Raspberry Pi)
3. Copy your pcap files to the pcapfiles folder. This folder can be changed in the config file.
4. Run csi-explorer: python3 csiexplorer.py

Enter the pcap's filename. Typing the .pcap part is optional. Type an index to show it's plot. Type indexes separated by - to play animation (example: 10-20).

Type help to see all options. Null-subcarriers are hidden by default, you can change this in the config file.

You can integrate csi-explorer into your python programs. Here is an example for the 'Interleaved' decoder for bcm43455c0 and bcm4339,

from decoders import interleaved as decoder

samples = decoder.read_pcap('pcapfiles/example-80.pcap')

samples.csi # Access all CSI samples as a numpy matrix

# Get the 200th sample, but remove Null
# and Pilot OFDM subcarriers.
csi = samples.get_csi(
    index=200,
    rm_nulls=True,
    rm_pilots=True
)

mac = samples.get_mac(index=200)    # Source Mac ID
sc, fn = samples.get_seq(index=200) # Sequence and fragment number
css = samples.get_css(index=200)    # Core and Spatial Stream

samples.bandwidth # Bandwidth, automatically inferred from the pcap file
samples.nsamples  # Number of samples

# Print info about the 200th sample
samples.print(200)

• @Gi-z - Glenn Forbes
• @zeroby0 - Aravind Voggu
• @tweigel-dev - Thomas Weigel