A **fast** channel state information parser for Intel, Atheros, Nexmon, ESP32 and PicoScenes in Python.
Project description
csiread
A fast channel state information parser for Intel, Atheros, Nexmon, ESP32 and PicoScenes in Python.
- Full support for Linux 802.11n CSI Tool, Atheros CSI Tool, nexmon_csi and ESP32-CSI-Tool
- Support for PicoScenes is experimental.
- At least 15 times faster than the implementation in Matlab
- Real-time parsing and visualization.
Introduction
Various CSI Tools only provide Matlab API parsing CSI data files. Those who want to process CSI with Python have to install Matlab to convert .dat
to .mat
. This process is redundant and inefficient. Therefore, Python API is recommended. Unfortunately, the API implemented in pure Python is inefficient. With this in mind, I implemented csiread in Cython(Pybind11 may be another great choice). The table below shows the performance of different implementations. They were all tested with 40k packets on the same computer.
Function | Matlab | Python3+Numpy | csiread | file size |
---|---|---|---|---|
Nexmon.read:bcm4339 | 3.2309s | 0.2739s | 0.0703s | 44.0MB |
Nexmon.read:bcm4358 | 3.5987s | 23.0025s | 0.1227s | 44.0MB |
Atheros.read | 3.3081s | 14.6021s | 0.0956s | 76.3MB |
Intel.read | 1.6102s | 7.6624s | 0.0479s | 21.0MB |
Intel.get_total_rss | 0.1786s | 0.0030s | 0.0030s | |
Intel.get_scaled_csi | 0.5497s | 0.1225s | 0.0376s/0.0278s | |
Intel.get_scaled_csi_sm | 5.0097s | 0.3627s | 0.0778s/0.0465s |
This tool is not only the translation of the Matlab API, but also a CSI toolbox. I added some utilities, real-time visualization and algorithms code in the examples
folder. These would be useful for Python-based CSI researchers.
Install
pip3 install csiread
Quickstart
import csiread
# Linux 802.11n CSI Tool
csifile = "../material/5300/dataset/sample_0x1_ap.dat"
csidata = csiread.Intel(csifile, nrxnum=3, ntxnum=2, pl_size=10)
csidata.read()
csi = csidata.get_scaled_csi()
print(csidata.csi.shape)
# Atheros CSI Tool
csifile = "../material/atheros/dataset/ath_csi_1.dat"
csidata = csiread.Atheros(csifile, nrxnum=3, ntxnum=2, pl_size=10, tones=56)
csidata.read(endian='little')
print(csidata.csi.shape)
# nexmon_csi
csifile = "../material/nexmon/dataset/example.pcap"
csidata = csiread.Nexmon(csifile, chip='4358', bw=80)
csidata.read()
print(csidata.csi.shape)
# ESP32-CSI-Tool
csifile = "../material/esp32/dataset/example_csi.csv"
csidata = csiread.ESP32(csifile, csi_only=True)
csidata.read()
print(csidata.csi.shape)
# PicoScenes
csifile = "../material/picoscenes/dataset/rx_by_iwl5300.csi"
csidata = csiread.Picoscenes(csifile, {'CSI': [30, 3, 2], 'MPDU': 1522})
csidata.read()
csidata.check()
print(csidata.raw['CSI']['CSI'].shape)
examples
are the best usage instructions. The API documentation can be found in docstring
of file core.py
, so we won't repeat them here.
Build from source
cd csiread
pip3 install -r requirements.txt
python3 setup.py sdist bdist_wheel
pip3 install -U dist/csiread*.whl
*
is a shell wildcard. After running python3 setup.py sdist bdist_wheel
,there will be a wheel file like csiread-1.3.4-cp36-cp36m-win_amd64.whl
in the dist
folder. Replace csiread*.whl
with it.
csiread is written in Cython, Cython requires a C compiler to be present on the system. You can refer to Installing Cython for more details. If you don't want to install a C compiler, just fork the project and push a tag to the latest commit. Then wheel files can be found in Github-Actions-Python package-Artifacts: csiread_dist
Design
csiread provides 7 classes: Intel, Atheros, Nexmon, AtherosPull10, NexmonPull46, ESP32 and Picoscenes
. Each class has 4 key methods: read(), seek()
, pmsg()
and display()
which are used for reading a file, reading a file from a specific position, real-time parsing and viewing the contents of a packet respectively. csiread.utils
provides some common functions.
Nexmon CSI
csiread.Nexmon
is based on the commit of nexmon_csi(Aug 29, 2020):ba99ce12a6a42d7e4ec75e6f8ace8f610ed2eb60
csiread.NexmonPull256
is the same ascsiread.NexmonPull46
. It works with the latest master branch (Dec 11, 2021):c037576b7035619e2716229c7622f4e8c511635f
- The
Nexmon.group
is experimental, it may be incorrect due tocore
andspatial
.core
andspatial
are ZERO or not recorded correctly in some files. I don't know how to solve it.
ESP32-CSI-Tool
pandas.read_csv
andcsiread.ESP32
have the similar performance, butpandas.read_csv
is much more flexible.
PicoScenes
The support for Picoscenes is an experimental feature. PicoScenes is still under active development, csiread cannot be updated synchronously.
csidata.raw
is a structured array in numpy and stores the parsed result.Mag
andPhase
fileds have been removed, usenp.abs
andnp.angle
instead.- Call
check()
method afterread()
, Then setpl_size
according to the report. - Edge padding are applied to
raw["xxx"]["SubcarrierIndex"]
for plotting. - The method
pmsg
has been implemented, but not yet ready. - Accessing CSI like
csidata.CSI.CSI
is only available after callingread
method. - 5-10 times faster than before
parseCSIMVM(...)
in_picoscenes.pyx
may be incorrect.
csiread.Picoscenes
is based on the PicoScenes MATLAB Toolbox(PMT)(Last modified at 2022-01-21).
Project details
Download files
Download the file for your platform. If you're not sure which to choose, learn more about installing packages.
Source Distribution
Built Distributions
File details
Details for the file csiread-1.4.1.tar.gz
.
File metadata
- Download URL: csiread-1.4.1.tar.gz
- Upload date:
- Size: 168.0 kB
- Tags: Source
- Uploaded using Trusted Publishing? No
- Uploaded via: twine/4.0.2 CPython/3.11.6
File hashes
Algorithm | Hash digest | |
---|---|---|
SHA256 | 5159c81d160c4fa581acc58944bf29d34d712f40721469125a8573bd0f5bbcb7 |
|
MD5 | a550582cd2a03abf8c31eaf8979f9a84 |
|
BLAKE2b-256 | 084c94f9d0d788548e6fffbed0491ea7dbc1e83864491e37101c1a2d7cdc9ef4 |
File details
Details for the file csiread-1.4.1-cp313-cp313-win_amd64.whl
.
File metadata
- Download URL: csiread-1.4.1-cp313-cp313-win_amd64.whl
- Upload date:
- Size: 268.1 kB
- Tags: CPython 3.13, Windows x86-64
- Uploaded using Trusted Publishing? No
- Uploaded via: twine/4.0.2 CPython/3.11.6
File hashes
Algorithm | Hash digest | |
---|---|---|
SHA256 | 45b06eb453725d5122347451e1e2150b90583c63ee003540abf54465983a3791 |
|
MD5 | 7b5e6a2ca8c3e8b0741487f0db8a09b9 |
|
BLAKE2b-256 | 26a4a1eb613b51dd1205f6de58b48ae73b819c780fbada80462e72a16c857258 |
File details
Details for the file csiread-1.4.1-cp313-cp313-manylinux_2_17_x86_64.manylinux2014_x86_64.whl
.
File metadata
- Download URL: csiread-1.4.1-cp313-cp313-manylinux_2_17_x86_64.manylinux2014_x86_64.whl
- Upload date:
- Size: 363.6 kB
- Tags: CPython 3.13, manylinux: glibc 2.17+ x86-64
- Uploaded using Trusted Publishing? No
- Uploaded via: twine/4.0.2 CPython/3.11.6
File hashes
Algorithm | Hash digest | |
---|---|---|
SHA256 | 66d709f51818c50ef35aff89728bc28b7ea30b123ffe071aca9d0d4be3101d35 |
|
MD5 | b6c821d21d002d38774fd4c1a12fa87a |
|
BLAKE2b-256 | c345add1fa6aa4f93f06e14b5f03faad16b40bf47c756684f3cadb9398fc410f |
File details
Details for the file csiread-1.4.1-cp313-cp313-macosx_10_13_universal2.whl
.
File metadata
- Download URL: csiread-1.4.1-cp313-cp313-macosx_10_13_universal2.whl
- Upload date:
- Size: 554.0 kB
- Tags: CPython 3.13, macOS 10.13+ universal2 (ARM64, x86-64)
- Uploaded using Trusted Publishing? No
- Uploaded via: twine/4.0.2 CPython/3.11.6
File hashes
Algorithm | Hash digest | |
---|---|---|
SHA256 | 554937422a91a1fc78aaad220f2816ab821508e23ff47282ac3c4fb72a058f49 |
|
MD5 | 0e97083d2a363a768e9ab829a4be7b7d |
|
BLAKE2b-256 | 52f7a0efc6b1ad52d638b4b317a903422a568466287bee16abd779bad001150d |
File details
Details for the file csiread-1.4.1-cp312-cp312-win_amd64.whl
.
File metadata
- Download URL: csiread-1.4.1-cp312-cp312-win_amd64.whl
- Upload date:
- Size: 268.5 kB
- Tags: CPython 3.12, Windows x86-64
- Uploaded using Trusted Publishing? No
- Uploaded via: twine/4.0.2 CPython/3.11.6
File hashes
Algorithm | Hash digest | |
---|---|---|
SHA256 | 87036d52613b673ba0b8ea53910b1033af9a873f0d853a4d0f97036cfe9bafaf |
|
MD5 | 2a7aea6c14c645e039b299701765e909 |
|
BLAKE2b-256 | ad3be36b8754129be81f669b59312696d49e8d6bb351c5f49e66409f2c0a4a95 |
File details
Details for the file csiread-1.4.1-cp312-cp312-manylinux_2_17_x86_64.manylinux2014_x86_64.whl
.
File metadata
- Download URL: csiread-1.4.1-cp312-cp312-manylinux_2_17_x86_64.manylinux2014_x86_64.whl
- Upload date:
- Size: 364.7 kB
- Tags: CPython 3.12, manylinux: glibc 2.17+ x86-64
- Uploaded using Trusted Publishing? No
- Uploaded via: twine/4.0.2 CPython/3.11.6
File hashes
Algorithm | Hash digest | |
---|---|---|
SHA256 | 21b71d571c4d706bd6d9210be3ec9a3dad901458e81b714c05b29d8eb8a83ba5 |
|
MD5 | 9c6ddc0922c5767858b3262c044dd183 |
|
BLAKE2b-256 | 8c259f100c9ce9bb6c8674500a10e9dbe604bf9adc00ea1ab932802522239c70 |
File details
Details for the file csiread-1.4.1-cp312-cp312-macosx_10_13_universal2.whl
.
File metadata
- Download URL: csiread-1.4.1-cp312-cp312-macosx_10_13_universal2.whl
- Upload date:
- Size: 559.2 kB
- Tags: CPython 3.12, macOS 10.13+ universal2 (ARM64, x86-64)
- Uploaded using Trusted Publishing? No
- Uploaded via: twine/4.0.2 CPython/3.11.6
File hashes
Algorithm | Hash digest | |
---|---|---|
SHA256 | bc9734630c174fb634c4edc4bf2daf06e48fb97baf0db09360d0a42245b5d6dc |
|
MD5 | c4e7870da32c222f220958b30e9db5a6 |
|
BLAKE2b-256 | c98ac640daaacb2a933a9680afe6e45cabaac386bb7ec29a5f406c8961eaea50 |
File details
Details for the file csiread-1.4.1-cp311-cp311-win_amd64.whl
.
File metadata
- Download URL: csiread-1.4.1-cp311-cp311-win_amd64.whl
- Upload date:
- Size: 273.8 kB
- Tags: CPython 3.11, Windows x86-64
- Uploaded using Trusted Publishing? No
- Uploaded via: twine/4.0.2 CPython/3.11.6
File hashes
Algorithm | Hash digest | |
---|---|---|
SHA256 | b7a8ab285f39ab9b7d32afcb52fc4dbef4b68540ae1e14edc9ad1d2fa9b7a94c |
|
MD5 | 45082b1a1f40ed8f47e967d466d4b116 |
|
BLAKE2b-256 | 474f3694008acfd31b7b6d82cc47864d6ecdfc05607cf5911425b2d865e6266c |
File details
Details for the file csiread-1.4.1-cp311-cp311-manylinux_2_17_x86_64.manylinux2014_x86_64.whl
.
File metadata
- Download URL: csiread-1.4.1-cp311-cp311-manylinux_2_17_x86_64.manylinux2014_x86_64.whl
- Upload date:
- Size: 378.8 kB
- Tags: CPython 3.11, manylinux: glibc 2.17+ x86-64
- Uploaded using Trusted Publishing? No
- Uploaded via: twine/4.0.2 CPython/3.11.6
File hashes
Algorithm | Hash digest | |
---|---|---|
SHA256 | 0499946134e703b422c4d9cc51f998338dfbf544a2b73ea96c8deef95e5e458b |
|
MD5 | 3b2d0e9af838ee9c314378b407c0eb4e |
|
BLAKE2b-256 | d14e30c282a732ae7c144887465600643938bb1b9ebe852495d1a629175f1078 |
File details
Details for the file csiread-1.4.1-cp311-cp311-macosx_10_9_universal2.whl
.
File metadata
- Download URL: csiread-1.4.1-cp311-cp311-macosx_10_9_universal2.whl
- Upload date:
- Size: 560.2 kB
- Tags: CPython 3.11, macOS 10.9+ universal2 (ARM64, x86-64)
- Uploaded using Trusted Publishing? No
- Uploaded via: twine/4.0.2 CPython/3.11.6
File hashes
Algorithm | Hash digest | |
---|---|---|
SHA256 | 5247d2c21b3f6eb15e36b2e57b308116b0c9991ac765ca5040085f889c7e378b |
|
MD5 | 2ad21d9358832a2c1143d6234d0013ca |
|
BLAKE2b-256 | f716b7ab650fe1de554a5f49081ae439634b1aaadfcc4db6bd82ce4be8e4d02d |
File details
Details for the file csiread-1.4.1-cp310-cp310-win_amd64.whl
.
File metadata
- Download URL: csiread-1.4.1-cp310-cp310-win_amd64.whl
- Upload date:
- Size: 272.7 kB
- Tags: CPython 3.10, Windows x86-64
- Uploaded using Trusted Publishing? No
- Uploaded via: twine/4.0.2 CPython/3.11.6
File hashes
Algorithm | Hash digest | |
---|---|---|
SHA256 | 603234077183ceba592a3176c9a75c0758f961688d9362eab12f07af095cdc64 |
|
MD5 | 60233aa10a1aea78826fb59a6c6282a0 |
|
BLAKE2b-256 | 6db14e028b526d3f926e8a50fe0ac4bc86a75f905cdf00999729176fac68ff5e |
File details
Details for the file csiread-1.4.1-cp310-cp310-manylinux_2_17_x86_64.manylinux2014_x86_64.whl
.
File metadata
- Download URL: csiread-1.4.1-cp310-cp310-manylinux_2_17_x86_64.manylinux2014_x86_64.whl
- Upload date:
- Size: 378.3 kB
- Tags: CPython 3.10, manylinux: glibc 2.17+ x86-64
- Uploaded using Trusted Publishing? No
- Uploaded via: twine/4.0.2 CPython/3.11.6
File hashes
Algorithm | Hash digest | |
---|---|---|
SHA256 | 1cc09bc47f2126ba46dff84c419f59037b0aaebc76e2486599032a6a59866178 |
|
MD5 | 5c87749736a2f109a397adbb237eead7 |
|
BLAKE2b-256 | 862ba9a7657617d20c4699f17b704c75cde36d3b12122a29949411523415c94f |
File details
Details for the file csiread-1.4.1-cp310-cp310-macosx_10_9_universal2.whl
.
File metadata
- Download URL: csiread-1.4.1-cp310-cp310-macosx_10_9_universal2.whl
- Upload date:
- Size: 559.1 kB
- Tags: CPython 3.10, macOS 10.9+ universal2 (ARM64, x86-64)
- Uploaded using Trusted Publishing? No
- Uploaded via: twine/4.0.2 CPython/3.11.6
File hashes
Algorithm | Hash digest | |
---|---|---|
SHA256 | 2915b2e5250bcd86a2ca6fb60f42d3c74e6a4d67a5f7aab4ae23d65741a81b91 |
|
MD5 | 480ea2a3f1874aa58ad3e49aba4ea9cc |
|
BLAKE2b-256 | fbdabee2cc062cd34fad42bd143d2e542b01241c09c4bb9a6e7dea700ab34126 |
File details
Details for the file csiread-1.4.1-cp39-cp39-win_amd64.whl
.
File metadata
- Download URL: csiread-1.4.1-cp39-cp39-win_amd64.whl
- Upload date:
- Size: 310.7 kB
- Tags: CPython 3.9, Windows x86-64
- Uploaded using Trusted Publishing? No
- Uploaded via: twine/4.0.2 CPython/3.11.6
File hashes
Algorithm | Hash digest | |
---|---|---|
SHA256 | e71402eb064f8376414bbff5a6858789e0b54845f844d9cf1b13ee1dff1441c8 |
|
MD5 | cfd37b876763f692f587a377f8b0f3f8 |
|
BLAKE2b-256 | bc94d749065d044211d302c2c6b3503c25a9bd0ba693c6f904c78ee4c3d73cad |
File details
Details for the file csiread-1.4.1-cp39-cp39-manylinux_2_17_x86_64.manylinux2014_x86_64.whl
.
File metadata
- Download URL: csiread-1.4.1-cp39-cp39-manylinux_2_17_x86_64.manylinux2014_x86_64.whl
- Upload date:
- Size: 379.2 kB
- Tags: CPython 3.9, manylinux: glibc 2.17+ x86-64
- Uploaded using Trusted Publishing? No
- Uploaded via: twine/4.0.2 CPython/3.11.6
File hashes
Algorithm | Hash digest | |
---|---|---|
SHA256 | 1a1dc1c1e40c74dfea518c5faaf02058274da724b228a004ca99d33c15ad9617 |
|
MD5 | d935297dc6e6076184ad3f4ff68ad15c |
|
BLAKE2b-256 | e0294d196455c665de327cd0a1f5e979c112ea851512311f8e3a4a0418a0a29f |
File details
Details for the file csiread-1.4.1-cp39-cp39-macosx_10_9_universal2.whl
.
File metadata
- Download URL: csiread-1.4.1-cp39-cp39-macosx_10_9_universal2.whl
- Upload date:
- Size: 561.2 kB
- Tags: CPython 3.9, macOS 10.9+ universal2 (ARM64, x86-64)
- Uploaded using Trusted Publishing? No
- Uploaded via: twine/4.0.2 CPython/3.11.6
File hashes
Algorithm | Hash digest | |
---|---|---|
SHA256 | f44bf15438097df2d7c0ff29f6833d30cc5e34ade128ad4f0fa1830b256f92f9 |
|
MD5 | fd4b5866cc118ed596e63e7420727edc |
|
BLAKE2b-256 | f08c7863e81e94a11c816a35f27292b5b58c6484c6244c4c1169590393199cb7 |
File details
Details for the file csiread-1.4.1-cp38-cp38-win_amd64.whl
.
File metadata
- Download URL: csiread-1.4.1-cp38-cp38-win_amd64.whl
- Upload date:
- Size: 311.8 kB
- Tags: CPython 3.8, Windows x86-64
- Uploaded using Trusted Publishing? No
- Uploaded via: twine/4.0.2 CPython/3.11.6
File hashes
Algorithm | Hash digest | |
---|---|---|
SHA256 | d32f2f97da0db030d467668bb9b2ea4094ea871eaec78d644ab5bc3852736e9f |
|
MD5 | cbb81ee702386e05d3686d5d7893be07 |
|
BLAKE2b-256 | fece291b7a140a744475b1c7dbe2075fe6e2b281fd90df98953927709a20e252 |
File details
Details for the file csiread-1.4.1-cp38-cp38-manylinux_2_17_x86_64.manylinux2014_x86_64.whl
.
File metadata
- Download URL: csiread-1.4.1-cp38-cp38-manylinux_2_17_x86_64.manylinux2014_x86_64.whl
- Upload date:
- Size: 384.6 kB
- Tags: CPython 3.8, manylinux: glibc 2.17+ x86-64
- Uploaded using Trusted Publishing? No
- Uploaded via: twine/4.0.2 CPython/3.11.6
File hashes
Algorithm | Hash digest | |
---|---|---|
SHA256 | d5d036096ef54b7140345fef8aa965d7690030dbd7adbd53463586dfb0b69160 |
|
MD5 | 4ce77935655ab75060462274355cbb98 |
|
BLAKE2b-256 | 7b9c0eaa12a6ca6dc33807a74680e8d70987197f8af1995cadf5071667d587d2 |
File details
Details for the file csiread-1.4.1-cp38-cp38-macosx_11_0_universal2.whl
.
File metadata
- Download URL: csiread-1.4.1-cp38-cp38-macosx_11_0_universal2.whl
- Upload date:
- Size: 558.2 kB
- Tags: CPython 3.8, macOS 11.0+ universal2 (ARM64, x86-64)
- Uploaded using Trusted Publishing? No
- Uploaded via: twine/4.0.2 CPython/3.11.6
File hashes
Algorithm | Hash digest | |
---|---|---|
SHA256 | fda9619c35111c646c6c4cfcf22732167b718736f067be6b199bb946b07e45c7 |
|
MD5 | 012220bc2ffe0bb3b1606449e30e7e72 |
|
BLAKE2b-256 | 42847f6fd30b158547e4088ee5c0461bc20a6de622b86275f84b6fe8776aff30 |