A vectorized pybind11 interface for liquid-dsp.
Uses cmake and scikit build system to provide a python extension module liquiddsp for python 3. To build one needs
- liquid-dsp (brew install)
- pybind11 (pip install)
- numpy (pip install)
Follow the typical steps,
git clone https://github.com/colbyAtCRI/python-liquiddsp.git
cd python-liquiddsp
do
pip install .
I've written several python interfaces for SDRs
- Rfspace's SDRIQ
- SDRPlay RSP2 and RSPduo (all are supported but I've only tested what I have)
- RTLSDRs
Each of these use python callbacks of the form of a class with
the __call__ member defined. A typical example of an AM radio reciever
is
import numpy as np
from liquiddsp import ComplexResampler, ComplexIIRFilter, AmpModem, AGC, DeemphasisFilter
from sdrplay import Radio
class AMRadio:
def __init__(self, bandwidth = 15000, iq_rate = 2000000, pcm_rate = 48000):
self.bandpass = ComplexIIRFilter (filter_type='cheby2', order = 8, Fc = bandwidth/iq_rate)
self.resample = ComplexRampler (rate = pcm_rate/iq_rate, Fc = pcm_rate/iq_rate)
self.am = AmpModem (modulation = 0.5, type = 'dsb', carrier = True)
self.audio_filter = DeemphasisFilter (pcm_rate)
self.agc = AGC()
self.agc.lock = False
self.agc.scale = 0.01
self.pcm = b''
def __call__(self, iq):
pcm = self.audio_filter ( self.am ( self.agc ( self.resample ( self.bandpass ((iq))))))
self.pcm = self.pcm + pcm.tobytes()
while len(self.pcm) > 4096:
# write self.pcm[0:4096] to pyaudio stream
self.pcm = self.pcm[4096:]
radio = Radio ()
radio.freq = 1170000
radio.onIQData = AMRadio()
radio.running = True
