mirror of
https://gitlab.science.ru.nl/mthesis-edeboone/m-thesis-introduction.git
synced 2024-11-14 02:23:32 +01:00
243 lines
8.1 KiB
Python
243 lines
8.1 KiB
Python
|
#!/usr/bin/env python3
|
||
|
|
||
|
|
||
|
from lib import util
|
||
|
from scipy import signal, interpolate
|
||
|
import matplotlib.pyplot as plt
|
||
|
import numpy as np
|
||
|
|
||
|
rng = np.random.default_rng()
|
||
|
|
||
|
class Waveform:
|
||
|
name = None
|
||
|
time = None
|
||
|
signal = None
|
||
|
|
||
|
def __init__(signal, time=None, name=None, dt=None):
|
||
|
self.signal = signal
|
||
|
self.time = time
|
||
|
self.name = name
|
||
|
|
||
|
if self.time is None and dt is not None:
|
||
|
self.time = dt * len(signal)
|
||
|
|
||
|
def white_noise_realisation(N_samples, noise_sigma=1, rng=rng):
|
||
|
return rng.normal(0, noise_sigma or 0, size=N_samples)
|
||
|
|
||
|
def antenna_bp(trace, low_bp, high_bp, dt, order=3):
|
||
|
# order < 6 for stability
|
||
|
fs = 1/dt
|
||
|
nyq = 1* fs
|
||
|
low_bp = low_bp / nyq
|
||
|
high_bp = high_bp / nyq
|
||
|
|
||
|
bp_filter = signal.butter(order, [low_bp, high_bp], 'band', fs=fs)
|
||
|
bandpassed = signal.lfilter(*bp_filter, trace)
|
||
|
|
||
|
return bandpassed
|
||
|
|
||
|
def my_correlation(in1, template):
|
||
|
#
|
||
|
in1_long = np.zeros( (len(in1)+2*len(template)) )
|
||
|
in1_long[len(template):-len(template)] = in1
|
||
|
|
||
|
# fill the template with zeros and copy template
|
||
|
template_long = np.zeros_like(in1_long)
|
||
|
template_long[len(template):2*len(template)] = template
|
||
|
|
||
|
lags = np.arange(-len(template), len(in1) ) - len(template)
|
||
|
|
||
|
# do the correlation jig
|
||
|
corrs = np.zeros_like(lags, dtype=float)
|
||
|
for i, l in enumerate(lags):
|
||
|
lagged_template = np.roll(template_long, l)
|
||
|
corrs[i] = np.dot(lagged_template, in1_long)
|
||
|
|
||
|
return corrs, (in1_long, template_long, lags)
|
||
|
|
||
|
def trace_upsampler(template_signal, trace, template_t, trace_t):
|
||
|
template_dt = template_t[1] - template_t[0]
|
||
|
trace_dt = trace_t[1] - trace_t[0]
|
||
|
|
||
|
upsample_factor = trace_dt/template_dt
|
||
|
upsampled_trace_N = np.ceil(len(trace) * upsample_factor)
|
||
|
|
||
|
upsample_factor = int(upsample_factor)
|
||
|
upsampled_trace_N = int(upsampled_trace_N)
|
||
|
|
||
|
# upsample trace
|
||
|
upsampled_trace = np.zeros(upsampled_trace_N)
|
||
|
upsampled_trace[::upsample_factor] = trace
|
||
|
|
||
|
#upsampled_t = np.arange(trace_t[0], trace_t[-1], template_dt)
|
||
|
upsampled_t = template_dt * np.arange(len(upsampled_trace)) + trace_t[0]
|
||
|
|
||
|
return upsampled_trace, upsampled_t
|
||
|
|
||
|
def template_downsampler(template_signal, trace, template_t, trace_t, offset):
|
||
|
pass
|
||
|
|
||
|
|
||
|
if __name__ == "__main__":
|
||
|
import os
|
||
|
import matplotlib
|
||
|
if os.name == 'posix' and "DISPLAY" not in os.environ:
|
||
|
matplotlib.use('Agg')
|
||
|
|
||
|
template_dt = 5e-2 # ns
|
||
|
bp_freq = (30e-3, 80e-3) # GHz
|
||
|
template_length = 100 # ns
|
||
|
noise_sigma_factor = 1e1 # keep between 10 and 0.1
|
||
|
|
||
|
antenna_dt = 2 # ns
|
||
|
antenna_timelength = 2048 # ns
|
||
|
|
||
|
_deltapeak = util.deltapeak(timelength=template_length, samplerate=1/template_dt, offset=10/template_dt)
|
||
|
template_t = util.sampled_time(1/template_dt, start=0, end=template_length)
|
||
|
template_signal = antenna_bp(_deltapeak[0], *bp_freq, (np.sqrt(antenna_dt/template_dt))*template_dt) # TODO: fix sqrt constant
|
||
|
template_peak_time = template_t[_deltapeak[1]]
|
||
|
|
||
|
if False: # show template
|
||
|
fig, ax = plt.subplots()
|
||
|
ax.set_title("Deltapeak and Bandpassed Template")
|
||
|
ax.set_xlabel("Time [ns]")
|
||
|
ax.set_ylabel("Amplitude")
|
||
|
ax.plot(template_t, max(template_signal)*_deltapeak[0])
|
||
|
ax.plot(template_t, template_signal)
|
||
|
fig.savefig('figures/11_template_deltapeak.pdf')
|
||
|
|
||
|
# receive at antenna
|
||
|
antenna_t = util.sampled_time(1/antenna_dt, start=0, end=antenna_timelength)
|
||
|
antenna_samplelength = len(antenna_t)
|
||
|
|
||
|
## place the deltapeak signal at a random location
|
||
|
antenna_true_signal, antenna_peak_location = util.deltapeak(timelength=antenna_timelength, samplerate=1/antenna_dt, offset=[0.2, 0.8], rng=rng)
|
||
|
antenna_peak_time = antenna_t[antenna_peak_location]
|
||
|
|
||
|
if not True: # flip polarisation
|
||
|
antenna_true_signal *= -1
|
||
|
|
||
|
## Add noise
|
||
|
noise_amplitude = max(template_signal) * noise_sigma_factor
|
||
|
noise_realisation = noise_amplitude * white_noise_realisation(len(antenna_true_signal))
|
||
|
antenna_unfiltered_signal = antenna_true_signal + noise_realisation
|
||
|
antenna_signal = antenna_bp(antenna_unfiltered_signal, *bp_freq, antenna_dt)
|
||
|
|
||
|
true_time_offset = antenna_peak_time - template_peak_time
|
||
|
|
||
|
if True: # show signals
|
||
|
fig, axs = plt.subplots(2, sharex=True)
|
||
|
axs[0].set_title("Antenna Waveform")
|
||
|
axs[-1].set_xlabel("Time [ns]")
|
||
|
axs[0].set_ylabel("Amplitude")
|
||
|
axs[0].plot(antenna_t, antenna_signal, label='bandpassed w/ noise')
|
||
|
axs[0].plot(antenna_t, antenna_unfiltered_signal, label='true signal w/ noise')
|
||
|
axs[0].plot(antenna_t, antenna_true_signal, label='true signal w/o noise')
|
||
|
axs[0].legend()
|
||
|
|
||
|
axs[1].set_title("Template")
|
||
|
axs[1].set_ylabel("Amplitude")
|
||
|
axs[1].plot(template_t, template_signal, label='orig')
|
||
|
axs[1].plot(template_t + true_time_offset, template_signal, label='moved orig')
|
||
|
axs[1].legend()
|
||
|
|
||
|
fig.savefig('figures/11_antenna_signals.pdf')
|
||
|
|
||
|
if True: # zoom
|
||
|
wx = 100
|
||
|
x0 = true_time_offset
|
||
|
|
||
|
old_xlims = axs[0].get_xlim()
|
||
|
axs[0].set_xlim( x0-wx, x0+wx)
|
||
|
fig.savefig('figures/11_antenna_signals_zoom.pdf')
|
||
|
|
||
|
# restore
|
||
|
axs[0].set_xlim(*old_xlims)
|
||
|
|
||
|
if True: # upsampled trace
|
||
|
upsampled_trace, upsampled_t = trace_upsampler(template_signal, antenna_signal, template_t, antenna_t)
|
||
|
|
||
|
if True: # Show upsampled traces
|
||
|
fig2, axs2 = plt.subplots(1, sharex=True)
|
||
|
if not hasattr(axs2, '__len__'):
|
||
|
axs2 = [axs2]
|
||
|
|
||
|
axs2[-1].set_xlabel("Time [ns]")
|
||
|
axs2[0].set_ylabel("Amplitude")
|
||
|
axs2[0].plot(antenna_t, antenna_signal, marker='o', label='orig')
|
||
|
axs2[0].plot(upsampled_t, upsampled_trace, label='upsampled')
|
||
|
axs2[0].legend(loc='upper right')
|
||
|
|
||
|
fig2.savefig('figures/11_upsampled.pdf')
|
||
|
|
||
|
wx = 1e2
|
||
|
x0 = upsampled_t[0] + wx - 5
|
||
|
axs2[0].set_xlim(x0-wx, x0+wx)
|
||
|
fig2.savefig('figures/11_upsampled_zoom.pdf')
|
||
|
|
||
|
# determine correlations with arguments
|
||
|
lag_dt = upsampled_t[1] - upsampled_t[0]
|
||
|
corrs, (out1_signal, out2_template, lags) = my_correlation(upsampled_trace, template_signal)
|
||
|
|
||
|
else: # downsampled template
|
||
|
raise NotImplementedError
|
||
|
|
||
|
corrs, (out1_signal, out2_signal, lags) = my_downsampling_correlation(template_signal, antenna_signal, template_t, antenna_t)
|
||
|
lag_dt = upsampled_t[1] - upsampled_t[0]
|
||
|
|
||
|
# Determine best correlation time
|
||
|
idx = np.argmax(abs(corrs))
|
||
|
best_sample_lag = lags[idx]
|
||
|
best_time_lag = best_sample_lag * lag_dt
|
||
|
if axs2:
|
||
|
axs2[-1].axvline(best_time_lag, color='r', alpha=0.5, linewidth=2)
|
||
|
|
||
|
# Show the final signals correlated
|
||
|
if True:
|
||
|
fig, axs = plt.subplots(3, sharex=True)
|
||
|
ylabel_kwargs = dict(
|
||
|
rotation=0,
|
||
|
ha='right',
|
||
|
va='center'
|
||
|
)
|
||
|
axs[-1].set_xlabel("Time [ns]")
|
||
|
|
||
|
# Signal
|
||
|
i=0
|
||
|
axs[i].set_ylabel("Signal\nAmplitude", **ylabel_kwargs)
|
||
|
axs[i].plot(antenna_t, antenna_signal)
|
||
|
|
||
|
# Template
|
||
|
i=1
|
||
|
axs[i].set_ylabel("Template\nAmplitude", **ylabel_kwargs)
|
||
|
for offset in [0, best_time_lag]:
|
||
|
axs[i].axvline(offset + len(template_signal) * (template_t[1] - template_t[0]), color='g')
|
||
|
axs[i].axvline(offset, color='g')
|
||
|
axs[i].plot(offset + template_t, template_signal)
|
||
|
|
||
|
|
||
|
# Correlation
|
||
|
i=2
|
||
|
axs[i].set_ylabel("Correlation", **ylabel_kwargs)
|
||
|
axs[i].plot(lags * lag_dt, corrs)
|
||
|
axs[i].axvline(best_time_lag, color='r', ls='--')
|
||
|
|
||
|
if True: # zoom
|
||
|
wx = len(template_signal) * (template_t[1] - template_t[0])/2
|
||
|
t0 = best_time_lag
|
||
|
|
||
|
for t in [t0-wx, t0+wx]:
|
||
|
axs[2].axvline(t, color='g')
|
||
|
|
||
|
fig.tight_layout()
|
||
|
fig.legend()
|
||
|
fig.savefig('figures/11_corrs.pdf')
|
||
|
|
||
|
#
|
||
|
time_residual = best_time_lag - true_time_offset
|
||
|
|
||
|
print(time_residual, template_dt, antenna_dt)
|
||
|
|
||
|
|
||
|
plt.show()
|