时域图
python
# 导入相关的库
import pickle
import matplotlib.pyplot as plt
import numpy as np
from pathlib import Path
import os
with open(r"C:\0-数据集\公开\RML2016\RML2016.10a_dict.pkl", 'rb') as file:
Xd = pickle.load(file, encoding='bytes')
snrs, mods = map(lambda j: sorted(list(set(map(lambda x: x[j], Xd.keys())))), [1, 0])
print(snrs)
print(mods)
def time_domain_rml2016(select_mod=b'PAM4',select_snr=16, select_num=10):
print(select_mod, select_snr)
path = Path("./time/" + str(select_mod) + '_' + str(select_snr))
# print(path)
if not path.exists():
os.makedirs(path)
for i in range(select_num):
fig = plt.figure()
plt.plot(Xd[select_mod, select_snr][i, 0])
plt.plot(Xd[select_mod, select_snr][i, 1])
name = str(select_mod) + '_' + str(select_snr) + '_' + str(i) + '.png'
print(name)
plt.axis('off') # 关闭坐标轴
plt.gca().set_frame_on(False) # 关闭图形边框
# 只保存散点图
plt.savefig(path/name)
# plt.show()
time_domain_rml2016()
星座图
把输入的IQ信号定义为信号的实部和虚部(但实际上IQ两路信号都是实信号)将I和Q分别作为横轴和纵轴,那么在复平面上每两个IQ值可以对应一个固定的点,将坐标图画出来就叫做星座图。
python
# 导入相关的库
import pickle
import matplotlib.pyplot as plt
import numpy as np
from pathlib import Path
import os
with open(r"C:\0-数据集\公开\RML2016\RML2016.10a_dict.pkl", 'rb') as file:
Xd = pickle.load(file, encoding='bytes')
snrs, mods = map(lambda j: sorted(list(set(map(lambda x: x[j], Xd.keys())))), [1, 0])
print(snrs)
print(mods)
def constellation_rml2016(select_mod=b'QPSK',select_snr=16, select_num=10):
data = []
# 遍历调制方式和信噪比,提取数据
for mod in mods:
for snr in snrs:
if mod == select_mod and snr == select_snr:
data.append(Xd[(mod,snr)])
data = np.vstack(data)
print(len(data))
print(select_mod, select_snr)
path = Path("./" + str(select_mod) + '_' + str(select_snr))
# print(path)
if not path.exists():
os.makedirs(path)
for i in range(select_num):
x, y = data[i]
fig = plt.figure()
plt.scatter(x, y, c='blue')
# plt.xlabel("I")
# plt.ylabel("Q")
name = str(select_mod) + '_' + str(select_snr) + '_' + str(i) + '.png'
print(name)
# 不显示坐标轴、标题等信息
plt.axis('off') # 关闭坐标轴
plt.gca().set_frame_on(False) # 关闭图形边框
# 只保存散点图
plt.savefig(path/name)
# plt.show()
constellation_rml2016()
功率谱
python
# 导入相关的库
import pickle
import matplotlib.pyplot as plt
import numpy as np
from pathlib import Path
import os
with open(r"C:\0-数据集\公开\RML2016\RML2016.10a_dict.pkl", 'rb') as file:
Xd = pickle.load(file, encoding='bytes')
snrs, mods = map(lambda j: sorted(list(set(map(lambda x: x[j], Xd.keys())))), [1, 0])
print(snrs)
print(mods)
def spectrum_power_rml2016(select_mod=b'AM-DSB',select_snr=0, select_num=10):
print(select_mod, select_snr)
path = Path("./spectrum/" + str(select_mod) + '_' + str(select_snr))
# print(path)
if not path.exists():
os.makedirs(path)
for i in range(select_num):
fig = plt.figure()
data_I = Xd[select_mod, select_snr][i, 0]
data_Q = Xd[select_mod, select_snr][i, 1]
# 合成IQ信号
data_IQ = data_I + 1j*data_Q
# 功率谱分析:对IQ信号进行FFT,计算功率谱
power_spectrum = np.abs(np.fft.fft(data_IQ)) ** 2
# 计算频率轴
frequencies = np.fft.fftfreq(len(power_spectrum), 1) # 假设 time_step 为 1
idx = np.argsort(frequencies)
plt.plot(frequencies[idx], power_spectrum[idx])
name = str(select_mod) + '_' + str(select_snr) + '_' + str(i) + '.png'
print(name)
plt.axis('off') # 关闭坐标轴
plt.gca().set_frame_on(False) # 关闭图形边框
# 只保存散点图
plt.savefig(path/name)
# plt.show()
spectrum_power_rml2016()
参考链接
RML2016.10a数据集画星座图、频域图、时域图
JoshiShamika/Deep-learning-for-Modulation-Recognition-on-RML2016.10a_dict-dataset