/
az_train_LSTM_AE_v1.py
252 lines (225 loc) · 9.93 KB
/
az_train_LSTM_AE_v1.py
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
import os
import numpy as np
import matplotlib.pyplot as plt
import time
#
import tensorflow as tf
from keras import Model, layers
from keras.models import Sequential
from keras.layers import Dense
from keras.layers import Flatten
from keras.layers import Dropout
from keras.layers import LSTM
from keras.layers import RepeatVector
from keras.layers import TimeDistributed
from keras.layers.convolutional import Conv1D
from keras.layers.convolutional import MaxPooling1D
t0 = time.time()
# ===================================================
Rician_factor = 5
power_B = 0.05 # Watts
M_Bobs = 4
power_E = 0.03 # Watts
N_Eves = 2
n_anten_Alice = 15
# RIS
K_RIS = 8
phi_list = [np.pi for i in range(K_RIS)]
phi_array = np.array(phi_list)
Phi_RIS = np.diag(np.exp(1j*2*np.pi*phi_array))
xi = 0.1 # temporal-correlation coefficient
# ===================================================
""" Fit a model & Predict the future """
# def LSTM_AE__sequential_model(trainX_3D, testX_3D):
# epochs, batch_size = 500, 32
# n_features = trainX_3D.shape[2]
# ''' define model using the Sequential API '''
# model = Sequential()
# model.add(LSTM(64, activation='relu', return_sequences=True))
# model.add(LSTM(32, activation='relu', return_sequences=True))
# model.add(LSTM(16, activation='relu', return_sequences=True))
# model.add(LSTM(32, activation='relu', return_sequences=True))
# model.add(LSTM(64, activation='relu', return_sequences=True))
# model.add(TimeDistributed(Dense(n_features)))
# model.compile(optimizer='adam', loss='mse')
# ''' fit model '''
# model.fit(trainX_3D, trainX_3D,
# epochs=epochs, batch_size=batch_size, verbose=0)
# testX_hat_3D = model.predict(testX_3D, verbose=0)
# return testX_hat_3D
def LSTM_AE__functional_model(trainX_3D):
epochs, batch_size = 70, 32
n_time_steps_IN = trainX_3D.shape[1]
n_features = trainX_3D.shape[2]
''' define model using the functional API '''
input_ = layers.Input(shape=(n_time_steps_IN, n_features))
hidden = LSTM(64, activation='relu', return_sequences=True)(input_)
hidden = LSTM(32, activation='relu', return_sequences=True)(hidden)
hidden = LSTM(16, activation='relu', return_sequences=True)(hidden)
hidden = LSTM(32, activation='relu', return_sequences=True)(hidden)
hidden = LSTM(64, activation='relu', return_sequences=True)(hidden)
output_ = TimeDistributed(Dense(n_features))(hidden)
model = Model(input=input_, output=output_)
model.compile(optimizer='adam', loss='mse')
''' fit model '''
history = model.fit(trainX_3D, trainX_3D,
epochs=epochs, batch_size=batch_size, verbose=0)
trainX_hat_3D = model.predict(trainX_3D, verbose=0)
return model, history, trainX_hat_3D
# ===================================================
n_time_steps_IN = 15
n_steps_OUT = 5
""" Create 'folder_name' in 'input' """
input_folder = '/input/NA' + str(n_anten_Alice)\
+ '_B' + str(power_B)\
+ '_E' + str(power_E)\
+ '_vB' + str(60)\
+ '_NE' + str(N_Eves)\
+ '_K' + str(Rician_factor)\
+ '_IN' + str(n_time_steps_IN)\
+ '_OUT' + str(n_steps_OUT)
input_path = os.path.abspath(os.getcwd()) + input_folder
path_to_input = os.path.join(input_path, '')
if not os.path.exists(path_to_input): # check if the subfolder exists
os.makedirs(path_to_input) # create the subfolder
""" Loading data """
trainX_3D_H0 = np.load(input_path+'/trainX_H0__LSTM_IN.npy')
testX_3D_H0 = np.load(input_path+'/testX_H0__LSTM_IN.npy')
testX_3D_H1 = np.load(input_path+'/testX_H1__LSTM_IN.npy')
testX_3D = np.vstack((testX_3D_H0, testX_3D_H1))
n_samples_train = trainX_3D_H0.shape[0]
n_samples_test_H0 = testX_3D_H0.shape[0]
n_samples_test_H1 = testX_3D_H1.shape[0]
n_samples_test = n_samples_test_H0 + n_samples_test_H1
n_features = trainX_3D_H0.shape[2]
print('n_time_steps_IN =', n_time_steps_IN)
cur_sample_i = 0
cur_anten_ell = 0
time_indices_for_a_sample = [(i+1) for i in range(n_time_steps_IN)]
plt.figure()
plt.plot(time_indices_for_a_sample,
testX_3D_H0[cur_sample_i, :, cur_anten_ell],
color='b')
plt.plot(time_indices_for_a_sample,
testX_3D_H1[cur_sample_i, :, cur_anten_ell],
color='r', linestyle='--')
plt.xlim((0, n_time_steps_IN))
plt.xlabel(r'$1\leq \mu\leq\Delta = %s$' % (n_time_steps_IN), fontsize=12)
plt.ylabel(r'$q^{[\mu]}_{%s}$' % (cur_anten_ell+1), fontsize=15)
plt.show()
""" Training and testing datasets to be reconstructed """
n_evaluations = 50
scaling_list = [0.5+0.1*i for i in range(27)]
acc_H0_arr = np.zeros([n_evaluations, len(scaling_list)])
acc_H1_arr = np.zeros([n_evaluations, len(scaling_list)])
for u in range(n_evaluations):
# name = 'aff'
model, hist, trainX_hat_3D_H0 = LSTM_AE__functional_model(trainX_3D_H0)
testX_hat_3D = model.predict(testX_3D, verbose=0)
testX_hat_3D_H0 = testX_hat_3D[:n_samples_test_H0]
testX_hat_3D_H1 = testX_hat_3D[n_samples_test_H0:]
""" Detection Algorithm """
diff_test_3D_H0 = abs(testX_hat_3D_H0 - testX_3D_H0)
mean_diff_test_H0 = np.empty([0, 1])
for n in range(n_samples_test_H0):
mean_diff_test_H0 = np.append(mean_diff_test_H0,
np.mean(diff_test_3D_H0[n, :, :]))
# end of 'for' loop
diff_test_3D_H1 = abs(testX_hat_3D_H1 - testX_3D_H1)
mean_diff_test_H1 = np.empty([0, 1])
for n in range(n_samples_test_H1):
mean_diff_test_H1 = np.append(mean_diff_test_H1,
np.mean(diff_test_3D_H1[n, :, :]))
# end of 'for' loop
""" Detection Threshold """
diff_train_3D_H0 = abs(trainX_hat_3D_H0 - trainX_3D_H0)
thrsh = np.mean(diff_train_3D_H0)
""" Calculate the probabilities """
acc_H0_list = []
acc_H1_list = []
for scaling in scaling_list:
thrsh_scaled = scaling*thrsh
H0_decoded_as_H0 = (mean_diff_test_H0 <= thrsh_scaled)*1
H1_decoded_as_H1 = (mean_diff_test_H1 > thrsh_scaled)*1
acc_H0 = np.sum(H0_decoded_as_H0)/len(H0_decoded_as_H0)
acc_H1 = np.sum(H1_decoded_as_H1)/len(H1_decoded_as_H1)
acc_H0_list.append(acc_H0)
acc_H1_list.append(acc_H1)
# end of 'for' loop
acc_H0_arr[u, :] = np.array(acc_H0_list)
acc_H1_arr[u, :] = np.array(acc_H1_list)
# ===================================================
""" Create 'folder_name' in 'output' """
# output_folder = '/output/NA' + str(n_anten_Alice)\
# + '_B' + str(power_B)\
# + '_E' + str(power_E)\
# + '_vB' + str(60)\
# + '_NE' + str(N_Eves)\
# + '_K' + str(Rician_factor)\
# + '_IN' + str(n_time_steps_IN)\
# + '_OUT' + str(n_steps_OUT)
# output_path = os.path.abspath(os.getcwd()) + output_folder
# path_to_output = os.path.join(output_path, '')
# if not os.path.exists(path_to_output): # check if the subfolder exists
# os.makedirs(path_to_output) # create the subfolder
""" Save the OUTPUT of LSTM """
# np.save(output_path+'/trainX_H0__LSTM_OUT.npy', trainX_hat_3D_H0)
# np.save(output_path+'/testX_H0__LSTM_OUT.npy', testX_hat_3D_H0)
# np.save(output_path+'/testX_H1__LSTM_OUT.npy', testX_hat_3D_H1)
# ===================================================
acc_H0_mean_arr = np.mean(acc_H0_arr, axis=0)
acc_H1_mean_arr = np.mean(acc_H1_arr, axis=0)
acc_H0_error_arr = np.sqrt(np.var(acc_H0_arr, axis=0))
acc_H1_error_arr = np.sqrt(np.var(acc_H1_arr, axis=0))
# ===================================================
""" Create the subfolder 'folder_name' in the folder 'input' """
result_folder = '/results/NA' + str(n_anten_Alice)\
+ '_B' + str(power_B)\
+ '_E' + str(power_E)\
+ '_vB' + str(60)\
+ '_NE' + str(N_Eves)\
+ '_K' + str(Rician_factor)\
+ '_IN' + str(n_time_steps_IN)\
+ '_OUT' + str(n_steps_OUT)
result_path = os.path.abspath(os.getcwd()) + result_folder
path_to_results = os.path.join(result_path, '')
if not os.path.exists(path_to_results): # check if the subfolder exists
os.makedirs(path_to_results) # create the results
""" Save the results """
# np.save(result_path+'/loss_vs_epochs.npy', hist.history['loss'])
# np.save(result_path+'/acc_H0_mean_arr.npy', acc_H0_mean_arr)
# np.save(result_path+'/acc_H1_mean_arr.npy', acc_H1_mean_arr)
# np.save(result_path+'/acc_H0_error_arr.npy', acc_H0_error_arr)
# np.save(result_path+'/acc_H1_error_arr.npy', acc_H1_error_arr)
# ===================================================
t1 = time.time()
print('Run time =', np.round((t1-t0)/60, 2), 'mins')
# The 3 following lines are used for generating LaTeX symbols
from matplotlib import rc
rc('text', usetex=True)
rc('text.latex', preamble=r'\usepackage{lmodern}')
""" Illustration """
plt.figure()
plt.plot(hist.history['loss'])
plt.ylim(ymax=20000)
plt.show()
fig = plt.figure()
plt.plot(scaling_list, acc_H0_mean_arr, 'b',
label=r'Prob. of identifying $H_0$ as $H_0$')
plt.plot(scaling_list, acc_H1_mean_arr, 'r--',
label=r'Prob. of identifying $H_1$ as $H_1$')
plt.fill_between(scaling_list,
acc_H0_mean_arr - acc_H0_error_arr,
acc_H0_mean_arr + acc_H0_error_arr,
alpha=0.5, edgecolor='#CC4F1B', facecolor='blue')
plt.fill_between(scaling_list,
acc_H1_mean_arr - acc_H1_error_arr,
acc_H1_mean_arr + acc_H1_error_arr,
alpha=0.5, edgecolor='#CC4F1B', facecolor='orange')
plt.xlim((0.5, 3))
plt.ylim((0, 1))
plt.legend(loc='best', fontsize=12)
plt.xlabel('Threshold-determining factor', fontsize=12)
plt.ylabel('Probability', fontsize=12)
plt.show()
# fig.savefig('saved_figs/' + 'Acc_vs_factor_after_50runs.png', dpi=300)