DA-MUSIC_ICASSP22/plotFigures.py at main · GitGudNow99/DA-MUSIC_ICASSP22 · GitHub

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
####################################################################################################
#                                          plotFigures.py                                          #
####################################################################################################
#                                                                                                  #
# Authors: J. M.                                                                                   #
#                                                                                                  #
# Created: 26/03/21                                                                                #
#                                                                                                  #
# Purpose: Plot synthetic examples to test correctness and performance of algorithms estimating    #
#          directions of arrival (DoA) of multiple signals.                                        #
#                                                                                                  #
####################################################################################################


#*************#
#   imports   #
#*************#
from tensorflow.keras.models import Model

from augMUSIC import augMUSIC
from beamformer import beamformer
from classicMUSIC import *
from errorMeasures import mean_min_perm_rmse
from models import *
from syntheticEx import *


#********************#
#   initialization   #
#********************#
np. set_printoptions(threshold=10)
np.set_printoptions(suppress=True)

# doa = [-0.1, 0.1]

x, s = construct_signal(doa)
# x, s = construct_coherent_signal(doa)


#*************************#
#   aug MUSIC algorithm   #
#*************************#
inX, outY = deepMUSICorig()
model = Model(inX, outY)
model.load_weights("model/deep_d5.h5")

# crate measurement for the augmentation
x_real = np.real(x)
x_imag = np.imag(x)
X = np.concatenate((x_real, x_imag), axis=0)
X = np.repeat(X[np.newaxis, :, :], 1, axis=0)

K = np.zeros((3, m, m))
Kx = np.cov(x)
K[0] = np.real(Kx)
K[1] = np.imag(Kx)
K[2] = np.angle(Kx)


# true if the augmentation outputs DoA and not subspace
DoAOut = False
spec = True

if DoAOut:
    inp = model.input
    outputs = [layer.output for layer in model.layers]
    # select output layer forming the spectrum and final layer (with DoA)
    functor = [K.function([inp], [outputs[-5]]), K.function([inp], [outputs[-1]])]

    pred_spec , DoA = [f([X]) for f in functor]
    pred_spec = pred_spec[0][0]

    # transform to spectrum indices
    DoA = ((np.array(DoA[0][0]) + np.pi / 2) / np.pi * r).astype(int)

elif spec:
    X = np.repeat(K[np.newaxis, :, :], 1, axis=0)
    pred_specs = model.predict(X)
    pred_spec = np.concatenate(pred_specs, axis=None)
    # DoA = []
    # for i, subspec in enumerate(pred_specs):
    #     angle, _ = signal.find_peaks(subspec[0], distance= r//36)
    #     DoA.append(angle + i * r//36)
    #
    # DoA = np.concatenate(DoA, axis=None)

    DoA, _ = signal.find_peaks(pred_spec, distance=10)

    # only keep d largest peaks
    DoA = DoA[np.argsort(pred_spec[DoA])[-d:]]

else:
    DoA, pred_spec = augMUSIC(model.predict(X), array, angles, d)


#*****************************#
#   classic MUSIC algorithm   #
#*****************************#
DoAMUSIC, spectrum = classicMUSIC(x, array, angles, d)
# DoA, pred_spec = testMUSIC(x, array, angles, d)


#****************#
#   beamformer   #
#****************#
DoABF, spectrumBF = beamformer(x, array, angles, d)


#***********************#
#   plot BF vs. MUSIC   #
#***********************#
def plotBFvMUSIC():
    plt.figure(figsize=(5, 4))
    plt.tight_layout()
    plt.plot(angles[0], spectrumBF, 'k--')
    # plt.plot(angles[0], spectrum)
    plt.plot(doa, np.mean(s, axis=1),'g*')
    plt.plot(angles[0, DoABF], spectrumBF[DoABF], color='grey', linestyle='', marker='x')
    # plt.plot(angles[0, DoAMUSIC], spectrum[DoAMUSIC], 'x')
    plt.xlabel('Azimuth angle (rad)')
    plt.ylabel('Spatial spectrum')
    plt.legend(['Beamformer', 'Actual DoA'], loc='center')


#************************#
#   plot classic MUSIC   #
#************************#
def plotMUSIC():
    plt.figure(figsize=(8, 4))
    plt.tight_layout()
    plt.plot(angles[0], spectrum)
    plt.plot(angles[0, DoAMUSIC], spectrum[DoAMUSIC], 'x')
    plt.plot(doa, np.mean(s, axis=1), '*')
    plt.xlabel('Azimuth angle (rad)')
    plt.ylabel('Spatial spectrum')
    plt.legend(['Classic MUSIC', 'Estimated DoA', 'Actual DoA'])


#************************#
#   plot est. spectrum   #
#************************#
def plotAugMUSIC():
    plt.figure(figsize=(8, 4))
    plt.plot(angles[0], pred_spec)
    plt.plot(angles[0, DoA], pred_spec[DoA],'x')
    plt.plot(doa, [0 for i in range(d)],'*')
    plt.xlabel('Azimuth angle (rad)')
    plt.ylabel('Spatial spectrum')
    plt.legend(['Aug MUSIC', 'Estimated DoA', 'Actual DoA'])


plt.style.use(['grid', 'science', 'no-latex'])

# plotBFvMUSIC()
plotMUSIC()
plotAugMUSIC()

plt.show()