-
Notifications
You must be signed in to change notification settings - Fork 5
/
Copy pathAirComp.py
83 lines (54 loc) · 2.93 KB
/
AirComp.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
import numpy as np
def AM(setup, d, signal, a_k1, a_k2, b_n, c_1, c_2):
rho = setup.rho
g_mean = np.mean(signal, axis=1)
global_g_mean = rho.T @ g_mean
g_var = np.var(signal, axis=1)
global_g_var = rho.T @ g_var
var_mean_sqrt = global_g_var ** 0.5
noise_1 = (np.random.randn(d) + 1j * np.random.randn(d)) / np.sqrt(2) * np.sqrt(setup.sigma)
noise_2 = (np.random.randn(d) + 1j * np.random.randn(d)) / np.sqrt(2) * np.sqrt(setup.sigma)
noise_N = (np.random.randn(setup.N, d) + 1j * np.random.randn(setup.N, d)) / np.sqrt(2) * np.sqrt(setup.sigma)
x_signal_1 = np.tile(a_k1, (d, 1)).T * (signal - np.tile(global_g_mean, (d, 1)).T) / var_mean_sqrt
x_signal_2 = np.tile(a_k2, (d, 1)).T * (signal - np.tile(global_g_mean, (d, 1)).T) / var_mean_sqrt
r_n = setup.g_kn.T @ x_signal_1 + noise_N
y_1 = setup.h_k.T @ x_signal_1 + noise_1
y_2 = setup.h_k.T @ x_signal_2 + setup.f_n.T @ (np.tile(b_n, (d, 1)).T * r_n) + noise_2
w = np.real((y_1 * c_1 + y_2 * c_2) * var_mean_sqrt + global_g_mean)
true_w = rho.T @ signal
avg_mse = np.linalg.norm((true_w - w)) ** 2 / np.linalg.norm(true_w) ** 2
mse2 = np.linalg.norm((true_w - w)) ** 2 / d
return w, true_w, avg_mse, mse2
def Single(setup, d, signal, a_k1, c_1):
rho = setup.rho
g_mean = np.mean(signal, axis=1)
global_g_mean = rho.T @ g_mean
g_var = np.var(signal, axis=1)
global_g_var = rho.T @ g_var
var_mean_sqrt = global_g_var ** 0.5
noise_1 = (np.random.randn(d) + 1j * np.random.randn(d)) / np.sqrt(2) * np.sqrt(setup.sigma)
x_signal = np.tile(a_k1, (d, 1)).T * (signal - np.tile(global_g_mean, (d, 1)).T) / var_mean_sqrt
y = setup.h_k.T @ x_signal + noise_1
w = np.real(y * c_1 * var_mean_sqrt + global_g_mean)
true_w = rho.T @ signal
avg_mse = np.linalg.norm((true_w - w)) ** 2 / np.linalg.norm(true_w) ** 2
mse2 = np.linalg.norm((true_w - w)) ** 2 / d
return w, true_w, avg_mse, mse2
def Xu(setup, d, signal, a_k1, b_n, c_2):
rho = setup.rho
g_mean = np.mean(signal, axis=1)
global_g_mean = rho.T @ g_mean
g_var = np.var(signal, axis=1)
global_g_var = rho.T @ g_var
var_mean_sqrt = global_g_var ** 0.5
# noise_1 = (np.random.randn(d) + 1j * np.random.randn(d)) / np.sqrt(2) * np.sqrt(setup.sigma)
noise_2 = (np.random.randn(d) + 1j * np.random.randn(d)) / np.sqrt(2) * np.sqrt(setup.sigma)
noise_N = (np.random.randn(setup.N, d) + 1j * np.random.randn(setup.N, d)) / np.sqrt(2) * np.sqrt(setup.sigma)
x_signal_1 = np.tile(a_k1, (d, 1)).T * (signal - np.tile(global_g_mean, (d, 1)).T) / var_mean_sqrt
r_n = setup.g_kn.T @ x_signal_1 + noise_N
y = setup.f_n.T @ (np.tile(b_n, (d, 1)).T * r_n) + noise_2
w = np.real(y / c_2 * var_mean_sqrt + global_g_mean)
true_w = rho.T @ signal
avg_mse = np.linalg.norm((true_w - w)) ** 2 / np.linalg.norm(true_w) ** 2
mse2 = np.linalg.norm((true_w - w)) ** 2 / d
return w, true_w, avg_mse, mse2