-
Notifications
You must be signed in to change notification settings - Fork 0
/
dup.py
402 lines (316 loc) · 14.2 KB
/
dup.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
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
#軟判定
from tkinter import LEFT
from turtle import color
import numpy as np
import csv
import operator
import matplotlib.pyplot as plt
import math
import random
from numpy.random import *
from scipy import special
S_REG = 3 # レジスタ数(前後半共通)
LENGTH = 259 # 符号長
TEST = 10 # テスト回数
OUT_BITS = 2 # 後半組は3
OUT_LEN = LENGTH * OUT_BITS #777
K = S_REG + 1 # 拘束長は4(前後半共通)
STATE_NUM = 8 # 後半組は16
def awgn(SNRdB, size):
No = OUT_BITS * 1 * 10 ** (-SNRdB / 10)
noise = np.random.normal(0, np.sqrt(No / 2), size) + 1j * np.random.normal(
0, np.sqrt(No / 2), size
)
return noise
def hamming(s1, s2):
# ハミング距離計算
return sum(map(operator.xor, s1, s2))
def distance(s1,s2,i,j):
np.place(s1, s1== 0, -1)
return np.dot(s1,s2,h_channel[i][j])
def convolutional_encoder(data, state):
# 状態と入力から，次の状態を返す
# 状態遷移図の規則h性を活用
state = (2 * state + data) % 8
return state
def combinations_count(n, r):
return math.factorial(n) // (math.factorial(n - r) * math.factorial(r))
# 初期化
tdata = np.zeros((TEST, LENGTH), dtype=int)
rdata = np.zeros((TEST, LENGTH), dtype=int)
tcode = np.zeros((TEST, OUT_LEN), dtype=int)
receive = np.zeros((TEST, OUT_LEN), dtype=float)
# TODO レイリーチャネルの生成
h_channel = np.random.rayleigh(scale = 1,size = (TEST,OUT_LEN))
h_nocode = np.random.rayleigh(scale = 1,size = (TEST,LENGTH))
state = 0
snr_list = []
ber_list = []
nocode_ber_list = []
nocode_rayleigh_ber_list = []
p_b_list = []
# 各時間，各状態において，ハミング距離を記録する
# metric[状態][時刻]
metric = -10000 * np.ones((STATE_NUM, LENGTH + 1), dtype=float)
metric[0][0] = 0
##各時間(260)，各状態(8)へのパス(2;どの状態からどの入力)を記録する
# path[状態][時刻][[前状態,入力]]
path = np.zeros((STATE_NUM, LENGTH, 2), dtype=int)
transmit = receive = np.zeros((TEST, OUT_LEN))
nocode_transmit = nocode_receive = nocode_demo = nocode_rayleigh_recieve = nocode_rayleith_demo= np.zeros((TEST, LENGTH))
# 状態と入力が決まると，出力が決まる3次元配列
# output[状態][入力][出力]
output = np.zeros((STATE_NUM, 2, OUT_BITS), dtype=int)
output[0, 0] = [0, 0]
output[0, 1] = [1, 1]
output[1, 0] = [1, 1]
output[1, 1] = [0, 0]
output[2, 0] = [0, 1]
output[2, 1] = [1, 0]
output[3, 0] = [1, 0]
output[3, 1] = [0, 1]
output[4, 0] = [1, 1]
output[4, 1] = [0, 0]
output[5, 0] = [0, 0]
output[5, 1] = [1, 1]
output[6, 0] = [1, 0]
output[6, 1] = [0, 1]
output[7, 0] = [0, 1]
output[7, 1] = [1, 0]
hugou = np.zeros((STATE_NUM, 2, OUT_BITS), dtype=int)
hugou[0, 0] = [-1, -1]
hugou[0, 1] = [1, 1]
hugou[1, 0] = [1, 1]
hugou[1, 1] = [-1, -1]
hugou[2, 0] = [-1, 1]
hugou[2, 1] = [1, -1]
hugou[3, 0] = [1, -1]
hugou[3, 1] = [-1, 1]
hugou[4, 0] = [1, 1]
hugou[4, 1] = [-1, -1]
hugou[5, 0] = [-1, -1]
hugou[5, 1] = [1, 1]
hugou[6, 0] = [1, -1]
hugou[6, 1] = [-1, 1]
hugou[7, 0] = [-1, 1]
hugou[7, 1] = [1, -1]
array = [["SNR", "BER", "NOCODE_BER", "p_k"]]
file_path = "./test.csv" # CSVの書き込みpath．任意で変えて．
# tdata: 符号化前の送信データ transmission
# tcode: 符号化後の送信データ
# rdata: 復号化前の受信データ receive
# rcode: 復号化後の受信データ
# transmit: 送信信号
# receive: 受信信号
if __name__ == "__dup__":
# 表示
print("# SNR BER:")
# 伝送シミュレーション
for SNRdB in np.arange(0, 6.25, 0.25):
# 送信データの生成
tdata = np.random.randint(0, 2, (TEST, LENGTH - S_REG)) # 送信データをランダムのバイナリで生成
rdata = np.zeros((TEST, LENGTH), dtype=float)
# 終端ビット系列の付加h
end = np.zeros((TEST, S_REG), dtype=int)
tdata = np.append(tdata, end, axis=1)
# 畳み込み符号化
for i in range(TEST):
for k in range(OUT_LEN):
test = (randn(1, 1) + 1j * randn(1, 1)) * 1 / np.sqrt(2)
for j in range(LENGTH):
if j == 0:
state = 0
else:
tcode[i][2 * j], tcode[i][2 * j + 1] = output[state][tdata[i][j]]
state = convolutional_encoder(tdata[i][j], state)
# BPSK変調
transmit[tcode == 0] = -1
transmit[tcode == 1] = 1
nocode_transmit[tdata == 0] = -1
nocode_transmit[tdata == 1] = 1
# 伝送
receive = h_channel * transmit + awgn(SNRdB, (TEST, OUT_LEN))
nocode_receive = nocode_transmit + awgn(SNRdB, (TEST, LENGTH))
nocode_rayleith = h_channel[:TEST,:LENGTH] * nocode_transmit+ awgn(SNRdB,(TEST,LENGTH))
# BPSK復調
#rcode[receive < 0] = 0
nocode_demo[nocode_receive < 0] = 0
nocode_demo[nocode_receive >= 0] = 1
nocode_rayleith_demo[nocode_rayleith < 0] = 0
nocode_rayleith_demo[nocode_rayleith >= 0] = 1
print(np.count_nonzero(nocode_demo != tdata))
# ビタビ復号
for i in range(TEST):
for j in range(LENGTH):
# TODO メトリックの書き換え
r_pair = [0] * OUT_BITS
r_pair = np.append(receive[i][2 * j], receive[i][2 * j + 1])
h_pair = np.append(h_channel[i][2*j], h_channel[i][2 * j + 1])
if j == 0:
metric[0][0] = 0
metric[0][1] = metric[0][0] + np.sum(hugou[0][0]*r_pair* h_pair)
metric[1][1] = metric[0][0] + np.sum(hugou[0][1]*r_pair *h_pair)
else:
# 8状態においてハミング距離更新かつパスの記録
# template
# if (metric[状態a][j-1]+distance(output[状態a][入力a],r_pair)) <(metric[状態b][j-1] + distance(output[状態b][入力b],r_pair)):
# metric[2][j] = 前者
# path[2][j] = [状態a,入力a]
# else:
# metric[2][j] =後者
# path[2][j] =[状態b,入力b]
# 状態0
# 左辺の方がパスメトリック小さい場合
if (metric[0][j - 1] + np.sum(hugou[0][0] * r_pair * h_pair)) > (
metric[4][j - 1] + np.sum(hugou[4][0] * r_pair * h_pair)
):
metric[0][j] = metric[0][j - 1] + np.sum(
hugou[0][0] * r_pair * h_pair
) # ハミング距離更新．(状態0時刻jのハミング距離を求める)
path[0][j] = [0, 0] # パスの記録(状態0からの入力0)
# 右辺の方がパスメトリック小さい場合
else:
metric[0][j] = metric[4][j - 1] + np.sum(
hugou[4][0]* r_pair * h_pair
) # ハミング距離更新
path[0][j] = [4, 0] # パスの記録，(状態4からの入力0)
# 状態1
if (metric[0][j - 1] + np.sum(hugou[0][1] * r_pair * h_pair)) > (
metric[4][j - 1] + np.sum(hugou[4][1] * r_pair * h_pair)
): # 状態1時刻jのハミング距離を求める
metric[1][j] = metric[0][j - 1] + np.sum(hugou[0][1] * r_pair * h_pair)
path[1][j] = [0, 1] # 状態1に来るパスは，状態0からの入力1
else:
metric[1][j] = metric[4][j - 1] + np.sum(
hugou[4][1] * r_pair * h_pair
) # 状態1時刻jのハミング距離を求める
path[1][j] = [4, 1] # 状態1に来るパスは，状態4からの入力1
##状態2
if (metric[1][j - 1] + np.sum(hugou[1][0] * r_pair * h_pair)) > (
metric[5][j - 1] + np.sum(hugou[5][0] * r_pair * h_pair)
):
metric[2][j] = metric[1][j - 1] + np.sum(hugou[1][0] * r_pair * h_pair)
path[2][j] = [1, 0]
else:
metric[2][j] = metric[5][j - 1] + np.sum(hugou[5][0] * r_pair * h_pair)
path[2][j] = [5, 0]
##状態3
if (metric[1][j - 1] + np.sum(hugou[1][1] * r_pair * h_pair)) > (
metric[5][j - 1] + np.sum(hugou[5][1] * r_pair * h_pair)
):
metric[3][j] = metric[1][j - 1] + np.sum(hugou[1][1] * r_pair * h_pair)
path[3][j] = [1, 1]
else:
metric[3][j] = metric[5][j - 1] + np.sum(hugou[5][1] * r_pair * h_pair)
path[3][j] = [5, 1]
##状態4
if (metric[2][j - 1] + np.sum(hugou[2][0] * r_pair * h_pair)) > (
metric[6][j - 1] + np.sum(hugou[6][0] * r_pair * h_pair)
):
metric[4][j] = metric[2][j - 1] + np.sum(hugou[2][0] * r_pair * h_pair)
path[4][j] = [2, 0]
else:
metric[4][j] = metric[6][j - 1] + np.sum(hugou[6][0] * r_pair * h_pair)
path[4][j] = [6, 0]
##状態5
if (metric[2][j - 1] + np.sum(hugou[2][1] * r_pair * h_pair)) > (
metric[6][j - 1] + np.sum(hugou[6][1] * r_pair * h_pair)
):
metric[5][j] = metric[2][j - 1] + np.sum(hugou[2][1] * r_pair * h_pair)
path[5][j] = [2, 1]
else:
metric[5][j] = metric[6][j - 1] + np.sum(hugou[6][1] * r_pair * h_pair)
path[5][j] = [6, 1]
##状態6
if (metric[3][j - 1] + np.sum(hugou[3][0] * r_pair * h_pair)) > (
metric[7][j - 1] + np.sum(hugou[7][0] * r_pair * h_pair)
):
metric[6][j] = metric[3][j - 1] + np.sum(hugou[3][0] * r_pair * h_pair)
path[6][j] = [3, 0]
else:
metric[6][j] = metric[7][j - 1] + np.sum(hugou[7][0] * r_pair * h_pair)
path[6][j] = [7, 0]
##状態7
if (metric[3][j - 1] + np.sum(hugou[3][1] * r_pair * h_pair)) > (
metric[7][j - 1] + np.sum(hugou[7][1] * r_pair * h_pair)
):
metric[7][j] = metric[3][j - 1] + np.sum(hugou[3][1] * r_pair * h_pair)
path[7][j] = [3, 1]
else:
metric[7][j] = metric[7][j - 1] + np.sum(hugou[7][1] * r_pair * h_pair)
path[7][j] = [7, 1]
# 復号系列を求める
for t in reversed(range(LENGTH)):
if t == LENGTH - 1:
rdata[i][t] = path[0][t][1]
prev = path[0][t][0]
elif t == 0:
rdata[i][t] = tdata[i][t]
else:
# 復元データはpath
rdata[i][t] = path[prev][t][1] # 状態prevの時刻t+1に向かってくるパスの入力
prev = path[prev][t][0] # 状態prevの時刻tに向かってくるパスの状態
# 誤り回数計算
ok = np.count_nonzero(rdata == tdata)
error = rdata.size - ok
nocode_ok = np.count_nonzero(nocode_demo == tdata)
nocode_rayleigh_ok =np.count_nonzero(nocode_rayleith_demo == tdata)
nocode_error = rdata.size - nocode_ok
nocode_rayleith_error = rdata.size - nocode_rayleigh_ok
# BER計算
BER = error / (ok + error)
NOCODE_BER = nocode_error / (nocode_ok + nocode_error)
NOCODE_RAYLEIGH_BER = nocode_rayleith_error / (nocode_rayleigh_ok + nocode_rayleith_error)
#TODO 軟判定理論上界
# 硬判定理論上界計算
p_k = [0] * 14 # 後半組は18にする
p = 1 / 2 * special.erfc(np.sqrt(1 / 2 * 10 ** (SNRdB / 10)))
for k in range(6, 14): # 後半組はrange(10,18)
p_k[k] = 1 / 2 * special.erfc(np.sqrt(k* (1 / 2) * 10 ** (SNRdB / 10)))
p_b = 0
p_b = (
2 * p_k[6]
+ 7 * p_k[7]
+ 18 * p_k[8]
+ 49 * p_k[9]
+ 130 * p_k[10]
+ 333 * p_k[11]
+ 836 * p_k[12]
+ 2069 * p_k[13]
)
# 後半組は以下のように書き換え
# p_b = 6 * p_k[10] + 0 * p_k[11] + 6 * p_k[12] + 0 * p_k[13] + 58 * p_k[14] + 0 * p_k[15] + 118 * p_k[16] + 0 * p_k[17]
if p_b >= 1 / 2:
p_b = 1 / 2
p_b_list.append(p_b)
snr_list.append(SNRdB)
ber_list.append(BER)
nocode_ber_list.append(NOCODE_BER)
nocode_rayleigh_ber_list.append(NOCODE_RAYLEIGH_BER)
# 結果表示
print(
"SNR: {0:.2f}, BER: {1:.4e}, NOCODE_BER:{2:.4e},NOCOXE_RAYLEIGH_BER:{3:.4e}, UPPER_BOUND:{3:.4e}".format(
SNRdB, BER, NOCODE_BER, NOCODE_RAYLEIGH_BER, p_b
)
)
# print('NOCODE_BER:{1:.4e}'.format(*NOCODE_BER))
# print(rdata)
# CSV書き込み．コメントアウト解除すれば書き込める
array.append([SNRdB, BER, NOCODE_BER,NOCODE_RAYLEIGH_BER])
# array.append([tdata,rdata])
with open(file_path, "w") as f:
writer = csv.writer(f, lineterminator="\n")
writer.writerows(array)
fig = plt.figure()
plt.plot(snr_list, ber_list, label="simulation(soft) with rayleigh", color="blue")
plt.plot(snr_list, nocode_ber_list, label="without coding no rayleigh", color="red")
#plt.plot(snr_list, nocode_rayleigh_ber_list, label="rayleigh without coding", color="black")
plt.plot(snr_list, p_b_list, label="upper bound no rayleigh", color="green")
ax = plt.gca()
ax.set_yscale("log")
ax.legend(loc=0)
fig.tight_layout()
plt.xlabel("E_b/N_0")
plt.ylabel("BER")
fig.savefig("img.png")
plt.show()