def std_distance(snr_db, tau=None, width_spec=None, tau_disc_fcm=None, mode='nonmod',
K=0.6):
"""(27-29)
Args:
snr_db: in dB
tau: in seconds
width_spec: in Hz
tau_disc_fcm: in seconds
mode ('nonmod', 'lfm', 'fcm'): Default nonmode
K:
Returns:
"""
snr_pow = db2pow(snr_db)
sigma = None
if mode == 'nonmod':
if tau is not None:
sigma = K * scc.c * tau / np.sqrt(snr_pow)
elif mode == 'lfm':
if width_spec is not None:
sigma = K * scc.c / width_spec / np.sqrt(snr_pow)
elif mode == 'fcm':
if tau_disc_fcm is not None:
sigma = K * scc.c * tau_disc_fcm / np.sqrt(snr_pow)
return sigma
def std_speed(snr_db=0, wavelength=None, tau=None, width_spec=None, T_lfm=None,
mode='nonmod',
K=0.6):
"""(27-29)
Args:
mode ('nonmod', 'lfm', 'fcm'): Default 'nonmode'
snr_db: in dB
wavelength: in m
tau: in seconds
width_spec: in Hz
T_lfm: in seconds
K:
Returns:
"""
snr_pow = db2pow(snr_db)
sigma_speed = None
if mode == 'nonmod' or mode == 'fcm':
if tau is not None:
sigma_speed = K * wavelength / tau / np.sqrt(snr_pow)
elif mode == 'lfm':
if width_spec is not None and T_lfm is not None:
sigma_speed = np.sqrt(2) * K * scc.c / width_spec / np.sqrt(snr_pow) / T_lfm
return sigma_speed
def test_main():
H = [[1, 0, 1, 0, 1, 0, 1], [0, 1, 1, 0, 0, 1, 1], [0, 0, 0, 1, 1, 1, 1]]
#[1 0 0 0 0 0 0] [0 0 0 0 0 0 0]
r = [
-0.34808075, 1.26110584, 0.49168568, 0.29461952, 1.1177641, 0.92850077,
1.25788416
]
#sigma = 1
sigma = bawgn.noise_sigma(db2pow(8), 4 / 7)
print(sigma)
cws = ml.all_codewords()
max_iter = 5
Ha = np.array(H)
ra = np.array(r)
# can be done once
msg_store = bpa_init(H)
c_hat = decode_bpa_awgn(msg_store, ra, Ha, sigma, max_iter, cws)
print(c_hat)
print(is_codeword(bawgn.demodulate_hard(ra), Ha))
print(is_codeword(c_hat, Ha))
def simulate_bpa_awgn():
#H = [[0, 1, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 1, 0, 0],
# [0, 0, 1, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 1, 0],
# [1, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 1],
# [0, 0, 0, 1, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 1, 0, 0, 1, 0, 0],
# [0, 0, 0, 0, 1, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 1, 0],
# [0, 0, 0, 0, 0, 1, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 1],
# [0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 1, 0, 0, 1, 0, 0, 0, 1, 0],
# [0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 1, 0, 0, 1, 0, 0, 0, 1],
# [0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 1, 0, 0, 1, 1, 0, 0]]
#n = 21
#k = 12
H = [[1, 0, 0, 1, 1, 0, 1], [0, 1, 0, 1, 0, 1, 1], [0, 0, 1, 0, 1, 1, 1]]
n = 7
k = 4
Ha = np.array(H)
n = Ha.shape[1]
k = n - Ha.shape[0]
# max number of iterations of iterative decoding session (per word!)
max_iter = 10
msg_store = bpa_init(H)
#print(msg_store)
# CW book
cws = ml.all_codewords()
# BCRJ trellis
trellis = bcjr.build_trellis(Ha)
# max runs of simulations (number of words to be transmitted)
max_runs = 1e7
snrDbs = np.arange(-2, 9)
bers = np.zeros(len(snrDbs))
i = 0
for snrDb in snrDbs:
snr = db2pow(snrDb)
sigma = bawgn.noise_sigma(snr, k / n)
#decode = lambda r : decode_bpa_awgn(msg_store, r, Ha, sigma, max_iter)
#decode = lambda r : ml.decode_ml(cws, r)
#decode = lambda r : decode_bpa_bec(msg_store, r, Ha, sigma, max_iter)
decode = lambda r: decide_from_llrs(bcjr.decode(r, trellis, sigma))
ber = simulate_transmission(n, n, max_runs, bawgn.modulate,
bawgn.transmit(sigma), decode)
bers[i] = ber
i += 1
print(snrDb, ber)
# save to mat file for plotting purposes
#scipy.io.savemat('ml_awgn_4', { 'ml_awgn_snrs_4' : snrDbs, 'ml_awgn_bers_4' : bers })
#scipy.io.savemat('bpa_awgn_4', { 'bpa_awgn_snrs_4' : snrDbs, 'bpa_awgn_bers_4' : bers })
#scipy.io.savemat('bpa_bec_2', { 'bpa_bec_snrs_2' : snrDbs, 'bpa_bec_bers_2' : bers })
#scipy.io.savemat('bpa_awgn_h21', { 'bpa_awgn_snrs_h21' : snrDbs, 'bpa_awgn_bers_h21' : bers })
scipy.io.savemat('bpa_awgn_bcjr_3', {
'bpa_awgn_snrs_bcjr_3': snrDbs,
'bpa_awgn_bers_bcjr_3': bers
})
def real_potential_rls(pt, gain_db, wavelength, loss_db, p_min, F=1):
"""(6)
Args:
pt (float): Power in Watts
gain_db (float): Faint in dB
wavelength (float): in m
loss_db (float): Loaa in dB
F (float): norm DNA. def=1
p_min: min power in Watts.
Returns:
Real potential in dB.
"""
gain_pow = db2pow(gain_db)
loss_pow = db2pow(loss_db)
potential_tmp = pt * gain_pow * gain_pow * np.power(wavelength, 2) * loss_pow * F * F
potential = potential_tmp / (np.power(4 * np.pi, 3) * p_min)
return pow2db(potential)
def simulate_turbo_7_5():
## PCCC
#trellis = build_5_7_trellis(True)
#punc1 = odd_puncturer()
#punc2 = even_puncturer()
#rate = 1/2
#punc1 = no_puncturer()
#punc2 = no_puncturer()
## SCCC
#trellis1 = build_acc_trellis()
#trellis1 = build_7_1_trellis(False)
#trellis1 = trellis_7_5_nonsys()
trellis1 = build_7_5_trellis(True)
#trellis1 = build_7_5_trellis(False) # rate-1 INNER ENCODER TRELLIS - NON-SYSTEMATIC
trellis2 = build_7_5_trellis(True) # rate-1/2 OUTER ENCODER TRELLIS - SYSTEMATIC
#rate = 1
#rate = 1 / 4
rate = 1 / 2
max_runs = 1e4
max_iter = 8
k = 1000
#k = 128
n = int(k / rate)
#snrDbs = np.arange(2, 5)
snrDbs = [0.5, 1, 1.5, 1.8, 2, 2.2, 2.5, 2.7, 3.0, 3.5]
#snrDbs = [2.4]
#snrDbs = [4,5,6,7,8,9]
bers = np.zeros(len(snrDbs))
fers = np.zeros(len(snrDbs))
i = 0
for snrDb in snrDbs:
snr = db2pow(snrDb)
sigma = bawgn.noise_sigma(snr, rate)
#decode = lambda r : decide_from_llrs(decode_pccc_7_5(r, k, sigma, max_iter, trellis, punc1, punc2))
decode = lambda r : decide_from_llrs(decode_sccc_7_5(r, k, sigma, max_iter, trellis1, trellis2))
ber, fer = simulate_transmission(n, k, max_runs, bawgn.modulate, bawgn.transmit(sigma), decode)
bers[i] = ber
fers[i] = fer
i += 1
print(snrDb, ber, fer)
# save to mat file for plotting purposes
scipy.io.savemat('sccc_7_5_fancypunc', {
'sccc_7_5_fancypunc_snrs' : snrDbs,
'sccc_7_5_fancypunc_bers' : bers,
'sccc_7_5_fancypunc_fers' : fers
})
def radareqrange(potential_db, rcs, ro):
"""(3)Max distance.
Args:
potential_db (float): In dB
rcs (float): In m^2
ro (float): probability limit log(Pa)/log(Pd) ?????????? in dB??????
Returns:
"""
potential_rls_power = db2pow(potential_db)
return np.power(potential_rls_power * rcs / ro, 0.25)
def probability_limit(potential_db, distance, rcs):
"""(20)
Args:
potential_db: in dB
distance: in m
rcs: in m^2
Returns:
"""
potential_pow = db2pow(potential_db)
return potential_pow * rcs / np.power(distance, 4)
def std_angle(wavelength, aperture, alpha, snr_db, K=0.6):
"""(23, 24)
Args:
wavelength: in m
aperture: in m
alpha: in radian
snr_db: in dB
K:
Returns:
"""
beam_width = beam_width_by_aperture(wavelength, aperture, alpha, K)
snr_pow = db2pow(snr_db)
return beam_width / np.sqrt(snr_pow)
def main():
max_runs = 100000
snrDbs = np.arange(-2, 10)
bers = np.zeros(len(snrDbs))
i = 0
for snrDb in snrDbs:
snr = db2pow(snrDb)
sigma = bawgn.noise_sigma(snr, 1) # uncoded transmission
n = 100
ber = simulate_transmission(n, n, max_runs, bawgn.modulate,
bawgn.transmit(sigma),
bawgn.demodulate_hard)
bers[i] = ber
i += 1
th = norm.sf(1 / sigma)
print(snrDb, ber, th)
# save to mat file for plotting purposes
scipy.io.savemat('uncoded', {'uncsnrs': snrDbs, 'uncbers': bers})