def test_002_t(self):
#OQPSK, no frequency offset
mod = dsss_modulator()
mod.shr_goldcode_reset_per_symbol = True
mod.shr_goldcode_seed = np.random.randint(500, 500000)
mod.shr_spreading_factor = 8
mod.sfd_present = False
mod.preamble_size = 32
mod.init()
sps = 2
sf = 2**mod.shr_spreading_factor
#init preamble block
self.preamble_search = dsss_preamble_search_cc(
sf, mod.shr_goldcode_seed, 0, 0, [0], mod.preamble_size, False,
sps, 0, 1000000, 'oqpsk', [0])
#generate data with preamble only
preamble = mod.oqpsk(mod.encode_header())
n = 6
data_in = []
positions = []
for i in range(n):
data_in = np.append(data_in,
np.zeros(np.random.randint(10000, 20000)))
data_in = np.append(data_in, preamble)
positions = np.append(positions,
len(data_in) +
self.preamble_search.rx_delay -
1) #delay in data signal in filterbank
data_in = np.append(data_in, np.zeros(np.random.randint(10000, 20000)))
data_in = data_in + np.random.randn(len(data_in)) * 5
# set up fg
self.src = blocks.vector_source_c_make(data_in, False, 1, [])
self.snk_corr = blocks.vector_sink_c_make(1)
self.snk_data = blocks.vector_sink_c_make(1)
self.tb.connect(self.src, self.preamble_search)
self.tb.connect((self.preamble_search, 0), self.snk_data)
self.tb.connect((self.preamble_search, 1), self.snk_corr)
self.tb.run()
# check data
tags = self.snk_data.tags()
searched_positions = np.array([])
for i in range(len(tags)):
searched_positions = np.append(searched_positions,
tags[i].offset + sf)
found = np.in1d(positions, searched_positions.astype(int))
self.assertTrue(np.all(found))
def test_006_t(self):
#OQPSK, frequency offset, multiple frequency paths
mod = dsss_modulator()
mod.shr_goldcode_reset_per_symbol = True
mod.shr_goldcode_seed = np.random.randint(500, 500000)
mod.shr_spreading_factor = 8
mod.sfd_present = False
mod.preamble_size = 32
mod.init()
sps = 2
sf = 2**mod.shr_spreading_factor
chiprate = 2000000
fo = 8000 # Hz
#init preamble block
self.preamble_search = dsss_preamble_search_cc(
sf, mod.shr_goldcode_seed, 0, 0, [0, 7900, -7900],
mod.preamble_size, False, sps, 0, chiprate, 'oqpsk', [0])
#generate data with preamble only
preamble = mod.oqpsk(mod.encode_header())
n = 20
data_in = []
positions = []
for i in range(n):
data_in = np.append(data_in,
np.zeros(np.random.randint(10000, 20000)))
data_in = np.append(data_in, preamble)
positions = np.append(positions,
len(data_in) +
self.preamble_search.rx_delay -
1) #delay in data signal in filterbank
data_in = np.append(data_in, np.zeros(np.random.randint(10000, 20000)))
#frequency offset
data_in = data_in * np.exp(
1j * 2.0 * np.pi * fo / chiprate * np.arange(0, len(data_in)))
#add some noise
data_in = data_in + np.random.randn(len(data_in)) * 1
# set up fg
self.src = blocks.vector_source_c_make(data_in, False, 1, [])
self.snk_corr = blocks.vector_sink_c_make(1)
self.snk_data = blocks.vector_sink_c_make(1)
self.tb.connect(self.src, self.preamble_search)
self.tb.connect((self.preamble_search, 0), self.snk_data)
self.tb.connect((self.preamble_search, 1), self.snk_corr)
self.tb.run()
# check data
#time offset estimation
tags = self.snk_data.tags()
searched_positions = np.array([])
for i in range(len(tags)):
searched_positions = np.append(searched_positions,
tags[i].offset + sf)
found = np.in1d(positions, searched_positions.astype(int))
self.assertTrue(np.all(found))
#freq offset estimation
mask = np.in1d(searched_positions.astype(int), positions)
v = np.asarray(tags)[mask]
for t in v:
self.assertTrue(pmt.to_python(t.value)['fo_est'] < fo * 1.2)
self.assertTrue(pmt.to_python(t.value)['fo_est'] > fo * 0.8)