def run_flow_graph(sync_sym1, sync_sym2, data_sym):
top_block = gr.top_block()
carr_offset = random.randint(-max_offset/2, max_offset/2) * 2
tx_data = shift_tuple(sync_sym1, carr_offset) + \
shift_tuple(sync_sym2, carr_offset) + \
shift_tuple(data_sym, carr_offset)
channel = [rand_range(min_chan_ampl, max_chan_ampl) * numpy.exp(1j * rand_range(0, 2 * numpy.pi)) for x in range(fft_len)]
src = gr.vector_source_c(tx_data, False, fft_len)
chan= gr.multiply_const_vcc(channel)
noise = gr.noise_source_c(gr.GR_GAUSSIAN, wgn_amplitude)
add = gr.add_cc(fft_len)
chanest = digital.ofdm_chanest_vcvc(sync_sym1, sync_sym2, 1)
sink = gr.vector_sink_c(fft_len)
top_block.connect(src, chan, (add, 0), chanest, sink)
top_block.connect(noise, gr.stream_to_vector(gr.sizeof_gr_complex, fft_len), (add, 1))
top_block.run()
channel_est = None
carr_offset_hat = 0
rx_sym_est = [0,] * fft_len
tags = sink.tags()
for tag in tags:
if pmt.pmt_symbol_to_string(tag.key) == 'ofdm_sync_carr_offset':
carr_offset_hat = pmt.pmt_to_long(tag.value)
self.assertEqual(carr_offset, carr_offset_hat)
if pmt.pmt_symbol_to_string(tag.key) == 'ofdm_sync_chan_taps':
channel_est = shift_tuple(pmt.pmt_c32vector_elements(tag.value), carr_offset)
shifted_carrier_mask = shift_tuple(carrier_mask, carr_offset)
for i in range(fft_len):
if shifted_carrier_mask[i] and channel_est[i]:
self.assertAlmostEqual(channel[i], channel_est[i], places=0)
rx_sym_est[i] = (sink.data()[i] / channel_est[i]).real
return (carr_offset, list(shift_tuple(rx_sym_est, -carr_offset_hat)))
def test_004_channel_no_carroffset_1sym (self):
""" Add a channel, check if it's correctly estimated.
Only uses 1 synchronisation symbol. """
fft_len = 16
carr_offset = 0
sync_symbol = (0, 0, 0, 1, 0, 1, 0, -1, 0, 1, 0, -1, 0, 1, 0, 0)
data_symbol = (0, 0, 0, 1, -1, 1, -1, 1, 0, 1, -1, -1, -1, 1, 0, 0)
tx_data = sync_symbol + data_symbol
channel = (0, 0, 0, 2, 2, 2, 2, 3, 3, 2.5, 2.5, -3, -3, 1j, 1j, 0)
#channel = (0, 0, 0, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 0)
src = blocks.vector_source_c(tx_data, False, fft_len)
chan = blocks.multiply_const_vcc(channel)
chanest = digital.ofdm_chanest_vcvc(sync_symbol, (), 1)
sink = blocks.vector_sink_c(fft_len)
sink_chanest = blocks.vector_sink_c(fft_len)
self.tb.connect(src, chan, chanest, sink)
self.tb.connect((chanest, 1), sink_chanest)
self.tb.run()
self.assertEqual(sink_chanest.data(), channel)
tags = sink.tags()
for tag in tags:
if pmt.pmt_symbol_to_string(tag.key) == 'ofdm_sync_carr_offset':
self.assertEqual(pmt.pmt_to_long(tag.value), carr_offset)
if pmt.pmt_symbol_to_string(tag.key) == 'ofdm_sync_chan_taps':
self.assertEqual(pmt.pmt_c32vector_elements(tag.value), channel)
def test_006_channel_and_carroffset (self):
""" Add a channel, check if it's correctly estimated """
fft_len = 16
carr_offset = 2
# Index 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
sync_symbol1 = (0, 0, 0, 1, 0, 1, 0, -1, 0, 1, 0, -1, 0, 1, 0, 0)
sync_symbol2 = (0, 0, 0, 1j, -1, 1, -1j, 1j, 0, 1, -1j, -1, -1j, 1, 0, 0)
data_symbol = (0, 0, 0, 1, -1, 1, -1, 1, 0, 1, -1, -1, -1, 1, 0, 0)
# Channel 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
# Shifted (0, 0, 0, 0, 0, 1j, -1, 1, -1j, 1j, 0, 1, -1j, -1, -1j, 1)
chanest_exp = (0, 0, 0, 5, 6, 7, 8, 9, 0, 11, 12, 13, 14, 15, 0, 0)
tx_data = shift_tuple(sync_symbol1, carr_offset) + \
shift_tuple(sync_symbol2, carr_offset) + \
shift_tuple(data_symbol, carr_offset)
channel = range(fft_len)
src = gr.vector_source_c(tx_data, False, fft_len)
chan = gr.multiply_const_vcc(channel)
chanest = digital.ofdm_chanest_vcvc(sync_symbol1, sync_symbol2, 1)
sink = gr.vector_sink_c(fft_len)
self.tb.connect(src, chan, chanest, sink)
self.tb.run()
tags = sink.tags()
chan_est = None
for tag in tags:
if pmt.pmt_symbol_to_string(tag.key) == 'ofdm_sync_carr_offset':
self.assertEqual(pmt.pmt_to_long(tag.value), carr_offset)
if pmt.pmt_symbol_to_string(tag.key) == 'ofdm_sync_chan_taps':
chan_est = pmt.pmt_c32vector_elements(tag.value)
self.assertEqual(chan_est, chanest_exp)
self.assertEqual(sink.data(), tuple(numpy.multiply(shift_tuple(data_symbol, carr_offset), channel)))
def test_006_channel_and_carroffset(self):
""" Add a channel, check if it's correctly estimated """
fft_len = 16
carr_offset = 2
# Index 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
sync_symbol1 = (0, 0, 0, 1, 0, 1, 0, -1, 0, 1, 0, -1, 0, 1, 0, 0)
sync_symbol2 = (0, 0, 0, 1j, -1, 1, -1j, 1j, 0, 1, -1j, -1, -1j, 1, 0,
0)
data_symbol = (0, 0, 0, 1, -1, 1, -1, 1, 0, 1, -1, -1, -1, 1, 0, 0)
# Channel 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
# Shifted (0, 0, 0, 0, 0, 1j, -1, 1, -1j, 1j, 0, 1, -1j, -1, -1j, 1)
chanest_exp = (0, 0, 0, 5, 6, 7, 8, 9, 0, 11, 12, 13, 14, 15, 0, 0)
tx_data = shift_tuple(sync_symbol1, carr_offset) + \
shift_tuple(sync_symbol2, carr_offset) + \
shift_tuple(data_symbol, carr_offset)
channel = range(fft_len)
src = gr.vector_source_c(tx_data, False, fft_len)
chan = gr.multiply_const_vcc(channel)
chanest = digital.ofdm_chanest_vcvc(sync_symbol1, sync_symbol2, 1)
sink = gr.vector_sink_c(fft_len)
self.tb.connect(src, chan, chanest, sink)
self.tb.run()
tags = sink.tags()
chan_est = None
for tag in tags:
if pmt.pmt_symbol_to_string(tag.key) == 'ofdm_sync_carr_offset':
self.assertEqual(pmt.pmt_to_long(tag.value), carr_offset)
if pmt.pmt_symbol_to_string(tag.key) == 'ofdm_sync_chan_taps':
chan_est = pmt.pmt_c32vector_elements(tag.value)
self.assertEqual(chan_est, chanest_exp)
self.assertEqual(
sink.data(),
tuple(
numpy.multiply(shift_tuple(data_symbol, carr_offset),
channel)))
def test_003_channel_no_carroffset (self):
""" Add a channel, check if it's correctly estimated """
fft_len = 16
carr_offset = 0
sync_symbol1 = (0, 0, 0, 1, 0, 1, 0, -1, 0, 1, 0, -1, 0, 1, 0, 0)
sync_symbol2 = (0, 0, 0, 1j, -1, 1, -1j, 1j, 0, 1, -1j, -1, -1j, 1, 0, 0)
data_symbol = (0, 0, 0, 1, -1, 1, -1, 1, 0, 1, -1, -1, -1, 1, 0, 0)
tx_data = sync_symbol1 + sync_symbol2 + data_symbol
channel = (0, 0, 0, 2, -2, 2, 3j, 2, 0, 2, 2, 2, 2, 3, 0, 0)
src = gr.vector_source_c(tx_data, False, fft_len)
chan = gr.multiply_const_vcc(channel)
chanest = digital.ofdm_chanest_vcvc(sync_symbol1, sync_symbol2, 1)
sink = gr.vector_sink_c(fft_len)
sink_chanest = gr.vector_sink_c(fft_len)
self.tb.connect(src, chan, chanest, sink)
self.tb.connect((chanest, 1), sink_chanest)
self.tb.run()
tags = sink.tags()
self.assertEqual(shift_tuple(sink.data(), -carr_offset), tuple(numpy.multiply(data_symbol, channel)))
for tag in tags:
if pmt.pmt_symbol_to_string(tag.key) == 'ofdm_sync_carr_offset':
self.assertEqual(pmt.pmt_to_long(tag.value), carr_offset)
if pmt.pmt_symbol_to_string(tag.key) == 'ofdm_sync_chan_taps':
self.assertEqual(pmt.pmt_c32vector_elements(tag.value), channel)
self.assertEqual(sink_chanest.data(), channel)
def test_004_channel_no_carroffset_1sym(self):
""" Add a channel, check if it's correctly estimated.
Only uses 1 synchronisation symbol. """
fft_len = 16
carr_offset = 0
sync_symbol = (0, 0, 0, 1, 0, 1, 0, -1, 0, 1, 0, -1, 0, 1, 0, 0)
data_symbol = (0, 0, 0, 1, -1, 1, -1, 1, 0, 1, -1, -1, -1, 1, 0, 0)
tx_data = sync_symbol + data_symbol
channel = (0, 0, 0, 2, 2, 2, 2, 3, 3, 2.5, 2.5, -3, -3, 1j, 1j, 0)
#channel = (0, 0, 0, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 0)
src = blocks.vector_source_c(tx_data, False, fft_len)
chan = blocks.multiply_const_vcc(channel)
chanest = digital.ofdm_chanest_vcvc(sync_symbol, (), 1)
sink = blocks.vector_sink_c(fft_len)
sink_chanest = blocks.vector_sink_c(fft_len)
self.tb.connect(src, chan, chanest, sink)
self.tb.connect((chanest, 1), sink_chanest)
self.tb.run()
self.assertEqual(sink_chanest.data(), channel)
tags = sink.tags()
for tag in tags:
if pmt.pmt_symbol_to_string(tag.key) == 'ofdm_sync_carr_offset':
self.assertEqual(pmt.pmt_to_long(tag.value), carr_offset)
if pmt.pmt_symbol_to_string(tag.key) == 'ofdm_sync_chan_taps':
self.assertEqual(pmt.pmt_c32vector_elements(tag.value),
channel)
def test_003_channel_no_carroffset(self):
""" Add a channel, check if it's correctly estimated """
fft_len = 16
carr_offset = 0
sync_symbol1 = (0, 0, 0, 1, 0, 1, 0, -1, 0, 1, 0, -1, 0, 1, 0, 0)
sync_symbol2 = (0, 0, 0, 1j, -1, 1, -1j, 1j, 0, 1, -1j, -1, -1j, 1, 0,
0)
data_symbol = (0, 0, 0, 1, -1, 1, -1, 1, 0, 1, -1, -1, -1, 1, 0, 0)
tx_data = sync_symbol1 + sync_symbol2 + data_symbol
channel = (0, 0, 0, 2, -2, 2, 3j, 2, 0, 2, 2, 2, 2, 3, 0, 0)
src = gr.vector_source_c(tx_data, False, fft_len)
chan = gr.multiply_const_vcc(channel)
chanest = digital.ofdm_chanest_vcvc(sync_symbol1, sync_symbol2, 1)
sink = gr.vector_sink_c(fft_len)
sink_chanest = gr.vector_sink_c(fft_len)
self.tb.connect(src, chan, chanest, sink)
self.tb.connect((chanest, 1), sink_chanest)
self.tb.run()
tags = sink.tags()
self.assertEqual(shift_tuple(sink.data(), -carr_offset),
tuple(numpy.multiply(data_symbol, channel)))
for tag in tags:
if pmt.pmt_symbol_to_string(tag.key) == 'ofdm_sync_carr_offset':
self.assertEqual(pmt.pmt_to_long(tag.value), carr_offset)
if pmt.pmt_symbol_to_string(tag.key) == 'ofdm_sync_chan_taps':
self.assertEqual(pmt.pmt_c32vector_elements(tag.value),
channel)
self.assertEqual(sink_chanest.data(), channel)
def test_002_simpledfe (self):
""" Use the simple DFE equalizer. """
fft_len = 8
# 4 5 6 7 0 1 2 3
tx_data = [-1, -1, 1, 2, -1, 3, 0, -1, # 0
-1, -1, 0, 2, -1, 2, 0, -1, # 8
-1, -1, 3, 0, -1, 1, 0, -1, # 16 (Pilot symbols)
-1, -1, 1, 1, -1, 0, 2, -1] # 24
cnst = digital.constellation_qpsk()
tx_signal = [cnst.map_to_points_v(x)[0] if x != -1 else 0 for x in tx_data]
occupied_carriers = ((1, 2, 6, 7),)
pilot_carriers = ((), (), (1, 2, 6, 7), ())
pilot_symbols = (
[], [], [cnst.map_to_points_v(x)[0] for x in (1, 0, 3, 0)], []
)
equalizer = digital.ofdm_equalizer_simpledfe(
fft_len, cnst.base(), occupied_carriers, pilot_carriers, pilot_symbols, 0, 0.01
)
channel = [
0, 0, 1, 1, 0, 1, 1, 0,
0, 0, 1, 1, 0, 1, 1, 0, # These coefficients will be rotated slightly...
0, 0, 1j, 1j, 0, 1j, 1j, 0, # Go crazy here!
0, 0, 1j, 1j, 0, 1j, 1j, 0 # ...and again here.
]
for idx in range(fft_len, 2*fft_len):
channel[idx] = channel[idx-fft_len] * numpy.exp(1j * .1 * numpy.pi * (numpy.random.rand()-.5))
idx2 = idx+2*fft_len
channel[idx2] = channel[idx2] * numpy.exp(1j * 0 * numpy.pi * (numpy.random.rand()-.5))
len_tag_key = "frame_len"
len_tag = gr.gr_tag_t()
len_tag.offset = 0
len_tag.key = pmt.pmt_string_to_symbol(len_tag_key)
len_tag.value = pmt.pmt_from_long(4)
chan_tag = gr.gr_tag_t()
chan_tag.offset = 0
chan_tag.key = pmt.pmt_string_to_symbol("ofdm_sync_chan_taps")
chan_tag.value = pmt.pmt_init_c32vector(fft_len, channel[:fft_len])
src = gr.vector_source_c(numpy.multiply(tx_signal, channel), False, fft_len, (len_tag, chan_tag))
eq = digital.ofdm_frame_equalizer_vcvc(equalizer.base(), 0, len_tag_key, True)
sink = gr.vector_sink_c(fft_len)
self.tb.connect(src, eq, sink)
self.tb.run ()
rx_data = [cnst.decision_maker_v((x,)) if x != 0 else -1 for x in sink.data()]
self.assertEqual(tx_data, rx_data)
for tag in sink.tags():
if pmt.pmt_symbol_to_string(tag.key) == len_tag_key:
self.assertEqual(pmt.pmt_to_long(tag.value), 4)
if pmt.pmt_symbol_to_string(tag.key) == "ofdm_sync_chan_taps":
self.assertComplexTuplesAlmostEqual(list(pmt.pmt_c32vector_elements(tag.value)), channel[-fft_len:], places=1)
def run_flow_graph(sync_sym1, sync_sym2, data_sym):
top_block = gr.top_block()
carr_offset = random.randint(-max_offset / 2, max_offset / 2) * 2
tx_data = shift_tuple(sync_sym1, carr_offset) + \
shift_tuple(sync_sym2, carr_offset) + \
shift_tuple(data_sym, carr_offset)
channel = [
rand_range(min_chan_ampl, max_chan_ampl) *
numpy.exp(1j * rand_range(0, 2 * numpy.pi))
for x in range(fft_len)
]
src = gr.vector_source_c(tx_data, False, fft_len)
chan = gr.multiply_const_vcc(channel)
noise = gr.noise_source_c(gr.GR_GAUSSIAN, wgn_amplitude)
add = gr.add_cc(fft_len)
chanest = digital.ofdm_chanest_vcvc(sync_sym1, sync_sym2, 1)
sink = gr.vector_sink_c(fft_len)
top_block.connect(src, chan, (add, 0), chanest, sink)
top_block.connect(
noise, gr.stream_to_vector(gr.sizeof_gr_complex, fft_len),
(add, 1))
top_block.run()
channel_est = None
carr_offset_hat = 0
rx_sym_est = [
0,
] * fft_len
tags = sink.tags()
for tag in tags:
if pmt.pmt_symbol_to_string(
tag.key) == 'ofdm_sync_carr_offset':
carr_offset_hat = pmt.pmt_to_long(tag.value)
self.assertEqual(carr_offset, carr_offset_hat)
if pmt.pmt_symbol_to_string(tag.key) == 'ofdm_sync_chan_taps':
channel_est = shift_tuple(
pmt.pmt_c32vector_elements(tag.value), carr_offset)
shifted_carrier_mask = shift_tuple(carrier_mask, carr_offset)
for i in range(fft_len):
if shifted_carrier_mask[i] and channel_est[i]:
self.assertAlmostEqual(channel[i],
channel_est[i],
places=0)
rx_sym_est[i] = (sink.data()[i] / channel_est[i]).real
return (carr_offset, list(shift_tuple(rx_sym_est,
-carr_offset_hat)))
def test_002_static (self):
"""
- Add a simple channel
- Make symbols QPSK
"""
fft_len = 8
# 4 5 6 7 0 1 2 3
tx_data = [-1, -1, 1, 2, -1, 3, 0, -1, # 0
-1, -1, 0, 2, -1, 2, 0, -1, # 8
-1, -1, 3, 0, -1, 1, 0, -1, # 16 (Pilot symbols)
-1, -1, 1, 1, -1, 0, 2, -1] # 24
cnst = digital.constellation_qpsk()
tx_signal = [cnst.map_to_points_v(x)[0] if x != -1 else 0 for x in tx_data]
occupied_carriers = ((1, 2, 6, 7),)
pilot_carriers = ((), (), (1, 2, 6, 7), ())
pilot_symbols = (
[], [], [cnst.map_to_points_v(x)[0] for x in (1, 0, 3, 0)], []
)
equalizer = digital.ofdm_equalizer_static(fft_len, occupied_carriers, pilot_carriers, pilot_symbols)
channel = [
0, 0, 1, 1, 0, 1, 1, 0,
0, 0, 1, 1, 0, 1, 1, 0, # These coefficients will be rotated slightly (but less than \pi/2)
0, 0, 1j, 1j, 0, 1j, 1j, 0, # Go crazy here!
0, 0, 1j, 1j, 0, 1j, 1j, 0
]
channel = [
0, 0, 1, 1, 0, 1, 1, 0,
0, 0, 1, 1, 0, 1, 1, 0, # These coefficients will be rotated slightly (but less than \pi/2)
0, 0, 1j, 1j, 0, 1j, 1j, 0, # Go crazy here!
0, 0, 1j, 1j, 0, 1j, 1j, 0
]
for idx in range(fft_len, 2*fft_len):
channel[idx] = channel[idx-fft_len] * numpy.exp(1j * .1 * numpy.pi * (numpy.random.rand()-.5))
len_tag_key = "frame_len"
len_tag = gr.gr_tag_t()
len_tag.offset = 0
len_tag.key = pmt.pmt_string_to_symbol(len_tag_key)
len_tag.value = pmt.pmt_from_long(4)
chan_tag = gr.gr_tag_t()
chan_tag.offset = 0
chan_tag.key = pmt.pmt_string_to_symbol("ofdm_sync_chan_taps")
chan_tag.value = pmt.pmt_init_c32vector(fft_len, channel[:fft_len])
src = gr.vector_source_c(numpy.multiply(tx_signal, channel), False, fft_len, (len_tag, chan_tag))
eq = digital.ofdm_frame_equalizer_vcvc(equalizer.base(), 0, len_tag_key, True)
sink = gr.vector_sink_c(fft_len)
self.tb.connect(src, eq, sink)
self.tb.run ()
rx_data = [cnst.decision_maker_v((x,)) if x != 0 else -1 for x in sink.data()]
# Check data
self.assertEqual(tx_data, rx_data)
# Check tags
tag_dict = dict()
for tag in sink.tags():
ptag = gr.tag_to_python(tag)
tag_dict[ptag.key] = ptag.value
if ptag.key == 'ofdm_sync_chan_taps':
tag_dict[ptag.key] = list(pmt.pmt_c32vector_elements(tag.value))
else:
tag_dict[ptag.key] = pmt.to_python(tag.value)
expected_dict = {
'frame_len': 4,
'ofdm_sync_chan_taps': channel[-fft_len:]
}
self.assertEqual(tag_dict, expected_dict)
def test_002_simpledfe(self):
""" Use the simple DFE equalizer. """
fft_len = 8
# 4 5 6 7 0 1 2 3
tx_data = [
-1,
-1,
1,
2,
-1,
3,
0,
-1, # 0
-1,
-1,
0,
2,
-1,
2,
0,
-1, # 8
-1,
-1,
3,
0,
-1,
1,
0,
-1, # 16 (Pilot symbols)
-1,
-1,
1,
1,
-1,
0,
2,
-1
] # 24
cnst = digital.constellation_qpsk()
tx_signal = [
cnst.map_to_points_v(x)[0] if x != -1 else 0 for x in tx_data
]
occupied_carriers = ((1, 2, 6, 7), )
pilot_carriers = ((), (), (1, 2, 6, 7), ())
pilot_symbols = ([], [],
[cnst.map_to_points_v(x)[0]
for x in (1, 0, 3, 0)], [])
equalizer = digital.ofdm_equalizer_simpledfe(fft_len, cnst.base(),
occupied_carriers,
pilot_carriers,
pilot_symbols, 0, 0.01)
channel = [
0,
0,
1,
1,
0,
1,
1,
0,
0,
0,
1,
1,
0,
1,
1,
0, # These coefficients will be rotated slightly...
0,
0,
1j,
1j,
0,
1j,
1j,
0, # Go crazy here!
0,
0,
1j,
1j,
0,
1j,
1j,
0 # ...and again here.
]
for idx in range(fft_len, 2 * fft_len):
channel[idx] = channel[idx - fft_len] * numpy.exp(
1j * .1 * numpy.pi * (numpy.random.rand() - .5))
idx2 = idx + 2 * fft_len
channel[idx2] = channel[idx2] * numpy.exp(
1j * 0 * numpy.pi * (numpy.random.rand() - .5))
len_tag_key = "frame_len"
len_tag = gr.gr_tag_t()
len_tag.offset = 0
len_tag.key = pmt.pmt_string_to_symbol(len_tag_key)
len_tag.value = pmt.pmt_from_long(4)
chan_tag = gr.gr_tag_t()
chan_tag.offset = 0
chan_tag.key = pmt.pmt_string_to_symbol("ofdm_sync_chan_taps")
chan_tag.value = pmt.pmt_init_c32vector(fft_len, channel[:fft_len])
src = gr.vector_source_c(numpy.multiply(tx_signal, channel), False,
fft_len, (len_tag, chan_tag))
eq = digital.ofdm_frame_equalizer_vcvc(equalizer.base(), 0,
len_tag_key, True)
sink = gr.vector_sink_c(fft_len)
self.tb.connect(src, eq, sink)
self.tb.run()
rx_data = [
cnst.decision_maker_v((x, )) if x != 0 else -1
for x in sink.data()
]
self.assertEqual(tx_data, rx_data)
for tag in sink.tags():
if pmt.pmt_symbol_to_string(tag.key) == len_tag_key:
self.assertEqual(pmt.pmt_to_long(tag.value), 4)
if pmt.pmt_symbol_to_string(tag.key) == "ofdm_sync_chan_taps":
self.assertComplexTuplesAlmostEqual(list(
pmt.pmt_c32vector_elements(tag.value)),
channel[-fft_len:],
places=1)
def test_002_static(self):
"""
- Add a simple channel
- Make symbols QPSK
"""
fft_len = 8
# 4 5 6 7 0 1 2 3
tx_data = [
-1,
-1,
1,
2,
-1,
3,
0,
-1, # 0
-1,
-1,
0,
2,
-1,
2,
0,
-1, # 8
-1,
-1,
3,
0,
-1,
1,
0,
-1, # 16 (Pilot symbols)
-1,
-1,
1,
1,
-1,
0,
2,
-1
] # 24
cnst = digital.constellation_qpsk()
tx_signal = [
cnst.map_to_points_v(x)[0] if x != -1 else 0 for x in tx_data
]
occupied_carriers = ((1, 2, 6, 7), )
pilot_carriers = ((), (), (1, 2, 6, 7), ())
pilot_symbols = ([], [],
[cnst.map_to_points_v(x)[0]
for x in (1, 0, 3, 0)], [])
equalizer = digital.ofdm_equalizer_static(fft_len, occupied_carriers,
pilot_carriers,
pilot_symbols)
channel = [
0,
0,
1,
1,
0,
1,
1,
0,
0,
0,
1,
1,
0,
1,
1,
0, # These coefficients will be rotated slightly (but less than \pi/2)
0,
0,
1j,
1j,
0,
1j,
1j,
0, # Go crazy here!
0,
0,
1j,
1j,
0,
1j,
1j,
0
]
channel = [
0,
0,
1,
1,
0,
1,
1,
0,
0,
0,
1,
1,
0,
1,
1,
0, # These coefficients will be rotated slightly (but less than \pi/2)
0,
0,
1j,
1j,
0,
1j,
1j,
0, # Go crazy here!
0,
0,
1j,
1j,
0,
1j,
1j,
0
]
for idx in range(fft_len, 2 * fft_len):
channel[idx] = channel[idx - fft_len] * numpy.exp(
1j * .1 * numpy.pi * (numpy.random.rand() - .5))
len_tag_key = "frame_len"
len_tag = gr.gr_tag_t()
len_tag.offset = 0
len_tag.key = pmt.pmt_string_to_symbol(len_tag_key)
len_tag.value = pmt.pmt_from_long(4)
chan_tag = gr.gr_tag_t()
chan_tag.offset = 0
chan_tag.key = pmt.pmt_string_to_symbol("ofdm_sync_chan_taps")
chan_tag.value = pmt.pmt_init_c32vector(fft_len, channel[:fft_len])
src = gr.vector_source_c(numpy.multiply(tx_signal, channel), False,
fft_len, (len_tag, chan_tag))
eq = digital.ofdm_frame_equalizer_vcvc(equalizer.base(), 0,
len_tag_key, True)
sink = gr.vector_sink_c(fft_len)
self.tb.connect(src, eq, sink)
self.tb.run()
rx_data = [
cnst.decision_maker_v((x, )) if x != 0 else -1
for x in sink.data()
]
# Check data
self.assertEqual(tx_data, rx_data)
# Check tags
tag_dict = dict()
for tag in sink.tags():
ptag = gr.tag_to_python(tag)
tag_dict[ptag.key] = ptag.value
if ptag.key == 'ofdm_sync_chan_taps':
tag_dict[ptag.key] = list(pmt.pmt_c32vector_elements(
tag.value))
else:
tag_dict[ptag.key] = pmt.to_python(tag.value)
expected_dict = {
'frame_len': 4,
'ofdm_sync_chan_taps': channel[-fft_len:]
}
self.assertEqual(tag_dict, expected_dict)