def __init__(self, vlen, channel_block, delay):
gr.hier_block2.__init__(self, 'apply_channel_to_vect',
gr.io_signature(1, 1, gr.sizeof_gr_complex * vlen),
gr.io_signature(1, 1, gr.sizeof_gr_complex * vlen))
self.vlen=vlen
self.channel_block = channel_block
self.delay = delay
pad_len = self.vlen + self.delay*2
if(delay>0):
self.padding = gr.vector_source_c((0,)*(delay*2),True,delay*2)
self.pad_cat = mlse.vector_concat_vv(vlen*gr.sizeof_gr_complex,gr.sizeof_gr_complex*2*delay)
self.connect(self, self.pad_cat)
self.connect(self.padding, (self.pad_cat,1))
else:
self.pad_cat = self
self.tostream = gr.vector_to_stream(gr.sizeof_gr_complex, pad_len)
self.connect(self.pad_cat, self.tostream)
# connect channel
self.connect(self.tostream, self.channel_block)
self.tovector = gr.stream_to_vector(gr.sizeof_gr_complex, pad_len)
self.connect(self.channel_block, self.tovector)
# cut out the proper part
self.trim = mlse.vector_slice_vv(
gr.sizeof_gr_complex,
pad_len,
delay,
vlen)
self.connect(self.tovector, self.trim, self)
def __init__(self, vlen, channel_block, delay):
gr.hier_block2.__init__(
self, 'apply_channel_to_vect',
gr.io_signature(1, 1, gr.sizeof_gr_complex * vlen),
gr.io_signature(1, 1, gr.sizeof_gr_complex * vlen))
self.vlen = vlen
self.channel_block = channel_block
self.delay = delay
pad_len = self.vlen + self.delay * 2
if (delay > 0):
self.padding = gr.vector_source_c((0, ) * (delay * 2), True,
delay * 2)
self.pad_cat = mlse.vector_concat_vv(
vlen * gr.sizeof_gr_complex, gr.sizeof_gr_complex * 2 * delay)
self.connect(self, self.pad_cat)
self.connect(self.padding, (self.pad_cat, 1))
else:
self.pad_cat = self
self.tostream = gr.vector_to_stream(gr.sizeof_gr_complex, pad_len)
self.connect(self.pad_cat, self.tostream)
# connect channel
self.connect(self.tostream, self.channel_block)
self.tovector = gr.stream_to_vector(gr.sizeof_gr_complex, pad_len)
self.connect(self.channel_block, self.tovector)
# cut out the proper part
self.trim = mlse.vector_slice_vv(gr.sizeof_gr_complex, pad_len, delay,
vlen)
self.connect(self.tovector, self.trim, self)
def test_001_vector_concat_vv (self): src1=gr.vector_source_f((1,2,3,4,5)); src2=gr.vector_source_f((11,12,13)); expected_result = (1,2,3,4,5,11,12,13) dst=gr.vector_sink_f() reor1=gr.stream_to_vector(gr.sizeof_float, 5) reor2=gr.stream_to_vector(gr.sizeof_float, 3) reor3=gr.vector_to_stream(gr.sizeof_float, 8) cat=mlse_swig.vector_concat_vv(gr.sizeof_float*5, gr.sizeof_float*3) self.tb.connect(src1,reor1,(cat,0)) self.tb.connect(src2,reor2,(cat,1)) self.tb.connect(cat,reor3,dst) self.tb.run() self.tb.run () result_data = dst.data () self.assertFloatTuplesAlmostEqual (expected_result, result_data, 6)
def test_001_vector_concat_vv(self):
src1 = gr.vector_source_f((1, 2, 3, 4, 5))
src2 = gr.vector_source_f((11, 12, 13))
expected_result = (1, 2, 3, 4, 5, 11, 12, 13)
dst = gr.vector_sink_f()
reor1 = gr.stream_to_vector(gr.sizeof_float, 5)
reor2 = gr.stream_to_vector(gr.sizeof_float, 3)
reor3 = gr.vector_to_stream(gr.sizeof_float, 8)
cat = mlse_swig.vector_concat_vv(gr.sizeof_float * 5,
gr.sizeof_float * 3)
self.tb.connect(src1, reor1, (cat, 0))
self.tb.connect(src2, reor2, (cat, 1))
self.tb.connect(cat, reor3, dst)
self.tb.run()
self.tb.run()
result_data = dst.data()
self.assertFloatTuplesAlmostEqual(expected_result, result_data, 6)
def __init__(self):
gr.hier_block2.__init__(self, 'gsm_decode_harddecision',
gr.io_signature(1, 1, gr.sizeof_gr_complex * 148 ),
gr.io_signature(1,1,58*2))
conf = mlse.make_packet_config_gsm()
slice1 = conf.make_slicer(conf.PAYLOAD_FRONT)
slice2 = conf.make_slicer(conf.PAYLOAD_REAR)
cat = mlse.vector_concat_vv(58*gr.sizeof_gr_complex, 58*gr.sizeof_gr_complex)
tostream = gr.vector_to_stream(gr.sizeof_gr_complex, 58*2)
real = gr.complex_to_real()
sign = gr.binary_slicer_fb()
demap = gr.map_bb([1,0]) # invert 0 and 1 because in gsm 0 is mapped to +1 and 1 to -1
tovect = gr.stream_to_vector(1, 58*2)
self.connect(self, slice1)
self.connect(self, slice2)
self.connect(slice1, (cat,0))
self.connect(slice2, (cat,1))
self.connect(cat, tostream, real, sign, demap, tovect, self)
def __init__(self, samples_per_symbol=1):
gr.hier_block2.__init__(self, 'gsm_modulate_burst',
gr.io_signature2(2, 2, 58 * 2, 1),
gr.io_signature(1, 1, gr.sizeof_gr_complex))
# take care of the packetizer
conf = mlse.make_packet_config_gsm()
self.packetbuilder = mlse.packetbuilder_midamble_vbb(conf)
self.connect(self, self.packetbuilder)
self.connect((self, 1), (self.packetbuilder, 1)) # tsc_index
# append 8 1-bits to the packet for the idle-time between bursts.
# This is necessary among others to flush the gmsk-modulator which is causal.
# (We use 1-bits because that's what the gsm-standard says)
# (we also only use 8 bits, because the specified value of 8.25 only works
# properly with samplerates that are a multiple of 4.)
self.guardbits = gr.vector_source_b((1, ) * 8, True, 8)
self.guard_cat = mlse.vector_concat_vv(148, 8)
self.connect(self.packetbuilder, self.guard_cat)
self.connect(self.guardbits, (self.guard_cat, 1))
# we now have a vector of bits, transform that into a stream
self.tostream = gr.vector_to_stream(1, 148 + 8)
# do the precoding:
self.diffcode = gr.diff_decoder_bb(2)
self.nrz = gr.map_bb([1, -1])
self.connect(self.guard_cat, self.tostream, self.diffcode, self.nrz)
# modulate
self.mod = digital.gmskmod_bc(samples_per_symbol, 0.3, 8)
# skip the first gmsk_length*samplerate/2 samples to make this block
# acausal.
# self.skiphead = gr.skiphead(gr.sizeof_gr_complex, int(samples_per_symbol*4.5));
# self.connect(self.nrz, self.mod, self.skiphead, self)
self.connect(self.nrz, self.mod,
self) # workaround: we need a negative delay later,
def __init__(self):
gr.hier_block2.__init__(
self, 'gsm_decode_harddecision',
gr.io_signature(1, 1, gr.sizeof_gr_complex * 148),
gr.io_signature(1, 1, 58 * 2))
conf = mlse.make_packet_config_gsm()
slice1 = conf.make_slicer(conf.PAYLOAD_FRONT)
slice2 = conf.make_slicer(conf.PAYLOAD_REAR)
cat = mlse.vector_concat_vv(58 * gr.sizeof_gr_complex,
58 * gr.sizeof_gr_complex)
tostream = gr.vector_to_stream(gr.sizeof_gr_complex, 58 * 2)
real = gr.complex_to_real()
sign = gr.binary_slicer_fb()
demap = gr.map_bb(
[1,
0]) # invert 0 and 1 because in gsm 0 is mapped to +1 and 1 to -1
tovect = gr.stream_to_vector(1, 58 * 2)
self.connect(self, slice1)
self.connect(self, slice2)
self.connect(slice1, (cat, 0))
self.connect(slice2, (cat, 1))
self.connect(cat, tostream, real, sign, demap, tovect, self)
def __init__(self, samples_per_symbol = 1):
gr.hier_block2.__init__(self, 'gsm_modulate_burst',
gr.io_signature2(2,2,58*2,1),
gr.io_signature(1, 1, gr.sizeof_gr_complex))
# take care of the packetizer
conf = mlse.make_packet_config_gsm()
self.packetbuilder = mlse.packetbuilder_midamble_vbb(conf)
self.connect(self, self.packetbuilder)
self.connect((self,1), (self.packetbuilder,1)) # tsc_index
# append 8 1-bits to the packet for the idle-time between bursts.
# This is necessary among others to flush the gmsk-modulator which is causal.
# (We use 1-bits because that's what the gsm-standard says)
# (we also only use 8 bits, because the specified value of 8.25 only works
# properly with samplerates that are a multiple of 4.)
self.guardbits = gr.vector_source_b((1,)*8,True,8)
self.guard_cat = mlse.vector_concat_vv(148,8)
self.connect(self.packetbuilder, self.guard_cat)
self.connect(self.guardbits, (self.guard_cat,1))
# we now have a vector of bits, transform that into a stream
self.tostream = gr.vector_to_stream(1, 148+8)
# do the precoding:
self.diffcode = gr.diff_decoder_bb(2);
self.nrz = gr.map_bb([1,-1])
self.connect(self.guard_cat, self.tostream, self.diffcode, self.nrz)
# modulate
self.mod = digital.gmskmod_bc(samples_per_symbol, 0.3, 8)
# skip the first gmsk_length*samplerate/2 samples to make this block
# acausal.
# self.skiphead = gr.skiphead(gr.sizeof_gr_complex, int(samples_per_symbol*4.5));
# self.connect(self.nrz, self.mod, self.skiphead, self)
self.connect(self.nrz, self.mod, self) # workaround: we need a negative delay later,
