def test_ofdm_rx(self):
z_out_test, H_test = (np.array([-3.11740028 - 0.90748269j, -3.11628187 - 0.88948888j,
2.88565859 + 1.13255112j, 2.89076997 + 3.16052588j,
2.90396853 + 1.19595053j, 2.93439648 + 1.23703401j,
-3.00724063 - 0.72880083j, 1.07519281 + 1.27075039j,
1.14472192 + 3.22099905j, -2.82962216 + 1.15148633j,
1.16245397 + 3.09533441j, -0.85799363 - 0.94063529j,
1.12036257 + 1.03825793j, 1.10109739 + 1.02622557j,
1.08488052 - 2.98041713j, 1.07132873 + 1.01625511j,
-0.92119499 + 3.01872286j, -2.91683903 - 0.9906338j,
-2.91213253 - 3.00295552j, 3.09229992 - 3.01974828j]),
np.array([1.42289223 - 1.43696423e-01j, 1.34580486 - 2.66705232e-01j,
1.23071709 - 3.51736667e-01j, 1.09530096 - 3.87688911e-01j,
0.95992898 - 3.71339473e-01j, 0.84428862 - 3.07656711e-01j,
0.76421410 - 2.08710707e-01j, 0.72928932 - 9.14213562e-02j,
0.74161551 + 2.54124114e-02j, 0.79590656 + 1.24244087e-01j,
0.88082345 + 1.91510354e-01j, 0.98123573 + 2.19504072e-01j,
1.08094630 + 2.07118520e-01j, 1.16536231 + 1.59418930e-01j,
1.22364402 + 8.62177170e-02j, 1.25000000 - 2.77555756e-17j,
1.25000000 + 2.77555756e-17j, 1.22364402 - 8.62177170e-02j,
1.16536231 - 1.59418930e-01j, 1.08094630 - 2.07118520e-01j,
0.98123573 - 2.19504072e-01j, 0.88082345 - 1.91510354e-01j,
0.79590656 - 1.24244087e-01j, 0.74161551 - 2.54124114e-02j,
0.72928932 + 9.14213562e-02j, 0.76421410 + 2.08710707e-01j,
0.84428862 + 3.07656711e-01j, 0.95992898 + 3.71339473e-01j,
1.09530096 + 3.87688911e-01j, 1.23071709 + 3.51736667e-01j,
1.34580486 + 2.66705232e-01j, 1.42289223 + 1.43696423e-01j]))
hc = np.array([1.0, 0.1, -0.05, 0.15, 0.2, 0.05])
x1, b1, IQ_data1 = dc.qam_bb(50000, 1, '16qam')
x_out = dc.ofdm_tx(IQ_data1, 32, 64, 0, True, 0)
c_out = signal.lfilter(hc, 1, x_out) # Apply channel distortion
r_out = dc.cpx_awgn(c_out, 100, 64 / 32) # Es/N0 = 100 dB
z_out, H = dc.ofdm_rx(r_out, 32, 64, -1, True, 0, alpha=0.95, ht=hc);
npt.assert_almost_equal(z_out[:20], z_out_test)
npt.assert_almost_equal(H, H_test)
def test_QAM_bb_qpsk_rc(self):
x_test, b_test, t_test = (np.array([ -2.22799382e-18 -2.22799382e-18j,
-4.07129297e-03 -4.07129297e-03j,
2.22160609e-19 +4.67814826e-18j,
-2.22059175e-03 +5.92199418e-03j,
6.43926133e-18 -2.91703518e-18j,
1.97462099e-02 +7.90222150e-03j,
-9.75189466e-18 -1.28297996e-17j,
-4.09888874e-02 -7.30785022e-02j,
1.05934616e-17 +3.71417032e-17j,
9.37971907e-02 +2.31071828e-01j,
-1.52549076e-18 -6.78758025e-17j,
-4.10719565e-01 -8.26448725e-01j,
-1.00000000e+00 -1.00000000e+00j,
-1.36231531e+00 +1.25147025e-01j,
-1.00000000e+00 +1.00000000e+00j,
4.09888874e-02 +2.75737546e-01j,
1.00000000e+00 -1.00000000e+00j,
1.24877191e+00 -1.36728049e+00j,
1.00000000e+00 -1.00000000e+00j, 8.23659721e-01 -7.64661456e-01j]),
np.array([ 1.11223990e-18, 2.03243583e-03, -1.22314501e-18,
-9.23891131e-04, -8.79167710e-19, -6.90120596e-03,
5.63651951e-18, 2.84718702e-02, -1.19149691e-17,
-8.10891577e-02, 1.73229581e-17, 3.08804252e-01,
4.99211393e-01, 3.08804252e-01, 1.73229581e-17,
-8.10891577e-02, -1.19149691e-17, 2.84718702e-02,
5.63651951e-18, -6.90120596e-03, -8.79167710e-19,
-9.23891131e-04, -1.22314501e-18, 2.03243583e-03,
1.11223990e-18]), np.array([-1.-1.j, -1.+1.j, 1.-1.j, 1.-1.j, 1.-1.j, 1.-1.j, -1.+1.j,
-1.-1.j, -1.-1.j, -1.+1.j]))
x, b, t = dc.qam_bb(10, 2, mod='qpsk', pulse='rc')
npt.assert_almost_equal(x, x_test)
npt.assert_almost_equal(b, b_test)
npt.assert_almost_equal(t, t_test)
def test_qam_sep_256qam(self):
Nsymb_test, Nerr_test, SEP_test = (4986, 2190, 0.43922984356197353)
x, b, tx_data = dc.qam_bb(5000, 10, '256qam', 'src')
x = dc.cpx_awgn(x, 20, 10)
y = signal.lfilter(b, 1, x)
Nsymb, Nerr, SEP = dc.qam_sep(tx_data, y[10 + 10 * 12::10], '256qam', n_transient=0)
self.assertEqual(Nsymb, Nsymb_test)
self.assertEqual(Nerr, Nerr_test)
self.assertEqual(SEP, SEP_test)
def test_qam_sep_16qam_error(self):
Nsymb_test, Nerr_test, SEP_test = (9976, 172, 0.017241379310344827)
x, b, tx_data = dc.qam_bb(10000, 1, '16qam', 'rect')
x = dc.cpx_awgn(x, 15, 1)
y = signal.lfilter(b, 1, x)
Nsymb, Nerr, SEP = dc.qam_sep(tx_data, y[1 * 12::1], '16qam', n_transient=0)
self.assertEqual(Nsymb, Nsymb_test)
self.assertEqual(Nerr, Nerr_test)
self.assertEqual(SEP, SEP_test)
def test_qam_sep_64qam(self):
Nsymb_test, Nerr_test, SEP_test = (4986, 245, 0.04913758523866827)
x, b, tx_data = dc.qam_bb(5000, 10, '64qam', 'src')
x = dc.cpx_awgn(x, 20, 10)
y = signal.lfilter(b, 1, x)
Nsymb, Nerr, SEP = dc.qam_sep(tx_data, y[10 + 10 * 12::10], '64qam', n_transient=0)
self.assertEqual(Nsymb, Nsymb_test)
self.assertEqual(Nerr, Nerr_test)
self.assertEqual(SEP, SEP_test)
def test_qam_sep_16qam_no_error(self):
Nsymb_test, Nerr_test, SEP_test = (4986, 0, 0.0)
x, b, tx_data = dc.qam_bb(5000, 10, '16qam', 'src')
x = dc.cpx_awgn(x, 20, 10)
y = signal.lfilter(b, 1, x)
Nsymb, Nerr, SEP = dc.qam_sep(tx_data, y[10 + 10 * 12::10], '16qam', n_transient=0)
self.assertEqual(Nsymb, Nsymb_test)
self.assertEqual(Nerr, Nerr_test)
self.assertEqual(SEP, SEP_test)
def test_QAM_bb_qpsk_rect(self):
x_test, b_test, t_test = (np.array([-1.-1.j, -1.-1.j, -1.+1.j, -1.+1.j, 1.-1.j, 1.-1.j, 1.-1.j,
1.-1.j, 1.-1.j, 1.-1.j, 1.-1.j, 1.-1.j, -1.+1.j, -1.+1.j,
-1.-1.j, -1.-1.j, -1.-1.j, -1.-1.j, -1.+1.j, -1.+1.j]), np.array([ 0.5, 0.5]),
np.array([-1.-1.j, -1.+1.j, 1.-1.j, 1.-1.j, 1.-1.j, 1.-1.j, -1.+1.j,
-1.-1.j, -1.-1.j, -1.+1.j]))
x, b, t = dc.qam_bb(10, 2, mod='qpsk', pulse='rect')
npt.assert_almost_equal(x, x_test)
npt.assert_almost_equal(b, b_test)
npt.assert_almost_equal(t, t_test)
def test_ofdm_rx_channel_estimate(self):
z_out_test, H_out_test = (np.array([-2.91356233-0.93854058j, -3.03083561-1.01177886j,
3.10687062+1.09962706j, 2.91679784+2.79392693j,
2.95621370+0.87789714j, 2.93521287+1.12869418j,
-3.17675560-1.0834705j , 1.25700626+1.19497994j,
1.16433902+2.62068101j, -3.10408334+1.08514004j,
1.02623864+3.01672402j, -0.98366297-1.21602375j,
0.89577012+1.07687508j, 1.05852406+1.05134363j,
0.93287609-3.11042385j, 0.99965390+0.88124784j,
-1.16293758+3.08562314j, -2.84891079-1.07199168j,
-3.22236927-2.90425199j, 3.07028549-2.88413491j,
-3.12192058+2.89625467j, 3.18017151-1.09375776j,
-2.78212772+3.05087219j, 1.13471595-2.89218144j,
-3.17092453-1.11298847j, 3.10927184+0.86801524j,
-0.76520964-3.32101721j, -0.94935570+2.86081052j,
0.93535950+1.10545223j, 1.09394518-1.17966519j,
3.10748055+1.12377382j, -3.12337017-0.89848715j,
-2.95725651+0.97491592j, 3.14041238-3.01998896j,
-1.05440640+3.04843936j, -0.94130790-0.82179287j,
-0.79049810-1.04083796j, 2.96004080+1.01692442j,
-3.13063510+1.32083138j, -2.58084447-3.28171534j,
3.09664605+0.82140179j, 2.87565015-1.17002378j,
2.82351021+2.83242155j, 2.99238994+3.06883778j,
-0.83601519-2.8886988j , 3.05383614+1.22402533j,
-0.92550302+0.92366226j, -0.97707573+3.08608891j,
0.73489228-2.99163649j, 2.89544691+2.76671634j]),
np.array([ 1.49261307-0.12886832j, 1.36399692-0.24831791j,
1.24438887-0.41524198j, 1.15276504-0.47480443j,
1.09981815-0.35438673j, 0.86684483-0.31710329j,
0.75885865-0.23542562j, 0.76309583-0.19374055j,
0.61556098+0.09731796j, 0.77281595+0.07096727j,
0.87593303+0.15642133j, 1.06728467+0.29788462j,
1.08613086+0.23650714j, 1.12082635+0.09129381j,
1.31026672+0.17419224j, 1.19459330+0.01027668j,
1.19745209+0.11471611j, 1.36689249-0.07997548j,
1.26471663-0.07505238j, 1.14356226-0.19961235j,
0.84149706-0.21609579j, 0.85489994-0.18101042j,
0.79502365-0.17155484j, 0.71666634-0.02650505j,
0.82384118+0.0565963j , 0.74313589+0.28403893j,
0.88570493+0.29345603j, 0.95203301+0.37888469j,
0.98676887+0.4108844j , 1.26869289+0.35672436j,
1.44594176+0.3296819j , 1.48817425+0.07577518j]))
hc = np.array([1.0, 0.1, -0.05, 0.15, 0.2, 0.05])
x1, b1, IQ_data1 = dc.qam_bb(50000, 1, '16qam')
x_out = dc.ofdm_tx(IQ_data1, 32, 64, 100, True, 10)
c_out = signal.lfilter(hc, 1, x_out) # Apply channel distortion
r_out = dc.cpx_awgn(c_out, 25, 64 / 32) # Es/N0 = 25 dB
z_out, H = dc.ofdm_rx(r_out, 32, 64, 100, True, 10, alpha=0.95, ht=hc)
npt.assert_almost_equal(z_out[:50], z_out_test)
npt.assert_almost_equal(H[:50], H_out_test)
def test_QAM_bb_256qam_rect(self):
x_test, b_test, t_test = (np.array([ 0.06666667-0.73333333j, 0.06666667-0.73333333j,
0.06666667-0.33333333j, 0.06666667-0.33333333j,
-0.60000000-0.73333333j, -0.60000000-0.73333333j,
-0.06666667-0.73333333j, -0.06666667-0.73333333j,
-0.06666667+0.86666667j, -0.06666667+0.86666667j,
1.00000000-0.73333333j, 1.00000000-0.73333333j,
-1.00000000-0.86666667j, -1.00000000-0.86666667j,
0.33333333-1.j , 0.33333333-1.j ,
0.86666667+0.06666667j, 0.86666667+0.06666667j,
-0.46666667+1.j , -0.46666667+1.j ]), np.array([ 0.5, 0.5]),
np.array([ 1.-11.j, 1. -5.j, -9.-11.j, -1.-11.j, -1.+13.j, 15.-11.j,
-15.-13.j, 5.-15.j, 13. +1.j, -7.+15.j]))
x, b, t = dc.qam_bb(10, 2, mod='256qam', pulse='rect')
npt.assert_almost_equal(x, x_test)
npt.assert_almost_equal(b, b_test)
npt.assert_almost_equal(t, t_test)
def test_QAM_bb_64qam_rect(self):
x_test, b_test, t_test = (np.array([-1.00000000-0.42857143j, -1.00000000-0.42857143j,
-1.00000000+0.42857143j, -1.00000000+0.42857143j,
-0.14285714-0.42857143j, -0.14285714-0.42857143j,
1.00000000-0.42857143j, 1.00000000-0.42857143j,
1.00000000+0.71428571j, 1.00000000+0.71428571j,
1.00000000-0.42857143j, 1.00000000-0.42857143j,
-1.00000000-0.71428571j, -1.00000000-0.71428571j,
-0.42857143-1.j , -0.42857143-1.j ,
0.71428571-1.j , 0.71428571-1.j ,
0.14285714+1.j , 0.14285714+1.j ]), np.array([ 0.5, 0.5]),
np.array([-7.-3.j, -7.+3.j, -1.-3.j, 7.-3.j, 7.+5.j, 7.-3.j, -7.-5.j,
-3.-7.j, 5.-7.j, 1.+7.j]))
x, b, t = dc.qam_bb(10, 2, mod='64qam', pulse='rect')
npt.assert_almost_equal(x, x_test)
npt.assert_almost_equal(b, b_test)
npt.assert_almost_equal(t, t_test)
def test_QAM_bb_16qam_rect(self):
x_test, b_test, t_test = (np.array([-1.00000000+0.33333333j, -1.00000000+0.33333333j,
-1.00000000-0.33333333j, -1.00000000-0.33333333j,
1.00000000+0.33333333j, 1.00000000+0.33333333j,
1.00000000+0.33333333j, 1.00000000+0.33333333j,
1.00000000+0.33333333j, 1.00000000+0.33333333j,
1.00000000+0.33333333j, 1.00000000+0.33333333j,
-1.00000000-0.33333333j, -1.00000000-0.33333333j,
0.33333333-1.j , 0.33333333-1.j ,
0.33333333-1.j , 0.33333333-1.j ,
-1.00000000+1.j , -1.00000000+1.j ]), np.array([ 0.5, 0.5]),
np.array([-3.+1.j, -3.-1.j, 3.+1.j, 3.+1.j, 3.+1.j, 3.+1.j, -3.-1.j,
1.-3.j, 1.-3.j, -3.+3.j]))
x, b, t = dc.qam_bb(10, 2, mod='16qam', pulse='rect')
npt.assert_almost_equal(x, x_test)
npt.assert_almost_equal(b, b_test)
npt.assert_almost_equal(t, t_test)
def test_ofdm_tx(self):
x_out_test = np.array([ 0.00000000+0.125j, -0.10185331+0.27369942j, -0.10291586+0.12529202j,
-0.05485981-0.1015143j, -0.02143872-0.09787268j, -0.06906044+0.05231368j,
-0.18815224+0.050888j, -0.26164122-0.15836327j, -0.21940048-0.36048543j,
-0.14486054-0.38169759j, -0.11830476-0.25561157j, -0.07250935-0.12760226j,
0.05301567-0.08413918j, 0.14316564-0.07020723j, 0.07590886+0.01736066j,
-0.04551924+0.15686941j, -0.03125000+0.21875j, 0.09755018+0.17168517j,
0.15431728+0.10974492j, 0.08889087+0.04259743j, 0.04284671-0.1107734j,
0.10071736-0.25986197j, 0.15582045-0.17226253j, 0.06652251+0.12312402j,
-0.15245874+0.29798543j, -0.32346606+0.23845079j, -0.25311017+0.21460293j,
0.07831717+0.3396657j, 0.43085592+0.30360811j, 0.48116320-0.0505655j,
0.16656460-0.32765262j, -0.20071609-0.16142259j, -0.25000000+0.1875j,
0.04290155+0.25900306j, 0.33313987+0.08484705j, 0.28478134-0.00986648j,
-0.05936711+0.00190181j, -0.30195965-0.0628197j, -0.12280721-0.1651266j,
0.31807654-0.16252886j, 0.53190048-0.13951457j, 0.31342228-0.20065005j,
0.00806130-0.17969398j, -0.00105255+0.03378639j, 0.15279016+0.16494501j,
0.09844557-0.009236j, -0.11589986-0.20597693j, -0.10438721-0.09983656j,
0.15625000+0.09375j, 0.22805837+0.03951473j,])
x1, b1, IQ_data1 = dc.qam_bb(50000, 1, '16qam')
x_out = dc.ofdm_tx(IQ_data1, 32, 64, 0, True, 0)
npt.assert_almost_equal(x_out[:50], x_out_test)
def test_QAM_bb_qpsk_src(self):
np.random.seed(100)
x_test, b_test, t_test = (np.array([ 0.00585723+0.00585723j, -0.00275016-0.00275016j,
-0.00164540-0.01335987j, 0.00887646+0.01437677j,
-0.01540288+0.01131686j, 0.00480440-0.02394915j,
0.02505607+0.02585128j, -0.04406383-0.00716616j,
-0.02797722-0.08626139j, 0.11024504+0.1600832j ,
0.00580570+0.16357483j, -0.44629859-0.76924864j,
-0.96387506-1.22739262j, -1.32990076+0.11435352j,
-1.06357060+1.2446656j , 0.04406383+0.22076409j,
1.06649175-1.18402117j, 1.21485132-1.24832839j,
0.97347224-0.94165619j, 0.88372072-0.89875699j]),
np.array([-0.00293625, 0.00137866, 0.00376109, -0.00582846, 0.00102417,
0.00479866, -0.01276 , 0.01284087, 0.02863407, -0.06775815,
-0.04245547, 0.30467864, 0.54924435, 0.30467864, -0.04245547,
-0.06775815, 0.02863407, 0.01284087, -0.01276 , 0.00479866,
0.00102417, -0.00582846, 0.00376109, 0.00137866, -0.00293625]),
np.array([-1.-1.j, -1.+1.j, 1.-1.j, 1.-1.j, 1.-1.j, 1.-1.j, -1.+1.j,
-1.-1.j, -1.-1.j, -1.+1.j]))
x, b, t = dc.qam_bb(10, 2, mod='qpsk', pulse='src')
npt.assert_almost_equal(x, x_test)
npt.assert_almost_equal(b, b_test)
npt.assert_almost_equal(t, t_test)
def test_QAM_bb_mod_error(self):
with self.assertRaisesRegexp(ValueError, 'Unknown mod_type'):
x, b, t = dc.qam_bb(10, 2, mod='unknown')
def test_QAM_bb_pulse_error(self):
with self.assertRaisesRegexp(ValueError, 'pulse shape must be src, rc, or rect'):
dc.qam_bb(10, 2, pulse='value')