getting back into sync with C version of fdmdv, 21/23 tests passing

git-svn-id: https://svn.code.sf.net/p/freetel/code@1651 01035d8c-6547-0410-b346-abe4f91aad63

diff --git a/codec2-dev/octave/fdmdv.m b/codec2-dev/octave/fdmdv.m
index 34e3586b0f30c88cccbb6ee647d2fd71cdaf5c07..4640a7fa035fbd7e8f9134f78ca9b362d70fba74 100644 (file)
--- a/codec2-dev/octave/fdmdv.m
+++ b/codec2-dev/octave/fdmdv.m
@@ -217,6 +217,14 @@ function tx_fdm = fdm_upconvert(tx_filt)
   % shifting for the purpose of testing easier
 
   tx_fdm = 2*tx_fdm;
+
+  % normalise digital oscilators as the magnitude can drift over time
+
+  for c=1:Nc+1
+    mag = abs(phase_tx(c));
+    phase_tx(c) /= mag;
+  end
+
 endfunction
 
 
@@ -290,9 +298,10 @@ endfunction
 
 % LPF and peak pick part of freq est, put in a function as we call it twice
 
-function [foff imax pilot_lpf S] = lpf_peak_pick(pilot_baseband, pilot_lpf, nin)
+function [foff imax pilot_lpf_out S] = lpf_peak_pick(pilot_baseband, pilot_lpf, nin)
   global M;
   global Npilotlpf;
+  global Npilotbaseband;
   global Npilotcoeff;
   global Fs;
   global Mpilotfft;
@@ -301,9 +310,10 @@ function [foff imax pilot_lpf S] = lpf_peak_pick(pilot_baseband, pilot_lpf, nin)
   % LPF cutoff 200Hz, so we can handle max +/- 200 Hz freq offset
 
   pilot_lpf(1:Npilotlpf-nin) = pilot_lpf(nin+1:Npilotlpf);
-  j = 1;
+  j = Npilotcoeff+1;
   for i = Npilotlpf-nin+1:Npilotlpf
-    pilot_lpf(i) = pilot_baseband(j:j+Npilotcoeff-1) * pilot_coeff';
+    pilot_lpf(i) = pilot_baseband(j-Npilotcoeff+1:j) * pilot_coeff';
+    %pilot_lpf(i) = pilot_baseband(j-Npilotcoeff+1);
     j++;
   end
 
@@ -326,6 +336,8 @@ function [foff imax pilot_lpf S] = lpf_peak_pick(pilot_baseband, pilot_lpf, nin)
     foff = (ix - 1)*r;
   endif
 
+  pilot_lpf_out = pilot_lpf;
+
 endfunction
 
 
@@ -340,10 +352,10 @@ function [foff S1 S2] = rx_est_freq_offset(rx_fdm, pilot, pilot_prev, nin)
   global pilot_lpf1;
   global pilot_lpf2;
 
-  % down convert latest nin samples of pilot by multiplying by
-  % ideal BPSK pilot signal we have generated locally.  This
-  % peak of the resulting signal is sensitive to the time shift between 
-  % the received and local version of the pilot, so we do it twice at
+  % down convert latest nin samples of pilot by multiplying by ideal
+  % BPSK pilot signal we have generated locally.  The peak of the DFT
+  % of the resulting signal is sensitive to the time shift between the
+  % received and local version of the pilot, so we do it twice at
   % different time shifts and choose the maximum.
  
   pilot_baseband1(1:Npilotbaseband-nin) = pilot_baseband1(nin+1:Npilotbaseband);

diff --git a/codec2-dev/octave/tfdmdv.m b/codec2-dev/octave/tfdmdv.m
index abadc41eaa2921933f5f5a41529c27602019d051..ed0368ebc126cd4381f1765a690366980c3837ce 100644 (file)
--- a/codec2-dev/octave/tfdmdv.m
+++ b/codec2-dev/octave/tfdmdv.m
@@ -9,7 +9,7 @@
 % Version 2
 %
 
-NumCarriers = 20;
+NumCarriers = 14;
 fdmdv; % load modem code
  
 % Generate reference vectors using Octave implementation of FDMDV modem
@@ -23,7 +23,7 @@ prev_rx_symbols = ones(Nc+1,1);
 foff_phase_rect = 1;
 coarse_fine = 0;
 fest_state = 0;
-channel = zeros(1,10*M);
+channel = [];
 channel_count = 0;
 next_nin = M;
 sig_est = zeros(Nc+1,1);
@@ -82,6 +82,7 @@ for f=1:frames
   %else
   %  nin = M;
   %end
+  nin = M;
   channel = [channel real(tx_fdm)];
   channel_count += M;
   rx_fdm = channel(1:nin);
@@ -93,7 +94,9 @@ for f=1:frames
   [pilot prev_pilot pilot_lut_index prev_pilot_lut_index] = get_pilot(pilot_lut_index, prev_pilot_lut_index, nin);
 
   [foff_coarse S1 S2] = rx_est_freq_offset(rx_fdm, pilot, prev_pilot, nin);
-  if coarse_fine == 0
+  foff_coarse = 0;
+  sync = 0;
+  if sync == 0
     foff = foff_coarse;
   end
   foff_coarse_log = [foff_coarse_log foff_coarse];
@@ -148,8 +151,8 @@ for f=1:frames
 
   % freq est state machine
 
-  [coarse_fine fest_state] = freq_state(sync_bit, fest_state);
-  coarse_fine_log = [coarse_fine_log coarse_fine];
+  [sync fest_state] = freq_state(sync_bit, fest_state);
+  sync_log = [coarse_fine_log sync];
 end
 
 % Compare to the output from the C version
@@ -161,9 +164,9 @@ load ../unittest/tfdmdv_out.txt
 function stem_sig_and_error(plotnum, subplotnum, sig, error, titlestr, axisvec)
   figure(plotnum)
   subplot(subplotnum)
-  stem(sig);
+  stem(sig,'g;C version;');
   hold on;
-  stem(error,'g');
+  stem(error,'r;Error between C and Octave;');
   hold off;
   if nargin == 6
     axis(axisvec);
@@ -174,9 +177,9 @@ endfunction
 function plot_sig_and_error(plotnum, subplotnum, sig, error, titlestr, axisvec)
   figure(plotnum)
   subplot(subplotnum)
-  plot(sig);
+  plot(sig,'g;C version;');
   hold on;
-  plot(error,'g');
+  plot(error,'r;Error between C and Octave;');
   hold off;
   if nargin == 6
     axis(axisvec);
@@ -197,9 +200,9 @@ stem_sig_and_error(1, 212, real(tx_symbols_log_c(1:n/2)), real(tx_symbols_log(1:
 % tx_filter()
 
 diff = tx_baseband_log - tx_baseband_log_c;
-c=15;
-plot_sig_and_error(2, 211, real(tx_baseband_log_c(c,:)), real(sum(diff)), 'tx baseband real')
-plot_sig_and_error(2, 212, imag(tx_baseband_log_c(c,:)), imag(sum(diff)), 'tx baseband imag')
+c=1;
+plot_sig_and_error(2, 211, real(tx_baseband_log_c(c,:)), real(tx_baseband_log(c,:) - tx_baseband_log_c(c,:)), 'tx baseband real')
+plot_sig_and_error(2, 212, imag(tx_baseband_log_c(c,:)), imag(tx_baseband_log(c,:) - tx_baseband_log_c(c,:)), 'tx baseband imag')
 
 % fdm_upconvert()
 
@@ -213,11 +216,11 @@ plot_sig_and_error(4, 212, imag(pilot_lut_c), imag(pilot_lut - pilot_lut_c), 'pi
 
 % rx_est_freq_offset()
 
-st=1;  en = 3*Npilotbaseband;
+st=1;  en = 5*Npilotbaseband;
 plot_sig_and_error(5, 211, real(pilot_baseband1_log(st:en)), real(pilot_baseband1_log(st:en) - pilot_baseband1_log_c(st:en)), 'pilot baseband1 real' )
 plot_sig_and_error(5, 212, real(pilot_baseband2_log(st:en)), real(pilot_baseband2_log(st:en) - pilot_baseband2_log_c(st:en)), 'pilot baseband2 real' )
 
-st=1;  en = 3*Npilotlpf;
+st=1;  en = 5*Npilotlpf;
 plot_sig_and_error(6, 211, real(pilot_lpf1_log(st:en)), real(pilot_lpf1_log(st:en) - pilot_lpf1_log_c(st:en)), 'pilot lpf1 real' )
 plot_sig_and_error(6, 212, real(pilot_lpf2_log(st:en)), real(pilot_lpf2_log(st:en) - pilot_lpf2_log_c(st:en)), 'pilot lpf2 real' )
 
@@ -231,7 +234,7 @@ plot_sig_and_error(9, 211, foff_coarse_log, foff_coarse_log - foff_coarse_log_c,
 plot_sig_and_error(9, 212, foff_fine_log, foff_fine_log - foff_fine_log_c, 'Fine Freq Offset' )
 
 plot_sig_and_error(10, 211, foff_log, foff_log - foff_log_c, 'Freq Offset' )
-plot_sig_and_error(10, 212, coarse_fine_log, coarse_fine_log - coarse_fine_log_c, 'Freq Est Coarse(0) Fine(1)', [1 frames -0.5 1.5] )
+plot_sig_and_error(10, 212, sync_log, sync_log - sync_log_c, 'Freq Est Coarse(0) Fine(1)', [1 frames -0.5 1.5] )
 
 c=15;
 plot_sig_and_error(11, 211, real(rx_baseband_log(c,:)), real(rx_baseband_log(c,:) - rx_baseband_log_c(c,:)), 'Rx baseband real' )
@@ -271,7 +274,8 @@ function check(a, b, test_name)
   end
   printf(": ");  
   
-  if sum(abs(a - b))/n < 1E-3
+  e = sum(abs(a - b))/n;
+  if e < 1E-3
     printf("OK\n");
     passes++;
   else
@@ -299,9 +303,8 @@ check(rx_timing_log, rx_timing_log_c, 'rx_timing');
 check(rx_symbols_log, rx_symbols_log_c, 'rx_symbols');
 check(rx_bits_log, rx_bits_log_c, 'rx bits');
 check(sync_bit_log, sync_bit_log_c, 'sync bit');
-check(coarse_fine_log, coarse_fine_log_c, 'coarse_fine');
+check(sync_log, sync_log_c, 'sync');
 check(nin_log, nin_log_c, 'nin');
 check(sig_est_log, sig_est_log_c, 'sig_est');
 check(noise_est_log, noise_est_log_c, 'noise_est');
-
 printf("\npasses: %d fails: %d\n", passes, fails);