% [X] tune perf/impl loss to get closer to ideal
% [X] linear interp of phase for better fading perf
% [X] freq offset/drift feedback loop
-% [ ] PAPR measurement and reduction
-% [ ] false sync
-% [ ] noise
-% [ ] similar but invalid signal like huge f off
-% [ ] check "unsync"
+% [X] PAPR measurement and reduction
+% [X] false sync
+% [X] doesn't sync up on noise (used EsNo = -12)
+% [X] similar but invalid signal like huge f off
+% [X] ability to "unsync" when signal dissapears
% [ ] some calibrated tests against FreeDV 1600
% + compare sound quality at various Es/Nos
+% [ ] way to handle eom w/o nasties
+% + like mute ouput when signal has gone or v low snr
+% + instantaneous snr
+% [ ] nasty rig filter passband
graphics_toolkit ("gnuplot");
more off;
clf
subplot(211)
plot(real(tx_fdm_frame_log))
+ title('tx fdm real');
subplot(212)
plot(imag(tx_fdm_frame_log))
- title('tx fdm');
+ title('tx fdm imag');
figure(2)
clf;
figure(7)
clf
- plot(tx_fdm_frame_log)
+ spec = 20*log10(abs(fft(tx_fdm_frame_log)));
+ l = length(spec);
+ plot((Fs/l)*(1:l), spec)
+ axis([1 Fs/2 0 max(spec)]);
+ title('tx spectrum');
+ ylabel('Amplitude (dB)')
+ xlabel('Frequency (Hz)')
end
endfunction\r
\r
\r
-freq_off_est_test_single;\r
-%freq_off_est_test_curves;\r
+%freq_off_est_test_single;\r
+freq_off_est_test_curves;\r
\r
% 1. start with +/- 20Hz offset\r
% 2. Measure frames to sync. How to define sync? Foff to withn 1 Hz. Sync state\r