mirror of
https://github.com/saitohirga/WSJT-X.git
synced 2024-08-21 14:06:32 -04:00
git-svn-id: svn+ssh://svn.code.sf.net/p/wsjt/wsjt/branches/map65@466 ab8295b8-cf94-4d9e-aec4-7959e3be5d79
135 lines
4.1 KiB
Fortran
135 lines
4.1 KiB
Fortran
subroutine decode1a(id,newdat,nfilt,freq,nflip,
|
|
+ mycall,hiscall,hisgrid,neme,ndepth,nqd,dphi,ndphi,
|
|
+ ipol,sync2,a,dt,pol,nkv,nhist,qual,decoded)
|
|
|
|
C Apply AFC corrections to a candidate JT65 signal, and then try
|
|
C to decode it.
|
|
|
|
parameter (NFFT1=77760,NFFT2=2430)
|
|
parameter (NMAX=60*96000) !Samples per 60 s
|
|
integer*2 id(4,NMAX) !46 MB: raw data from Linrad timf2
|
|
complex c2x(NMAX/4), c2y(NMAX/4) !After 1/4 filter and downsample
|
|
complex c3x(NMAX/16),c3y(NMAX/16) !After 1/16 filter and downsample
|
|
complex c4x(NMAX/64),c4y(NMAX/64) !After 1/64 filter and downsample
|
|
complex cx(NMAX/64), cy(NMAX/64) !Data at 1378.125 samples/s
|
|
complex c5x(NMAX/256),c5y(NMAX/256),c5y0(NMAX/256)
|
|
complex c5a(256), c5b(256)
|
|
complex z
|
|
|
|
real s2(256,126)
|
|
real a(5)
|
|
real*8 samratio
|
|
integer resample
|
|
logical first
|
|
character decoded*22
|
|
character mycall*12,hiscall*12,hisgrid*6
|
|
data first/.true./,jjjmin/1000/,jjjmax/-1000/
|
|
save
|
|
|
|
C Mix sync tone to baseband, low-pass filter, and decimate by 64
|
|
dt00=dt
|
|
C If freq=125.0 kHz, f0=48000 Hz.
|
|
f0=1000*(freq-77.0) !Freq of sync tone (0-96000 Hz)
|
|
if(nfilt.eq.1) then
|
|
call filbig(id,NMAX,f0,newdat,cx,cy,n5)
|
|
joff=0
|
|
else
|
|
call fil659(id,NMAX,f0,c2x,c2y,n2) !Pass 1: mix and filter both pol'ns
|
|
call fil658(c2x,n2,c3x,n3) !Pass 2
|
|
call fil658(c2y,n2,c3y,n3)
|
|
call fil658(c3x,n3,c4x,n4) !Pass 3
|
|
call fil658(c3y,n3,c4y,n4)
|
|
joff=-8
|
|
|
|
C Resample from 96000/64 = 1500 Hz to 1378.125 Hz
|
|
C Converter type: 0=Best quality sinc (band limited), BW=97%
|
|
C 1=medium quality sinc, BW=90%
|
|
C 2=fastest sinc, BW=80%
|
|
C 3=stepwise (very fast)
|
|
C 4=linear (very fast)
|
|
nconv_type=2 !### test! ###
|
|
nchans=2
|
|
samratio=1378.125d0/1500.d0
|
|
i1=resample(c4x,n4,nconv_type,nchans,samratio,cx,n5)
|
|
i2=resample(c4y,n4,nconv_type,nchans,samratio,cy,n5)
|
|
endif
|
|
|
|
sqa=0.
|
|
sqb=0.
|
|
do i=1,n5
|
|
sqa=sqa + real(cx(i))**2 + aimag(cx(i))**2
|
|
sqb=sqb + real(cy(i))**2 + aimag(cy(i))**2
|
|
enddo
|
|
sqa=sqa/n5
|
|
sqb=sqb/n5
|
|
|
|
C Find best DF, f1, f2, DT, and pol
|
|
|
|
! a(5)=dt00
|
|
! fsample=1378.125
|
|
! i0=nint((a(5)+0.5)*fsample)
|
|
! if(i0.lt.1) i0=1
|
|
! nz=n5+1-i0
|
|
! call afc65b(cx(i0),cy(i0),nz,fsample,nflip,ipol,a,dt,
|
|
! + ccfbest,dtbest)
|
|
|
|
call fil6521(cx,n5,c5x,n6)
|
|
call fil6521(cy,n5,c5y0,n6)
|
|
|
|
! Adjust for cable length difference:
|
|
z=cmplx(cos(dphi),sin(dphi))
|
|
do i=1,n6
|
|
c5y(i)=z*c5y0(i)
|
|
enddo
|
|
|
|
fsample=1378.125/4.
|
|
a(5)=dt00
|
|
i0=nint((a(5)+0.5)*fsample) - 2
|
|
if(i0.lt.1) i0=1
|
|
nz=n6+1-i0
|
|
|
|
call afc65b(c5x(i0),c5y(i0),nz,fsample,nflip,ipol,a,dt,
|
|
+ ccfbest,dtbest)
|
|
|
|
pol=a(4)/57.2957795
|
|
aa=cos(pol)
|
|
bb=sin(pol)
|
|
sq0=aa*aa*sqa + bb*bb*sqb
|
|
sync2=3.7*ccfbest/sq0
|
|
|
|
C Apply AFC corrections to the time-domain signal
|
|
call twkfreq(cx,cy,n5,a)
|
|
|
|
C Compute spectrum at best polarization for each half symbol.
|
|
C Adding or subtracting a small number (e.g., 5) to j may make it decode.
|
|
nsym=126
|
|
nfft=256
|
|
j=(dt00+dtbest+2.685)*1378.125 + joff
|
|
if(j.lt.0) j=0
|
|
j0=j
|
|
do k=1,nsym
|
|
do i=1,nfft
|
|
j=j+1
|
|
c5a(i)=aa*cx(j) + bb*cy(j)
|
|
enddo
|
|
call four2a(c5a,nfft,1,1,1)
|
|
do i=1,nfft
|
|
j=j+1
|
|
c5b(i)=aa*cx(j) + bb*cy(j)
|
|
enddo
|
|
call four2a(c5b,nfft,1,1,1)
|
|
|
|
do i=1,256
|
|
s2(i,k)=real(c5a(i))**2 + aimag(c5a(i))**2 +
|
|
+ real(c5b(i))**2 + aimag(c5b(i))**2
|
|
enddo
|
|
enddo
|
|
|
|
flip=nflip
|
|
call decode65b(s2,flip,mycall,hiscall,hisgrid,neme,ndepth,
|
|
+ nqd,nkv,nhist,qual,decoded)
|
|
dt=dt00 + dtbest
|
|
|
|
return
|
|
end
|