WSJT-X/lib/timf2.f90

226 lines
6.5 KiB
Fortran

subroutine timf2(k,nxpol,nfft,nwindow,nb,peaklimit,iqadjust,iqapply,faclim, &
cx0,cy0,gainx,gainy,phasex,phasey,cx1,cy1,slimit,lstrong,px,py,nzap)
! Sequential processing of time-domain I/Q data, using Linrad-like
! "first FFT" and "first backward FFT".
! cx0,cy0 - complex input data
! nfft - length of FFTs
! nwindow - 0 for no window, 2 for sin^2 window
! iqapply - 0/1 determines if I/Q phase and amplitude corrections applied
! gainx,y - gain error in Q channel, relative to I
! phasex,y - phase error
! cx1,cy1 - output data
! Non-windowed processing means no overlap, so kstep=nfft.
! Sin^2 window has 50% overlap, kstep=nfft/2.
! Frequencies with strong signals are identified and separated. The back
! transforms are done separately for weak and strong signals, so that
! noise blanking can be applied to the weak-signal portion. Strong and
! weak are finally re-combined in the time domain.
parameter (MAXFFT=1024,MAXNH=MAXFFT/2)
parameter (MAXSIGS=100)
complex cx0(0:nfft-1),cx1(0:nfft-1)
complex cy0(0:nfft-1),cy1(0:nfft-1)
complex cx(0:MAXFFT-1),cxt(0:MAXFFT-1)
complex cy(0:MAXFFT-1),cyt(0:MAXFFT-1)
complex cxs(0:MAXFFT-1),covxs(0:MAXNH-1) !Strong X signals
complex cys(0:MAXFFT-1),covys(0:MAXNH-1) !Strong Y signals
complex cxw(0:MAXFFT-1),covxw(0:MAXNH-1) !Weak X signals
complex cyw(0:MAXFFT-1),covyw(0:MAXNH-1) !Weak Y signals
real*4 w(0:MAXFFT-1)
real*4 s(0:MAXFFT-1),stmp(0:MAXFFT-1)
logical*1 lstrong(0:MAXFFT-1),lprev
integer ia(MAXSIGS),ib(MAXSIGS)
complex h,u,v
logical first
data first/.true./
data k0/99999999/
save w,covxs,covxw,covys,covyw,s,ntc,ntot,nh,kstep,fac,first,k0
if(first) then
pi=4.0*atan(1.0)
do i=0,nfft-1
w(i)=(sin(i*pi/nfft))**2
enddo
s=0.
ntc=0
ntot=0
nh=nfft/2
kstep=nfft
if(nwindow.eq.2) kstep=nh
fac=1.0/nfft
slimit=1.e30
first=.false.
endif
if(k.lt.k0) then
covxs=0.
covxw=0.
covys=0.
covyw=0.
endif
k0=k
cx(0:nfft-1)=cx0
if(nwindow.eq.2) cx(0:nfft-1)=w(0:nfft-1)*cx(0:nfft-1)
call four2a(cx,nfft,1,1,1) !First forward FFT
if(nxpol.ne.0) then
cy(0:nfft-1)=cy0
if(nwindow.eq.2) cy(0:nfft-1)=w(0:nfft-1)*cy(0:nfft-1)
call four2a(cy,nfft,1,1,1) !First forward FFT
endif
if(iqapply.ne.0) then !Apply I/Q corrections
h=gainx*cmplx(cos(phasex),sin(phasex))
v=0.
do i=0,nfft-1
u=cx(i)
if(i.gt.0) v=cx(nfft-i)
x=real(u) + real(v) - (aimag(u) + aimag(v))*aimag(h) + &
(real(u) - real(v))*real(h)
y=aimag(u) - aimag(v) + (aimag(u) + aimag(v))*real(h) + &
(real(u) - real(v))*aimag(h)
cxt(i)=0.5*cmplx(x,y)
enddo
else
cxt(0:nfft-1)=cx(0:nfft-1)
endif
if(nxpol.ne.0) then
if(iqapply.ne.0) then !Apply I/Q corrections
h=gainy*cmplx(cos(phasey),sin(phasey))
v=0.
do i=0,nfft-1
u=cy(i)
if(i.gt.0) v=cy(nfft-i)
x=real(u) + real(v) - (aimag(u) + aimag(v))*aimag(h) + &
(real(u) - real(v))*real(h)
y=aimag(u) - aimag(v) + (aimag(u) + aimag(v))*real(h) + &
(real(u) - real(v))*aimag(h)
cyt(i)=0.5*cmplx(x,y)
enddo
else
cyt(0:nfft-1)=cy(0:nfft-1)
endif
endif
! Identify frequencies with strong signals, copy frequency-domain
! data into array cs (strong) or cw (weak).
ntot=ntot+1
if(mod(ntot,128).eq.5) then
call pctile(s,stmp,1024,50,xmedian)
slimit=faclim*xmedian
endif
if(ntc.lt.96000/nfft) ntc=ntc+1
uu=1.0/ntc
smax=0.
do i=0,nfft-1
p=real(cxt(i))**2 + aimag(cxt(i))**2
if(nxpol.ne.0) p=p + real(cyt(i))**2 + aimag(cyt(i))**2
s(i)=(1.0-uu)*s(i) + uu*p
lstrong(i)=(s(i).gt.slimit)
if(s(i).gt.smax) smax=s(i)
enddo
nsigs=0
lprev=.false.
iwid=1
ib=-99
do i=0,nfft-1
if(lstrong(i) .and. (.not.lprev)) then
if(nsigs.lt.MAXSIGS) nsigs=nsigs+1
ia(nsigs)=i-iwid
if(ia(nsigs).lt.0) ia(nsigs)=0
endif
if(.not.lstrong(i) .and. lprev) then
ib(nsigs)=i-1+iwid
if(ib(nsigs).gt.nfft-1) ib(nsigs)=nfft-1
endif
lprev=lstrong(i)
enddo
if(nsigs.gt.0) then
do i=1,nsigs
ja=ia(i)
jb=ib(i)
if(ja.lt.0 .or. ja.gt.nfft-1 .or. jb.lt.0 .or. jb.gt.nfft-1) then
cycle
endif
if(jb.eq.-99) jb=ja + min(2*iwid,nfft-1)
lstrong(ja:jb)=.true.
enddo
endif
do i=0,nfft-1
if(lstrong(i)) then
cxs(i)=fac*cxt(i)
cxw(i)=0.
if(nxpol.ne.0) then
cys(i)=fac*cyt(i)
cyw(i)=0.
endif
else
cxw(i)=fac*cxt(i)
cxs(i)=0.
if(nxpol.ne.0) then
cyw(i)=fac*cyt(i)
cys(i)=0.
endif
endif
enddo
call four2a(cxw,nfft,1,-1,1) !Transform weak and strong X
call four2a(cxs,nfft,1,-1,1) !back to time domain, separately
if(nxpol.ne.0) then
call four2a(cyw,nfft,1,-1,1) !Transform weak and strong Y
call four2a(cys,nfft,1,-1,1) !back to time domain, separately
endif
if(nwindow.eq.2) then
cxw(0:nh-1)=cxw(0:nh-1)+covxw(0:nh-1) !Add previous segment's 2nd half
covxw(0:nh-1)=cxw(nh:nfft-1) !Save 2nd half
cxs(0:nh-1)=cxs(0:nh-1)+covxs(0:nh-1) !Ditto for strong signals
covxs(0:nh-1)=cxs(nh:nfft-1)
if(nxpol.ne.0) then
cyw(0:nh-1)=cyw(0:nh-1)+covyw(0:nh-1) !Add previous segment's 2nd half
covyw(0:nh-1)=cyw(nh:nfft-1) !Save 2nd half
cys(0:nh-1)=cys(0:nh-1)+covys(0:nh-1) !Ditto for strong signals
covys(0:nh-1)=cys(nh:nfft-1)
endif
endif
! Apply noise blanking to weak data
if(nb.ne.0) then
do i=0,kstep-1
peak=abs(cxw(i))
if(nxpol.ne.0) peak=max(peak,abs(cyw(i)))
if(peak.gt.peaklimit) then
cxw(i)=0.
if(nxpol.ne.0) cyw(i)=0.
nzap=nzap+1
endif
enddo
endif
! Compute power levels from weak data only
do i=0,kstep-1
px=px + real(cxw(i))**2 + aimag(cxw(i))**2
if(nxpol.ne.0) py=py + real(cyw(i))**2 + aimag(cyw(i))**2
enddo
cx1(0:kstep-1)=cxw(0:kstep-1) + cxs(0:kstep-1) !Recombine weak + strong
if(nxpol.ne.0) then
cy1(0:kstep-1)=cyw(0:kstep-1) + cys(0:kstep-1) !Weak + strong
endif
return
end subroutine timf2