113)
where d(t) is the information symbol, c(t) is the concatenated scrambling plus spreading code, h(t) is
the channel response, and n(t) is the AWGN. The late correlator output signal is given as
(7.114)
(7.115)
(7.116)
Since it is commonly assumed that the pilot channel is used to perform fine time tracking, then
d(k) is either known to be depiloted or constant. In either case, it can be removed from Eq. (7.116),
so we now have
(7.117)
Hence the late correlator output consists of a shifted autocorrelation function weighed by the channel
response. Similarly, we can write down the early correlation output signal as
(7.118)
(7.119) rE(k) h(k t) # Rcca
Tc
2 b noise terms
rE(k) ak
r(k) # cak t
Tc
2 b noise terms
rL(k) h(k t) # RccaTc
2 b noise terms
rL(k) d(k t) # h(k t) # RccaTc
2 b noise terms
rL(k) d(k t) # h(k t) # ak
c(k t) # cak t
Tc
2 b noise terms
rL(k) rL(t)Ztk
Tc
4
ak
r(k) # cak t
Tc
2 b noise terms
r(t) d(t t) # c(t t) # h(t t) n(t)
410 CHAPTER SEVEN
FIGURE 7.83 Noncoherent DLL RAKE finger-block diagram.
X
O
Finger
Input
| |2
| |2 X
PN
Generator
+
LPF
??“
E L
Error
Signal
Finger
Output
X X
X
CCCH
CDCH
Buffering
MA
X
*
e(t)
C(t ??“ t ??“ Tc/2)
C(t ??“ t + Tc/2)
rL(t)
rE(t)
C(t??“t)
??‘
??‘
??‘
??‘
As shown above, we wish to subtract these two correlator outputs in order to determine if the optimal
sampling point is the one presently being used or not.
Pages:
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677