DCH Despreader
X X
+
w1(t) = h1(t)
w2(t) = h2(t)
wN(t) = hN(t)
DCH Despreader
X X
S(t ??“ t2)
S(t ??“ tN)
S(t ??“ t1)
DCH Despreader
X X
MRC Combiner
.
.
.
.
.
.
d(t ??“ t1) = x1(t)
d(t ??“ t2) = x2(t)
d(t ??“ tN) = xN(t)
RAKE
Input
RAKE
Output
y(t)
t1+PG
t1
dt ??«
t2+PG
t1
dt ??«
t2+PG
t1
dt ??«
*
*
*
The question we wish to answer in this section is ???Can alternative combining techniques be used
in the RAKE???? The answer is an overwhelming YES. If we treat the combining function as a multisensor
problem, then whatever was discussed in the receiver spatial diversity section in Chap. 5 can
be applied in the RAKE receiver (i.e., SS, EGC, MRC, etc.). We wish to focus on what is called the
OC technique [8, 29].
The MRC combiner performs the following operations:
(7.95)
In this case, the weights are the channel estimates themselves [ ]. We could apply the
MMSE-based cost function and arrive with the following combining weight equation:
(7.96)
Where the estimated covariance matrix of the received and despread data symbols and
estimated cross-correlation vector between the received, despread data symbols, and the desired
signals.
Recall MRC is a special case of OC in that MRC assumes the covariance matrix to be a diagonal.
Pages:
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664