In this case, a simple block diagram is shown in Fig. 7.21.
w (I UI # U*I
) # rxd
RIN UI # DI # U*I
UN # DN # U*
N
R RD RI RN
358 CHAPTER SEVEN
FIGURE 7.20 GRAKE architecture.
RAKE
Output, I(t)
X X X X
. . .
. . .
. . .
. . .
+
Received
Signal
PN
Code
PG
. Tc
dt
0
??«
PG
. Tc
dt
0
??«
PG
. Tc
dt
0
??«
PG
. Tc
dt
0
??«
??† ??†
??†
Weight Calculation, w
X1 X2 X3 XN
If we have to combine at the chip level, then knowledge of the arrival and interarrival times of the
multipaths is necessary. This information is obtained from the multipath-searcher function and is used
in the time-deskewing operation. In addition to knowing the arrival time information, the channel
response is required before coherent combining can take place. As a simple example, consider a
2-ray channel model, where the separation between the 2 rays is 4 chips. A very simple implementation
would be to delay the first arriving multipath by 4-chip times and combine this sequence with a
nondelayed sequence (which represents the second ray). In this architecture, one needs to know the
maximum expected delay spread to be used in the matched filter. The benefit in using this method is
that a single despreader function is required to extract the desired signal.
Pages:
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593