6.19). This is expected since we assumed no interference was present
and the desired signal experienced flat fading.
L(k) Re5h ^
*(k) # s*h
(k) # r(k)6
M(k) aD
j1 Prj(k) aJ1
i0
h ^
j(i) # sh(k i) P2
M(k) aD
j1 Prj(k) h ^
j(k) # sh(k) P2
^h
(k)
RECEIVER DIGITAL SIGNAL PROCESSING 315
T
MLSE | |2
n(k)
h0,1(k)
X X
X
h0(k) h1(k)
+
s(k)
+ +
s1(k)
r(k)
??“
+
X X h1(k) ?† ?†h
0(k)
?†h
0(k)
T
Channel
Estimation Z??“q
sh(k)
h1(k) ?†
FIGURE 6.17 Overall system block diagram using an MLSE-based receiver.
Now let??™s return to the presence of interference which is not white, then the previously provided
ML delay-spread metric should be minimized. Alternatively, it is easy to show that the above metric
is equivalent to maximizing the following relationship.
(6.90)
Which can be alternatively shown next (see Fig. 6.20).
L(k) Re5h ^
*(k) # R1
In(k) # r(k) # s*h
(k)6
316 CHAPTER SIX
FIGURE 6.18 Simplified block diagram of an MLSE-based receiver.
r(k)
L(k)
X X Re{}
Choose
Largest
S*
h h* ?†
FIGURE 6.19 ML equivalent representation of maximal ratio combining.
h*
r(k)
L(k)
X X
R??“1
I+n(k)
Re{}
Choose
Largest
S*
h
?† .
FIGURE 6.20 ML equivalent representation of optimum combining using the
MSINR-based weights.
Pages:
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523