60)
With the above defined block matrices we can see that the MMSE array weight vector can be written as
(6.61)
(6.62)
where H1A is the channel matrix for the first arriving ray, H1B is the channel matrix for the second
arriving ray, H1C and H1D are the channel matrices involving both the first and second arriving rays.
In Fig. 6.13 we plot the BER performance of a K 3 tap fractionally spaced STE using M 2
antennas. The results are for the flat-fading and frequency-selective two-ray channel models. For the
wMMSE [H1A H1B H1C H1D s2n
I ]1 # rd
wMMSE [H1 H2 s2n
I ]1 # rd
F Rd (T )hx2(k T )hy1(k)
RdQ
3T
2 Rhx2Qk
3T
2 Rhy1(k)
Rd (2T )hx2(k 2T )hy1(k)
RdQ
T
2 Rhx2(k T )hy1Qk
T
2 R
Rd (T )hx2Qk
3T
2 Rhy1Qk
T
2 R
RdQ
3T
2 Rhx2(k 2T )hy1Qk
T
2 R
Rd (0)hx2(k T )hy1(k T )
RdQ
T
2 Rhx2Qk
3T
2 Rhy1(k T )
Rd (T )hx2(k 2T )hy1(k T ) V
F Rd (T )hx1(k)hy2(k T )
RdQT
2 Rhx1Qk
T
2 Rhy2(k T )
Rd(0)hx1(k T )hy2(k T )
RdQ3T
2 Rhx1(k)hy2Qk
3T
2 R
Rd (T )hx1Qk
T
2 Rhy2Qk
3T
2 R
RdQT
2 Rhx1(k T )hy2Qk
3T
2 R
Rd(2T )hx1(k)hy2(k 2T )
RdQ3T
2 Rhx1Qk
T
2 Rhy2(k 2T )
Rd (T )hx1(k T )hy2(k 2T ) V
RECEIVER DIGITAL SIGNAL PROCESSING 309
Space-Time Decomposition, M = 2 Antennas, fd = 190 Hz
K = 3 Taps (T/2-Spaced), MMSE Combine
1.
Pages:
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514