SEARCH
0-9 A B C D E F G H I J K L M N O P Q R S T U V W X Y Z
Prev | Current Page 499 | Next

Benny Bing

"Emerging Technologies in Wireless LANs: Theory, Design, and Deployment"

11n WLAN 212
9.7.4 LSD Max-log-MAP
On the average, only hypotheses for s are considered when computing the bit
LLRs, and only ??‘ operations are performed. The corresponding
upper bounds on these quantities are given by and , respectively. The
resulting average complexity, expressed as the equivalent number of real MUL and ADD
operations, is summarized in Table 9.3, where the indicator function 1full is 1 only in the
first iteration (otherwise 0), and 1it is 1 if a-priori LLRs are available, as they would
normally be in the second and subsequent iterations.
1 ( ) S N
ave i N i = ??‘
1 ( ) 2 S N
ave i N i = ??’ {} max ?‹…
S N
cand N 2 S N
cand N ??’
9.7.5 Per-Stream LLR Computation
When performing per-stream LLR computation as in the suboptimal MMSE-based
receiver, the number of hypotheses that need to be considered is reduced from
(for full-complexity max-log-MAP) to ??‘ . Further complexity reduction is made
possible by the fact that the cost functions take scalar inputs rather than vectors and
matrices. The complexity of per-stream max-log-MAP LLR computation is summarized in
Table 9.3.
1 2
NS
i i K = ??‘
1 2 S i N K
i=
9.


Pages:
487 488 489 490 491 492 493 494 495 496 497 498 499 500 501 502 503 504 505 506 507 508 509 510 511