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 444 | Next

Joseph Boccuzzi

"Signal Processing for Wireless Communications"

132)
where NI is the dimension of the interference vector space of ?†R
IN.
From this spectral decomposition many insightful array weights can be derived and used under
different conditions. Below we will list three of them:
1. MMSE signal subspace. Here we use only the eigen vectors that correspond to the signal subspace,
which is of size Ns. The reasoning behind this weight is as follows. The cross-correlation vector is
a member of the signal subspace and as such is collectively orthogonal to the interference  noise
subspace. Hence the second summation would result in contributing ideally a value of zero. In
practice, it is not zero, but rather a small value and so by forcing the property performance
improvements can be observed.
(5.133)
2. MSINR noise subspace. This technique is also referred to as the Eigen Cancellor [71].
The reasoning behind this array weight is as follows. Since we are dealing with the RIN covariance
matrix the noise subspace is collectively orthogonal to the interference subspace. Hence in
using this noise vector space we can attempt to cancel/suppress the interference that belongs to the
interference subspace. That is why the second half of the summation is used.
(5.134)
3.


Pages:
432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456
hotel jelenia góra Russian bride Free English grammar and study guid powiekszenia wielkoformatowe counter strike 1.6