Digital Signal Processing Reference
In-Depth Information
H
i
)
=
H
i
)
= ··· =
H
i
(
1
(
2
(
N
)
(2.85)
G
i
)
=
G
i
)
= ··· =
G
i
(
1
(
2
(
N
)
(2.86)
respectively. In order to make the symbols of Users 1 and 2 transmitted in two
orthogonal subspaces, i.e., the first
N
columns of
H
are orthogonal to the second
N
columns of
H
,welet
G
i
=
η
i
−
H
i
∗
(2.87)
(
1
,
1
)
(
2
,
1
)
G
i
H
i
(
,
)
(
,
)
2
1
1
1
Step 3
: Designing low-complexity algorithms to calculate the parameters of the
precoders:
From (
2.87
), we have
⎛
⎞
⎛
⎞
∗
b
i
11
b
i
21
.
b
i
N
1
a
i
11
a
i
21
.
a
i
N
1
g
11
g
12
···
∗
=
η
i
−
⎝
⎠
⎝
⎠
g
1
N
h
21
−
h
22
··· −
h
2
N
(2.88)
g
21
g
22
···
g
2
N
h
11
h
12
···
h
1
N
with normalization equations
1
N
a
i
11
|
2
a
i
21
|
2
a
i
N
1
|
2
|
+|
+···+|
=
1
N
b
i
11
|
2
b
i
21
|
2
b
i
N
1
|
2
|
+|
+···+|
=
(2.89)
Note that the channel matrices in (
2.88
) are not square matrices. Therefore, we cannot
use the reverse matrix directly as we did for the users with 2 transmit antennas in
Sect.
2.1
. Instead, in order to simplify the precoder design, at the first 2 time slots,
we let all the elements in complex vector
=
a
i
11
a
i
21
···
a
i
N
1
T
a
i
,
=
,
i
1
2
(2.90)
be zero except for the first 2 elements and also let all the elements in
=
b
i
11
b
i
21
···
b
i
N
1
T
b
i
,
i
=
1
,
2
(2.91)
be zero except for the first 2 elements. By the above choices for
a
i
and
b
i
,Eq.(
2.88
)
results in
g
11
g
12
g
21
g
22
∗
b
i
11
b
i
21
∗
=
η
i
−
a
i
11
a
i
21
h
21
−
h
22
(2.92)
h
11
h
12
Search WWH ::
Custom Search