Enhancing MIMO transmission
View Patent ↗A multi-rank beamforming (MRBF) scheme in which the downlink channel is estimated and an optimal precoding matrix to be used by the MRBF transmitter is determined accordingly. The optimal precoding matrix is selected from a codebook of matrices having a recursive structure which allows for efficient computation of the optimal precoding matrix and corresponding Signal to Interference and Noise Ratio (SINR). The codebook also enjoys a small storage footprint. Due to the computational efficiency and modest memory requirements, the optimal precoding determination can be made at user equipment (UE) and communicated to a transmitting base station over a limited uplink channel for implementation over the downlink channel.
1. A method of operating a multi-rank beam forming (MRBF) system comprising:
modulating a data signal to generate a modulated symbol;
precoding the modulated symbol using a precoding matrix selected from a plurality of matrices; and
transmitting the precoded modulated symbol using at least two transmit antennas,
wherein the plurality of matrices have a nested structure and are derived from a set of vector codebooks determined in accordance with the following expression:
A
(
v
i
1
1
,
v
i
2
2
,
v
i
3
3
,
…
)
=
[
v
i
1
1
,
HH
(
v
i
1
1
-
e
1
M
)
[
0
v
i
2
2
]
,
HH
(
v
i
1
1
-
e
1
M
)
[
0
HH
(
v
i
2
2
-
e
1
M
-
1
)
[
0
v
i
3
3
]
]
,
…
]
,
wherein e 1 N =[1, 0, . . . , 0] T εC N , M is the number of transmit antennas, C N is an N-dimensional complex space,
HH
(
w
)
=
{
I
-
2
ww
H
w
2
if
w
≠
0
I
if
w
=
0
and
{
v
i
j
j
∈
C
M
-
j
+
1
}
is a vector codebook and v i a is selected from a set V a where V 1 ={v i 1 εC M } i=1 N 1 , V 2 ={e 1 M-1 }, V 3 ={e 1 M-2 }, . . . .
2. The method of claim 1 , wherein deriving the precoding matrix of rank-k comprises;
generating a square matrix using the expression in claim 1 ; and
picking k columns of the square matrix.
3. The method of claim 1 , further comprising:
estimating a channel over which the precoded modulated symbol is to be transmitted;
determining a channel quality metric for each of the plurality of matrices; and
selecting one of the matrices according to the channel quality metric.
4. The method of claim 3 wherein determining the channel quality metric for each of the plurality of matrices exploits the nested structure of said matrices.
5. The method of claim 4 , wherein exploiting the nested structure of a first matrix involves reusing a computation made in determining the channel quality metric of a second matrix which is a sub-matrix of the first matrix.
6. The method of claim 1 , wherein a scaled version of the precoding matrix is used.
7. The method of claim 1 , wherein the number of transmit antennas is 4.
8. The method of claim 1 , wherein the number of transmit antennas is 2.
9. The method of claim 2 , wherein the number of selected columns to form the precoding matrix is less than the number of transmit antennas.
10. The method of claim 2 , wherein all the columns are selected to form the precoding matrix.