Multi-resolution beamforming based on codebooks in MIMO systems
Certain aspects of the present disclosure relate to methods for beamforming that achieve beamforming optimality criterions. Some proposed beamforming techniques are based on antenna directions with multiple resolutions.
1. A method for wireless communications, comprising:
receiving beamforming capability information from a device either directly or via a piconet controller (PNC);
receiving training signals using a first subset of a first set of codebooks;
employing a second subset of the first set of codebooks to acquire channel state information (CSI), wherein the second subset is same or different than the first subset;
estimating a first preferred vector of weights from the first subset and a second preferred vector of weights from the second subset;
providing, as a feedback to the device, the first preferred vector of weights; and
using the second preferred vector of weights to communicate with the device on a receive direction from a set of receive directions, wherein
the beamforming capability information comprises an indication if codebooks are used for beamforming, a total number of N transmit antennas, and a total number of M transmit directions from a first set of transmit directions,
the first set of codebooks is employed for transmitting the training signals using transmit directions from the first set of transmit directions, and
the first set of codebooks is represented in a matrix form for the total of N transmit antennas and the total of M transmit directions from the first set of transmit directions as:
W
(
n
,
m
)
=
j
fix
[
n
×
mod
(
m
+
(
M
/
2
)
,
M
)
(
M
/
4
)
]
for
n
=
0
:
N
-
1
and
m
=
0
:
M
-
1
wherein fix(•) is a function that returns the integer part of its argument.
2. The method of claim 1 , wherein the first preferred vector of weights from the first subset of the first set of codebooks and the second preferred vector of weights from the second subset of the first set of codebooks are determined based on a signal-quality metric.
3. The method of claim 2 , wherein the signal-quality metric comprises an effective signal-to-noise ratio (ESNR) at the device.
4. A method for wireless communications, comprising:
receiving beamforming capability information from a device either directly or via a piconet controller (PNC), wherein the beamforming capability information comprises an indication if codebooks are used for beamforming, a total number of N transmit antennas, and a total number of M transmit directions from a first set of transmit directions
receiving training signals using a first subset of a first set of codebooks;
employing a second subset of the first set of codebooks to acquire channel state information (CSI), wherein the second subset is same or different than the first subset;
estimating a first preferred vector of weights from the first subset and a second preferred vector of weights from the second subset;
providing, as a feedback to the device, the first preferred vector of weights; and
using the second preferred vector of weights to communicate with the device on a receive direction from a set of receive directions,
wherein a second set of codebooks is employed for transmitting the training signals using transmit directions from a second set of transmit directions, wherein transmit directions from the first set of transmit directions have higher resolution than transmit directions from the second set of transmit directions, and
wherein the second set of codebooks is represented in a matrix form for the total number of N transmit antennas, N is an even number and P=N/2 transmit directions from the second set of transmit directions as:
W
(
n
,
m
)
=
{
(
-
j
)
mod
(
n
,
2
)
m
=
0
(
-
1
)
fix
[
n
×
mod
(
m
+
(
P
/
2
)
,
P
)
(
P
/
2
)
]
n
=
0
:
N
-
1
m
=
1
:
P
-
1
wherein each transmit direction from the second set of transmit directions comprises two transmit directions from the first set of transmit directions of a codebook from a first set of codebooks having N transmit directions from the first set of transmit directions, and wherein fix(•) is a function that returns the integer part of its argument.
5. The method of claim 1 , wherein at least a part of each training signal is based on Golay sequences.
6. An apparatus for wireless communications, comprising:
a receiver for receiving training signals using a first subset of a first set of codebooks and receiving beamforming capability information from a device either directly or via a piconet controller (PNC);
a circuit for employing a second subset of the first set of codebooks to acquire channel state information (CSI), wherein the second subset is same or different than the first subset;
an estimator for estimating a first preferred vector of weights from the first subset and a second preferred vector of weights from the second subset;
a circuit for providing, as a feedback to the device, the first preferred vector of weights; and
a circuit for using the second preferred vector of weights to communicate with the device on a receive direction from a set of receive directions, wherein
the beamforming capability information comprises an indication if codebooks are used for beamforming, a total number of N transmit antennas, and a total number of M transmit directions from a first set of transmit directions,
the first set of codebooks is employed for transmitting the training signals using transmit directions from the first set of transmit directions, and
the first set of codebooks is represented in a matrix form for the total of N transmit antennas and the total of M transmit directions from the first set of transmit directions as:
W
(
n
,
m
)
=
j
fix
[
n
×
mod
(
m
+
(
M
/
2
)
,
M
)
(
M
/
4
)
]
for
n
=
0
:
N
-
1
and
m
=
0
:
M
-
1
wherein fix(•) is a function that returns the integer part of its argument.
7. The apparatus of claim 6 , wherein the first preferred vector of weights from the first subset of the first set of codebooks and the second preferred vector of weights from the second subset of the first set of codebooks are determined based on a signal-quality metric.
8. The apparatus of claim 7 , wherein the signal-quality metric comprises an effective signal-to-noise ratio (ESNR) at the device.
9. An apparatus for wireless communications, comprising:
a receiver for receiving training signals using a first subset of a first set of codebooks and receiving beamforming capability information from a device either directly or via a piconet controller (PNC), wherein the beamforming capability information comprises an indication if codebooks are used for beamforming, a total number of N transmit antennas, and a total number of M transmit directions from a first set of transmit directions;
a circuit for employing a second subset of the first set of codebooks to acquire channel state information (CSI), wherein the second subset is same or different than the first subset;
an estimator for estimating a first preferred vector of weights from the first subset and a second preferred vector of weights from the second subset;
a circuit for providing, as a feedback to the device, the first preferred vector of weights; and
a circuit for using the second preferred vector of weights to communicate with the device on a receive direction from a set of receive directions,
wherein a second set of codebooks is employed for transmitting the training signals using transmit directions from a second set of transmit directions, wherein transmit directions from the first set of transmit directions have higher resolution than transmit directions from the second set of transmit directions, and
wherein the second set of codebooks is represented in a matrix form for the total number of N transmit antennas, N is an even number and P=N/2 transmit directions from the second set of transmit directions as:
W
(
n
,
m
)
=
{
(
-
j
)
mod
(
n
,
2
)
m
=
0
(
-
1
)
fix
[
n
×
mod
(
m
+
(
P
/
2
)
,
P
)
(
P
/
2
)
]
n
=
0
:
N
-
1
m
=
1
:
P
-
1
wherein each transmit direction from the second set of transmit directions comprises two transmit directions from the first set of transmit directions of a codebook from a first set of codebooks having N transmit directions from the first set of transmit directions, and wherein fix(•) is a function that returns the integer part of its argument.
10. The apparatus of claim 6 , wherein at least a part of each training signal is based on Golay sequences.
11. An apparatus for wireless communications, comprising:
means for receiving training signals using a first subset of a first set of codebooks and receiving beamforming capability information from a device either directly or via a piconet controller (PNC);
means for employing a second subset of the first set of codebooks to acquire channel state information (CSI), wherein the second subset is same or different than the first subset;
means for estimating a first preferred vector of weights from the first subset and a second preferred vector of weights from the second subset;
means for providing, as a feedback to the device, the first preferred vector of weights; and
means for using the second preferred vector of weights to communicate with the device on a receive direction from a set of receive directions, wherein
the beamforming capability information comprises an indication if codebooks are used for beamforming, a total number of N transmit antennas, and a total number of M transmit directions from a first set of transmit directions,
the first set of codebooks is employed for transmitting the training signals using transmit directions from the first set of transmit directions, and
the first set of codebooks is represented in a matrix form for the total of N transmit antennas and the total of M transmit directions from the first set of transmit directions as:
W
(
n
,
m
)
=
j
fix
[
n
×
mod
(
m
+
(
M
/
2
)
,
M
)
(
M
/
4
)
]
for
n
=
0
:
N
-
1
and
m
=
0
:
M
-
1
wherein fix(•) is a function that returns the integer part of its argument.
12. The apparatus of claim 11 , wherein the first preferred vector of weights from the first subset of the first set of codebooks and the second preferred vector of weights from the second subset of the first set of codebooks are determined based on a signal-quality metric.
13. The apparatus of claim 12 , wherein the signal-quality metric comprises an effective signal-to-noise ratio (ESNR) at the device.
14. An apparatus for wireless communications, comprising:
means for receiving training signals using a first subset of a first set of codebooks and receiving beamforming capability information from a device either directly or via a piconet controller (PNC), wherein the beamforming capability information comprises an indication if codebooks are used for beamforming, a total number of N transmit antennas, and a total number of M transmit directions from a first set of transmit directions;
means for employing a second subset of the first set of codebooks to acquire channel state information (CSI), wherein the second subset is same or different than the first subset;
means for estimating a first preferred vector of weights from the first subset and a second preferred vector of weights from the second subset;
means for providing, as a feedback to a device, the first preferred vector of weights; and
means for using the second preferred vector of weights to communicate with the device on a receive direction from a set of receive directions,
wherein a second set of codebooks is employed for transmitting the training signals using transmit directions from a second set of transmit directions, wherein transmit directions from the first set of transmit directions have higher resolution than transmit directions from the second set of transmit directions, and
wherein the second set of codebooks is represented in a matrix form for the total number of N transmit antennas, N is an even number and P=N/2 transmit directions from the second set of transmit directions as:
W
(
n
,
m
)
=
{
(
-
j
)
mod
(
n
,
2
)
m
=
0
(
-
1
)
fix
[
n
×
mod
(
m
+
(
P
/
2
)
,
P
)
(
P
/
2
)
]
n
=
0
:
N
-
1
m
=
1
:
P
-
1
wherein each transmit direction from the second set of transmit directions comprises two transmit directions from the first set of transmit directions of a codebook from a first set of codebooks having N transmit directions from the first set of transmit directions, and wherein fix(•) is a function that returns the integer part of its argument.
15. The apparatus of claim 11 , wherein at least a part of each training signal is based on Golay sequences.
16. A computer-program product for wireless communications, comprising a computer readable storage medium encoded with instructions executable to:
receive training signals using a first subset of a first set of codebooks;
receive beamforming capability information from a device either directly or via a piconet controller (PNC);
employ a second subset of the first set of codebooks to acquire channel state information (CSI), wherein the second subset is same or different than the first subset;
estimate a first preferred vector of weights from the first subset and a second preferred vector of weights from the second subset;
provide, as a feedback to the device, the first preferred vector of weights; and
use the second preferred vector of weights to communicate with the device on a receive direction from a set of receive directions, wherein
the beamforming capability information comprises an indication if codebooks are used for beamforming, a total number of N transmit antennas, and a total number of M transmit directions from a first set of transmit directions,
the first set of codebooks is employed for transmitting the training signals using transmit directions from the first set of transmit directions, and
the first set of codebooks is represented in a matrix form for the total of N transmit antennas and the total of M transmit directions from the first set of transmit directions as:
W
(
n
,
m
)
=
j
fix
[
n
×
mod
(
m
+
(
M
/
2
)
,
M
)
(
M
/
4
)
]
for
n
=
0
:
N
-
1
and
m
=
0
:
M
-
1
wherein fix(•) is a function that returns the integer part of its argument.
17. An access point, comprising:
at least one antenna;
a receiver for receiving via the at least one antenna training signals using a first subset of a first set of codebooks and receiving beamforming capability information from a device either directly or via a piconet controller (PNC);
a circuit for employing a second subset of the first set of codebooks to acquire channel state information (CSI), wherein the second subset is same or different than the first subset;
an estimator for estimating a first preferred vector of weights from the first subset and a second preferred vector of weights from the second subset;
a circuit for providing, as a feedback to the device, the first preferred vector of weights; and
a circuit for using the second preferred vector of weights to communicate with the device on a receive direction from a set of receive directions, wherein
the beamforming capability information comprises an indication if codebooks are used for beamforming, a total number of N transmit antennas, and a total number of M transmit directions from a first set of transmit directions,
the first set of codebooks is employed for transmitting the training signals using transmit directions from the first set of transmit directions, and
the first set of codebooks is represented in a matrix form for the total of N transmit antennas and the total of M transmit directions from the first set of transmit directions as:
W
(
n
,
m
)
=
j
fix
[
n
×
mod
(
m
+
(
M
/
2
)
,
M
)
(
M
/
4
)
]
for
n
=
0
:
N
-
1
and
m
=
0
:
M
-
1
wherein fix(•) is a function that returns the integer part of its argument.
18. A method for wireless communications, comprising:
transmitting beamforming capability information to a device either directly or via a piconet controller (PNC);
transmitting training signals using a first subset of a first set of codebooks;
receiving, as a feedback from the device, a first preferred vector of weights; and
using the first preferred vector of weights to communicate with the device on a transmit direction from a first set of transmit directions, wherein
the beamforming capability information comprises an indication if codebooks are used for beamforming, a total number of N transmit antennas, and a total number of M transmit directions from the first set of transmit directions,
the first set of codebooks is employed for transmitting the training signals using transmit directions from the first set of transmit directions, and
the first set of codebooks is represented in a matrix form for a total of N transmit antennas and a total of M transmit directions from the first set of transmit directions as:
W
(
n
,
m
)
=
j
fix
[
n
×
mod
(
m
+
(
M
/
2
)
,
M
)
(
M
/
4
)
]
for
n
=
0
:
N
-
1
and
m
=
0
:
M
-
1
wherein fix(•) is a function that returns the integer part of its argument.
19. An apparatus for wireless communications, comprising:
a transmitter for transmitting training signals using a first subset of a first set of codebooks and transmitting beamforming capability information to a device either directly or via a piconet controller (PNC);
a receiver for receiving, as a feedback from the device, a first preferred vector of weights; and
a circuit for using the first preferred vector of weights to communicate with the device on a transmit direction from a first set of transmit directions, wherein
the beamforming capability information comprises an indication if codebooks are used for beamforming, a total number of N transmit antennas, and a total number of M transmit directions from the first set of transmit directions,
the first set of codebooks is employed for transmitting the training signals using transmit directions from the first set of transmit directions, and
the first set of codebooks is represented in a matrix form for a total of N transmit antennas and a total of M transmit directions from the first set of transmit directions as:
W
(
n
,
m
)
=
j
fix
[
n
×
mod
(
m
+
(
M
/
2
)
,
M
(
M
/
4
)
]
for
n
=
0
:
N
-
1
and
m
=
0
:
M
-
1
wherein fix(•) is a function that returns the integer part of its argument.
20. An apparatus for wireless communications, comprising:
means for transmitting training signals using a first subset of a first set of codebooks and transmitting beamforming capability information to a device either directly or via a piconet controller (PNC);
means for receiving, as a feedback from the device, a first preferred vector of weights; and
means for using the first preferred vector of weights to communicate with the device on a transmit direction from a first set of transmit directions, wherein
the beamforming capability information comprises an indication if codebooks are used for beamforming, a total number of N transmit antennas, and a total number of M transmit directions from the first set of transmit directions,
the first set of codebooks is employed for transmitting the training signals using transmit directions from the first set of transmit directions, and
the first set of codebooks is represented in a matrix form for a total of N transmit antennas and a total of M transmit directions from the first set of transmit directions as:
W
(
n
,
m
)
=
j
fix
[
n
×
mod
(
m
+
(
M
/
2
)
,
M
(
M
/
4
)
]
for
n
=
0
:
N
-
1
and
m
=
0
:
M
-
1
wherein fix(•) is a function that returns the integer part of its argument.
21. A computer-program product for wireless communications, comprising a computer readable storage medium encoded with instructions executable to:
transmit beamforming capability information to a device either directly or via a piconet controller (PNC);
transmit training signals using a first subset of a first set of codebooks;
receive, as a feedback from the device, a first preferred vector of weights; and
use the first preferred vector of weights to communicate with the device on a transmit direction from a first set of transmit directions, wherein
the beamforming capability information comprises an indication if codebooks are used for beamforming, a total number of N transmit antennas, and a total number of M transmit directions from the first set of transmit directions,
the first set of codebooks is employed for transmitting the training signals using transmit directions from the first set of transmit directions, and
the first set of codebooks is represented in a matrix form for a total of N transmit antennas and a total of M transmit directions from the first set of transmit directions as:
W
(
n
,
m
)
=
j
fix
[
n
×
mod
(
m
+
(
M
/
2
)
,
M
(
M
/
4
)
]
for
n
=
0
:
N
-
1
and
m
=
0
:
M
-
1
wherein fix(•) is a function that returns the integer part of its argument.
22. An access point, comprising:
at least one antenna;
a transmitter for transmitting via the at least one antenna training signals using a first subset of a first set of codebooks and transmitting beamforming capability information to a device either directly or via a piconet controller (PNC);
a receiver for receiving via the at least one antenna, as a feedback from the device, a first preferred vector of weights; and
a circuit for using the first preferred vector of weights to communicate with the device on a transmit direction from a first set of transmit directions, wherein
the beamforming capability information comprises an indication if codebooks are used for beamforming, a total number of N transmit antennas, and a total number of M transmit directions from the first set of transmit directions,
the first set of codebooks is employed for transmitting the training signals using transmit directions from the first set of transmit directions, and
the first set of codebooks is represented in a matrix form for a total of N transmit antennas and a total of M transmit directions from the first set of transmit directions as:
W
(
n
,
m
)
=
j
fix
[
n
×
mod
(
m
+
(
M
/
2
)
,
M
(
M
/
4
)
]
for
n
=
0
:
N
-
1
and
m
=
0
:
M
-
1
wherein fix(•) is a function that returns the integer part of its argument.