System and method for transceivers in a wireless network
View Patent ↗In accordance with an embodiment, a method of operating a multi-output wireless transmitter includes determining a precoded data stream based on input data, where the determining includes determining a signal to interference plus noise ratio (SINR) of a plurality of receivers, determining a channel gain to each of the plurality of receivers, and maximizing a sum of utility functions, wherein each utility function is proportional to the SINR of each of the plurality of receivers at low SINR values and asymptotically approaches a constant value at high SINR values. The precoded data stream is transmitted on the multi-output wireless transmitter.
1. A method of operating a multi-output wireless transmitter, the method comprising:
determining a precoded data stream in accordance with input data, the determining comprising,
determining a signal to interference plus noise ratio (SNIR) of a plurality of receivers,
determining a channel gain to each of the plurality of receivers, and
maximizing a sum of utility functions without using a log detector, wherein each utility function is proportional to the SNIR of each of the plurality of receivers at low SNIR values and asymptotically approaches a constant value at high SNIR values; and
transmitting the precoded data stream on the multi-output wireless transmitter, wherein determining the precoded data stream further comprises:
updating auxiliary variable W k according to:
W
k
=
(
I
-
H
kk
Q
k
H
kk
T
(
σ
2
I
+
∑
l
=
1
K
H
kl
Q
l
H
kl
T
)
-
1
)
-
1
,
∀
k
=
1
,
2
,
…
,
K
;
and solving the following optimization problem
min
Q
k
Tr
[
B
k
Q
k
]
s
.
t
.
Tr
(
Q
k
)
=
p
k
Q
k
μ
0
,
where H kk is a channel matrix from transmitter k (after the precoded data) to receiver k, K is a number of transmitters, H kl is a channel matrix from transmitter l (after the precoded data) to receiver k, H jj is a channel matrix from transmitter j (after the precoded data) to receiver j, H jk is a channel matrix from transmitter k (after the precoded data) to receiver j, Q k is a precoded transmit data covariance matrix of user k, Q l is a precoded transmit data covariance matrix of user l, {tilde over (Q)} k is a previous value for Q k , Q j is a precoded transmit data covariance matrix of user j, and p k is a power constraint of precoded data of user k, B k , represents received interference power from all receivers, I is an identity matrix and σ is a noise standard deviation.
2. The method of claim 1 , wherein each utility function is expressed as:
γ
p
+
γ
,
where p comprises a constant and γ comprises a SNIR.
3. The method of claim 1 , wherein the multi-output transmitter comprises a multi-antenna beam forming transmitter and a multi-time slot transmitter, and
wherein the multi-output transmitter operates using multiple subcarriers.
4. The method of claim 1 , further comprising receiving the transmitted precoded data stream with a linear minimum mean square error receiver.
5. The method of claim 1 , wherein determining the precoded data stream is performed by a base station processor.
6. The method of claim 1 , wherein determining the SNIR of a plurality of receivers comprises polling receivers.
7. A method of operating a multi-output wireless transmitter, the method comprising:
determining a precoded data stream in accordance with input data, the determining comprising,
determining a signal to interference plus noise ratio (SNIR) of a plurality of receivers,
determining a channel gain to each of the plurality of receivers, and
maximizing a sum of utility functions without using a log detector, wherein each utility function is proportional to the SNIR of each of the plurality of receivers at low SNIR values and asymptotically approaches a constant value at high SNIR values; and
transmitting the precoded data stream on the multi-output wireless transmitter, wherein determining the precoded data stream further comprises solving the following optimization problem:
max
Q
k
Tr
[
H
kk
Q
k
H
kk
T
(
σ
2
I
+
∑
l
=
1
K
H
kl
Q
l
H
kl
T
)
-
1
]
s
.
t
Tr
(
Q
k
)
≤
p
k
Q
k
≽
0
,
where H kk is a channel matrix from transmitter k (after the precoded data) to receiver k, H kl is a channel from transmitter l (after the precoded data) to receiver k, Q k is a precoded transmit data covariance matrix of user k, Q l is the precoded transmit data covariance matrix of user l, and p k is a power constraint of precoded data of user k, I is an identity matrix, σ is a noise standard deviation, and the superscript T denotes a transpose operation in real signaling and hermitian transpose in complex signaling.
8. A method of operating a multi-output wireless transmitter, the method comprising:
determining a precoded data stream in accordance with input data, the determining comprising,
determining a signal to interference plus noise ratio (SNIR) of a plurality of receivers,
determining a channel gain to each of the plurality of receivers, and
maximizing a sum of utility functions without using a log detector, wherein each utility function is proportional to the SNIR of each of the plurality of receivers at low SNIR values and asymptotically approaches a constant value at high SNIR values; and
transmitting the precoded data stream on the multi-output wireless transmitter, wherein determining the precoded data stream comprises solving the following optimization problem:
max
Q
k
α
k
Tr
[
H
kk
Q
k
H
kk
T
(
σ
2
I
+
∑
l
=
1
K
H
kl
Q
l
H
kl
T
)
-
1
]
+
∑
j
≠
k
α
j
Tr
[
H
jj
Q
j
H
jj
T
(
σ
2
I
+
∑
l
=
1
K
H
jl
Q
l
H
jl
T
)
-
1
]
s
.
t
Tr
(
Q
k
)
≤
p
k
Q
k
≽
0
,
where H kk is a channel matrix from transmitter k (after the precoded data) to receiver k, H kl is a channel matrix from transmitter/(after the precoded data) to receiver k, H jj is a channel matrix from transmitter j (after the precoded data) to receiver j, H jl is a channel matrix from transmitter l (after the precoded data) to receiver j, Q k is a precoded transmit data covariance matrix of user k, Q l is a precoded transmit data covariance matrix of user l, Q j is a precoded transmit data covariance matrix of user j, p k is a power constraint of precoded data of user k, I is an identity matrix, α j is a priority weight of user j, α k is a priority weight of user k, and σ is a noise standard deviation.
9. The method of claim 8 , further comprising updating precoded transmit data covariance matrix Q k according to
Q k ←αQ k *+(1−α) {tilde over (Q)} k ,
where 0<α≦1 is a regularization factor, Q k * is a value found for the transmit covariance matrix after solving the optimization problem, and {tilde over (Q)} k is a previous value for the transmit covariance matrix Q k .
10. A method of operating a multi-output wireless transmitter, the method comprising:
determining a precoded data stream in accordance with input data, determining comprising,
determining a signal to interference plus noise ratio (SNIR) of a plurality of receivers,
determining a channel gain to each of the plurality of receivers, and
maximizing a sum of utility functions without using a log detector, wherein each utility function is proportional to the SNIR of each of the plurality of receivers at low SNIR values and asymptotically approaches a constant value at high SNIR values; and
transmitting the precoded data stream on the multi-output wireless transmitter, wherein determining the precoded data stream comprises solving the following optimization problem:
min
Q
k
,
Y
α
k
Tr
[
Y
]
+
Tr
[
B
k
Q
k
]
s
.
t
Tr
(
Q
k
)
≤
p
k
,
Q
k
μ
0
[
C
kk
+
H
kk
Q
k
H
kk
T
(
W
k
C
kk
)
1
2
(
W
k
C
kk
)
1
2
Y
]
μ
0
,
where
,
C
k
,
k
Δ
=
∑
l
≠
k
K
H
kl
Q
l
H
kl
T
and
B
k
Δ
=
∑
j
≠
k
a
j
H
jk
T
(
C
j
,
k
+
H
jk
Q
~
k
H
jk
T
)
-
1
W
j
H
jj
Q
j
H
jj
T
(
C
j
,
k
+
H
jk
Q
~
k
H
jk
T
)
-
1
H
jk
,
where W k is an auxilliary variable, H kk is a channel matrix from transmitter k (after the precoded data) to receiver k, H kl is a channel matrix from transmitter l (after the precoded data) to receiver k, H jj is a channel matrix from transmitter j (after the precoded data) to receiver j, H jk is a channel matrix from transmitter k (after the precoded data) to receiver j, Q k is a precoded transmit data covariance matrix of user k, Q l is a precoded transmit data covariance matrix of user l, {tilde over (Q)} k is a previous value for Q k , Q j is a precoded transmit data covariance matrix of user j, and p k is a power constraint of precoded data of user k, C k,k represents received power, C j,k represents the received signal covariance matrix of receiver j excluding the k-th user's signal, B k , represents received interference power from all receivers, α j is a priority weight of user j, α k is a priority weight of user k, and Y is an auxiliary variable.
11. The method according to claim 10 , further comprising updating the auxiliary variable W k according to:
W
k
=
(
I
-
H
kk
Q
k
H
kk
T
(
σ
2
I
+
∑
l
=
1
K
H
kl
Q
l
H
kl
T
)
-
1
)
-
1
,
∀
k
=
1
,
2
,
…
,
K
;
where K is a number of transmitters, σ is a noise standard deviation, and I is an identity matrix.
12. A method of operating a multi-output wireless transmitter, the method comprising:
determining a precoded data stream in accordance with input data; and
transmitting the precoded data stream on the multi-output wireless transmitter,
wherein determining the precoded data stream comprises:
determining a signal to interference plus noise ratio (SNIR) of a plurality of receivers;
determining a channel gain to each of the plurality of receivers;
maximizing a sum of utility functions, wherein each utility function is proportional to the SNIR of each of the plurality of receivers at low SNIR values and asymptotically approaches a constant value at high SNIR values; and
solving the following optimization problem
min
U
,
V
,
W
∑
k
α
k
(
Tr
[
W
k
E
k
]
-
log
det
W
k
)
,
where E k is a mean square error for used k according to
E
k
=
(
I
-
U
k
T
H
kk
V
k
)
(
I
-
U
k
T
H
kk
V
k
)
T
+
∑
j
≠
k
U
k
T
H
kj
V
j
V
j
T
H
kj
T
U
k
+
σ
k
2
U
k
T
U
k
,
where V k is a transmit beamformer, U k is a receive beamformer, H kk is a channel matrix from transmitter k (after the precoded data) to receiver k, H kl is a channel matrix from transmitter l (after the precoded data) to receiver k, σ is a noise standard deviation, and W k is an auxiliary variable.
13. The method according to claim 12 , further comprising updating auxiliary variable W k according to: W k =(I−U k T H kk V k ) −1 ,
where K is a number of transmitters, Q k is a precoded transmit data covariance matrix of user k, Q l is a precoded transmit data covariance matrix of user l, σ is a noise standard deviation, and I is an identity matrix.
14. The method of claim 12 , further comprising using the following constraint when solving the optimization problem:
Tr ( V k V k T )≦ P k ,
where P k is a maximum available power of transmitter k.
15. A method of operating a wireless base station, the method comprising:
determining a signal to interference plus noise ratio (SNIR) of a plurality of receivers;
determining a channel gain to each of the plurality of receivers;
determining a precoded data stream in accordance with input data, the determining comprising solving the following optimization problem:
max
Q
k
Tr
[
H
kk
Q
k
H
kk
T
(
σ
2
I
+
∑
l
=
1
K
H
kl
Q
l
H
kl
T
)
-
1
]
s
.
t
Tr
(
Q
k
)
≤
p
k
Q
k
≽
0
,
where H kk is a channel matrix from transmitter k (after the precoded data) to receiver k, H kl is a channel from transmitter l (after the precoded data) to receiver k, Q k is a precoded transmit data covariance matrix of user k, Q l is a precoded transmit data covariance matrix of user l, p k is a power constraint of precoded data of user k, I is an identity matrix and σ is a noise standard deviation; and
transmitting the precoded data stream on a multi-output wireless transmitter.
16. The method of claim 15 , wherein the multi-output wireless transmitter comprises a multi-antenna beam forming transmitter.
17. The method of claim 15 , wherein at least one of the plurality of receivers comprises a cellular telephone.
18. A multi-output transmission device comprising:
a base station processor configured to:
determine a precoded data stream in accordance with input data;
determine a signal to interference plus noise ratio (SNIR) of a plurality of receivers;
determine a channel gain to each of the plurality of receivers; and
maximize a sum of utility functions without using a log detector, wherein each utility function is proportional to the SNIR of each of the plurality of receivers at low SNIR values and asymptotically approaches a constant value at high SNIR values, and wherein the base station processor determines the precoded data stream by solving the following optimization problem:
max
Q
k
Tr
[
H
kk
Q
k
H
kk
T
(
σ
2
I
+
∑
l
=
1
K
H
kl
Q
l
H
kl
T
)
-
1
]
s
.
t
Tr
(
Q
k
)
≤
p
k
Q
k
≽
0
,
where H kk is a channel matrix from transmitter k (after the precoded data) to receiver k, H kl is a channel from transmitter l (after the precoded data) to receiver k, Q k is a precoded transmit data covariance matrix of user k, Q l is a precoded transmit data covariance matrix of user l, and p k is a power constraint of precoded data of user k, I is an identity matrix and σ is a noise standard deviation.
19. The transmission device of claim 18 , wherein each utility function is expressed as:
γ
p
+
γ
,
where p comprises a constant and γ comprises a SNIR.
20. The transmission device of claim 18 , wherein the multi-output transmission device comprises a multi-antenna beam forming transmitter and a multi-time slot transmitter, and
wherein the multi-output transmission device operates using multiple subcarriers.
21. The transmission device of claim 18 , further comprising a multi-output transmitter for transmitting the precoded data.
22. The transmission device of claim 18 , further comprising a receiver, the receiver configured to obtain SNIR values from the plurality of receivers.
23. A multi-output transmission device comprising:
a base station processor configured to:
determine a precoded data stream in accordance with input data;
determine a signal to interference plus noise ratio (SNIR) of a plurality of receivers;
determine a channel gain to each of the plurality of receivers; and
maximize a sum of utility functions without using a log detector, wherein each utility function is proportional to the SNIR of each of the plurality of receivers at low SNIR values and asymptotically approaches a constant value at high SNIR values, and wherein the base station processor determines the precoded data stream by solving the following optimization problem:
max
Q
k
α
k
Tr
[
H
kk
Q
k
H
kk
T
(
σ
2
I
+
∑
l
=
1
K
H
kl
Q
l
H
kl
T
)
-
1
]
+
∑
j
≠
k
α
j
Tr
[
H
jj
Q
j
H
jj
T
(
σ
2
I
+
∑
l
=
1
K
H
jl
Q
l
H
jl
T
)
-
1
]
s
.
t
Tr
(
Q
k
)
≤
p
k
Q
k
≽
0
,
where H kk is a channel matrix from transmitter k (after the precoded data) to receiver k, H kl is a channel matrix from transmitter l (after the precoded data) to receiver k, H jj is a channel matrix from transmitter j (after the precoded data) to receiver j, H jl is a channel matrix from transmitter l (after the precoded data) to receiver j, Q k is a precoded transmit data covariance matrix of user k, Q l is a precoded transmit data covariance matrix of user l, and p k is a power constraint of precoded data of user k, I is an identity matrix, α j is a priority weight of user j, α k is a priority weight of user k, and σ is a noise standard deviation.