Channel and noise estimation method and associated apparatus
View Patent ↗A channel and noise estimation method includes: performing channel estimation on a received signal to obtain a real channel estimation value; filtering the real channel estimation value to obtain a filtered channel estimation value; calculating a biased noise power value, a biased signal power value, a biased noise correlation and a biased signal correlation according to the filtered channel estimation value; calculating an unbiased noise power value, an unbiased signal power value, an unbiased noise correlation and an unbiased signal correlation according to the biased noise power value, the biased signal power value, the biased noise correlation and the biased signal correlation.
1. A channel and noise estimation method, applied to a channel and noise estimation apparatus of a wireless communication system, the channel and noise estimation method comprising:
performing channel estimation on a received signal to obtain a real channel estimation value;
filtering the real channel estimation value to obtain a filtered channel estimation value;
calculating a biased noise power value, a biased signal power value, a biased noise correlation and a biased signal correlation according to the filtered channel estimation value; and
respectively calculating an unbiased noise power value, an unbiased signal power value, an unbiased noise correlation and an unbiased signal correlation according to the biased noise power value, the biased signal power value, the biased noise correlation and the biased signal correlation by using a first inverted matrix;
wherein the step of performing the channel estimation on the received signal to obtain the real channel estimation value comprises:
performing the channel estimation on at least one signal received by a first reception antenna and a second reception antenna, respectively, to obtain corresponding real channel estimation values H LS 1 and H LS 2 , where H LS 1 represents the real channel estimation value corresponding to the first reception antenna, and H LS 2 represents the real channel estimation value corresponding to the second reception antenna;
wherein the step of filtering the real channel estimation value to obtain the filtered channel estimation value comprises:
filtering the real channel estimation values H LS 1 and H LS 2 by utilizing a first filter and a second filter, respectively, to obtain corresponding filtered channel estimation values Y w 1 (k), Y w 2 (k), Y u 1 (k) and Y u 2 (k), where Y w 1 (k) represents the filtered channel estimation value obtained from filtering H LS 1 by the first filter, Y w 2 (k) represents the filtered channel estimation value obtained from filtering H LS 2 by the first filter, Y u 1 (k) represents the filtered channel estimation value obtained from filtering H LS 1 by the second filter, and Y u 2 (k) represents the filtered channel estimation value obtained from filtering H LS 2 by the second filter;
wherein values of H LS 1 and H LS 2 are:
H LS 1 =H ideal 1 +v 1 (1)
H LS 2 =H ideal 2 +v 2 (2)
where H ideal 1 represents an ideal channel estimation value of the first reception antenna, v 1 represents a noise value of the first reception antenna, H ideal 2 represents an ideal channel estimation value of the second reception antenna, and v 2 represents a noise value of the second reception antenna; and
values of Y w 1 (k), Y w 2 (k), Y u 1 (k) and Y u 2 (k) are calculated through following equations:
Y w 1 ( k )= wH LS 1 (3)
Y w 2 ( k )= wH LS 2 (4)
Y u 1 ( k )= uH LS 1 (5)
Y u 2 ( k )= uH LS 2 (6)
where w represents a coefficient row vector of the first filter, k represents a k th subcarrier on a frequency domain, and u represents a coefficient row vector of the second filter; and
wherein the step of calculating the biased noise power value, the biased signal power value, the biased noise correlation and the biased signal correlation according to the filtered channel estimation value comprises:
calculating biased noise power values S w 1 and S w 2 , biased signal power values S u 1 and S u 2 , a biased noise correlation P w 12 between Y w 1 (k) and Y w 2 (k), and a biased signal correlation P u 12 between Y u 1 (k) and Y u 2 (k) corresponding to the first reception antenna and the second reception antenna according to Y w 1 (k), Y w 2 (k), Y u 1 (k) and Y u 1 (k), respectively; and
values of S w 1 , S w 2 , S u 1 , S u 2 , P w 12 and P u 12 are calculated through following equations:
S w 1 =E k ( Y w 1 ( k )· Y w 1 ( k )*) (7)
S w 2 =E k ( Y w 2 ( k )· Y w 2 ( k )*) (8)
S u 1 =E k ( Y u 1 ( k )· Y u 1 ( k )*) (9)
S u 2 =E k ( Y u 2 ( k )· Y u 2 ( k )*) (10)
P w 12 =E k ( Y w 1 ( k )· Y w 2 ( k )*) (11)
P u 12 =E k ( Y u 1 ( k )· Y u 2 ( k )*) (12)
where S w 1 represents the biased noise power value corresponding to the first reception antenna, S w 2 represents the biased noise power value corresponding to the second reception antenna, S u 1 represents the biased signal power value corresponding to the first reception antenna, S u 2 represents the biased signal power value corresponding to the second reception antenna, P w 12 represents the biased noise correlation between Y w 1 (k) and Y w 2 (k), P u 12 represents the biased signal correlation between Y u 1 (k) and Y u 2 (k), E k (●) represents a frequency-domain average, and (●)* represents a conjugation; and
wherein after the step of calculating the unbiased noise power value, the unbiased signal power value, the unbiased noise correlation and the unbiased signal correlation according to the biased noise power value, the biased signal power value, the biased noise correlation and the biased signal correlation, the method further comprising:
frequency-domain smoothing the unbiased noise power value, the unbiased signal power value, the unbiased noise correlation and the unbiased signal correlation;
time-domain filtering the frequency-domain smoothed unbiased noise power value, unbiased signal power value, unbiased noise correlation and unbiased signal correlation; and
demodulating the filtered unbiased noise power value, unbiased signal power value, unbiased noise correlation and unbiased signal correlation to restore an original transmission signal.
2. The channel and noise estimation method according to claim 1 , wherein the first and second filters are linearity-unassociated.
3. The channel and noise estimation method according to claim 1 , wherein the step of calculating the unbiased noise power value, the unbiased signal power value, the unbiased noise correlation and the unbiased signal correlation according to the biased noise power value, the biased signal power value, the biased noise correlation and the biased signal correlation comprises:
obtaining according to equations (1), (3) and (7) that:
S
w
1
=
E
k
(
Y
w
1
(
k
)
·
Y
w
1
(
k
)
*
)
=
wE
[
(
H
ideal
1
+
v
1
)
(
H
ideal
1
+
v
1
)
H
]
w
H
=
w
(
α
1
2
R
HH
+
σ
1
2
I
)
w
H
=
α
1
2
wR
HH
w
H
+
σ
1
2
ww
H
;
(
13
)
obtaining according to equations (1), (5) and (9) that:
S
u
1
=
E
k
(
Y
u
1
(
k
)
·
Y
u
1
(
k
)
*
)
=
uE
[
(
H
ideal
1
+
v
1
)
(
H
ideal
1
+
v
1
)
H
]
u
H
=
u
(
α
1
2
R
HH
+
σ
1
2
I
)
u
H
=
α
1
2
uR
HH
u
H
+
σ
1
2
uu
H
;
(
14
)
by establishing a simultaneous equation from equations (13) and (14), obtaining that:
(
S
w
1
S
u
1
)
=
(
wR
HH
w
H
ww
H
uR
HH
u
H
uu
H
)
(
α
1
2
σ
1
2
)
;
(
15
)
(
wR
HH
w
H
ww
H
uR
HH
u
H
uu
H
)
by inverting the matrix in equation (15), according to the biased noise power value S w 1 corresponding to the first reception antenna and the biased signal power value S u 1 corresponding to the first reception antenna, obtaining that:
(
α
1
2
σ
1
2
)
=
(
wR
HH
w
H
ww
H
uR
HH
u
H
uu
H
)
-
1
(
S
w
1
S
u
1
)
,
(
16
)
where E(●) represents an average, (●) H represents a conjugate transpose, I represents a unit matrix, R HH represents a correlation matrix of H LS 1 and H LS 2 on the subcarriers, α 1 2 the unbiased signal power value corresponding to the first reception antenna, and σ 1 2 represents the unbiased noise power value corresponding to the first reception antenna;
for an unbiased signal power value α 2 2 corresponding to the second reception antenna and an unbiased noise power value σ 2 2 corresponding to the second reception antenna, obtaining according to equations (1), (2), (3), (4) and (11) that:
P
w
12
=
E
k
(
Y
w
1
(
k
)
·
Y
w
2
(
k
)
*
)
=
wE
[
(
H
ideal
1
+
v
1
)
(
H
ideal
2
+
v
2
)
H
]
w
H
=
w
(
β
12
R
HH
+
γ
12
I
)
w
H
=
β
12
wR
HH
w
H
+
γ
12
ww
H
;
(
17
)
obtaining according to equations (1), (2), (5), (6) and (12) that:
P
u
12
=
E
k
(
Y
u
1
(
k
)
·
Y
u
2
(
k
)
*
)
=
uE
[
(
H
ideal
1
+
v
1
)
(
H
ideal
2
+
v
2
)
H
]
u
H
=
u
(
β
12
R
HH
+
γ
12
I
)
u
H
=
β
12
uR
HH
u
H
+
γ
12
uu
H
;
(
18
)
by establishing a simultaneous equation from equations (17) and (18), obtaining that:
(
P
w
12
P
u
12
)
=
(
wR
HH
w
H
ww
H
uR
HH
u
H
uu
H
)
(
β
12
γ
12
)
;
(
19
)
and
by inverting the matrix
(
wR
HH
w
H
ww
H
uR
HH
u
H
uu
H
)
in equation (19), according to the biased noise correlation P w 12 and the biased signal correlation P u 12 , obtaining that:
(
β
12
γ
12
)
=
(
wR
HH
w
H
ww
H
uR
HH
u
H
uu
H
)
-
1
(
P
w
12
P
u
12
)
,
(
20
)
where β 12 represents the unbiased signal correlation between Y u 1 (k) and Y u 2 (k), and γ 12 represents the unbiased noise correlation between Y w 1 (k) and Y w 2 (k).
4. A channel and noise estimation apparatus of a wireless communication system, the channel and noise estimation apparatus comprising:
a channel estimation module, wherein the channel estimation module is stored in a non-transitory computer-readable storage media, for performing channel estimation on a received signal to obtain a real channel estimation value;
a first filter module and a second filter module, wherein the first filter module and second filter module are stored in a non-transitory computer-readable storage media, for filtering the real channel estimation value to obtain a filtered channel estimation value, respectively;
a first calculation module, wherein the first calculation module is stored in a non-transitory computer-readable storage media, for calculating a biased noise power value, a biased signal power value, a biased noise correlation and a biased signal correlation according to the filtered channel estimation values; and
a second calculation module, wherein the second calculation module is stored in a non-transitory computer-readable storage media, for respectively calculating an unbiased noise power value, an unbiased signal power value, an unbiased noise correlation and an unbiased signal correlation according to the biased noise power value, the biased signal power value, the biased noise correlation and the biased signal correlation by using first inverted matrix;
a frequency-domain smoothing module, wherein the frequency-domain smoothing module is stored in non-transitory computer-readable storage media, for frequency-domain smoothing the unbiased noise power value, the unbiased signal power value, the unbiased noise correlation and the unbiased signal correlation;
a time-domain filter module, wherein the time-domain filter module is stored in non-transitory computer readable storage media, for time-domain filtering the frequency-domain smoothed unbiased noise power value, unbiased signal power value, unbiased noise correlation and unbiased signal correlation; and
a demodulation circuit, for demodulating the filtered unbiased noise power value, unbiased signal power value, unbiased noise correlation and unbiased signal correlation to restore an original transmission signal;
wherein the channel estimation module performs the channel estimation on at least one signal received by a first reception antenna and a second reception antenna, respectively, to obtain corresponding real channel estimation values H LS 1 and H LS 2 , where H LS 1 represents the real channel estimation value corresponding to the first reception antenna, and H LS 2 represents the real channel estimation value corresponding to the second reception antenna;
wherein the first and second filter modules filter the real channel estimation values H LS 1 and H LS 2 , respectively, to obtain corresponding filtered channel estimation values Y w 1 (k), Y w 2 (k), Y w 1 (k) and Y u 2 (k), where Y w 1 (k) represents the filtered channel estimation value obtained from filtering H LS 1 by the first filter module, Y w 2 (k) represents the filtered channel estimation value obtained from filtering H LS 2 by the first filter module, Y u 1 (k) represents the filtered channel estimation value obtained from filtering H LS 1 by the second filter module, and Y u 2 (k) represents the filtered channel estimation value obtained from filtering H LS 2 by the second filter module; and the first and second filter modules are linearity-unassociated;
wherein values of H LS 1 and H LS 2 are:
H LS 1 =H ideal 1 +v 1 (1)
H LS 2 =H ideal 2 +v 2 (2)
where H ideal 1 represents an ideal channel estimation value of the first reception antenna, v 1 represents a noise value of the first reception antenna, H ideal represents an ideal channel estimation value of the second reception antenna, and v 2 represents a noise value of the second reception antenna; and
values of Y w 1 (k), Y w 2 (k), Y u 1 (k) are calculated through following equations:
Y w 1 ( k )= wH LS 1 (3)
Y w 2 ( k )= wH LS 2 (4)
Y u 1 ( k )= uH LS 1 (5)
Y u 2 ( k )= uH LS 2 (6)
where w represents a coefficient row vector of the first filter, k represents a k th subcarrier on a frequency domain, and u represents a coefficient row vector of the second filter; and
wherein the first calculation modules calculates biased noise power values S w 1 and S w 2 , biased signal power values S u 1 and S u 2 , a biased noise correlation P w 12 between Y w 1 (k) and Y w 2 (k), and a biased signal correlation P u 12 between Y w 1 (k) and Y w 1 (k) corresponding to the first reception antenna and the second reception antenna according to Y w 1 (k), Y w 1 (k), Y w 1 (k) and Y w 2 (k), respectively; and
values of S w 1 , S w 2 , S u 1 , S u 2 , P w 12 and P u 12 are calculated through following equations:
S w 1 =E k ( Y w 1 ( k )· Y w 1 ( k )*) (7)
S w 2 =E k ( Y w 2 ( k )· Y w 2 ( k )*) (8)
S u 1 =E k ( Y u 1 ( k )· Y u 1 ( k )*) (9)
S u 2 =E k ( Y u 2 ( k )· Y u 2 ( k )*) (10)
P w 12 =E k ( Y w 1 ( k )· Y w 2 ( k )*) (11)
P u 12 =E k ( Y u 1 ( k )· Y u 2 ( k )*) (12)
where S w 1 represents the biased noise power value corresponding to the first reception antenna, S w 2 represents the biased noise power value corresponding to the second reception antenna, S u 1 represents the biased signal power value corresponding to the first reception antenna, S u 2 represents the biased signal power value corresponding to the second reception antenna, P w 12 represents the biased noise correlation between Y w 1 (k) and Y w 1 (k), P u 12 represents the biased signal correlation between Y w 1 (k) and Y w 1 (k), E k (●) represents a frequency-domain average, and (●)* represents a conjugation.
5. The channel and noise estimation apparatus according to claim 4 , wherein the second calculation modules obtains according to equations (1), (3) and (7) that:
S
w
1
=
E
k
(
Y
w
1
(
k
)
·
Y
w
1
(
k
)
*
)
=
wE
[
(
H
ideal
1
+
v
1
)
(
H
ideal
1
+
v
1
)
H
]
w
H
=
w
(
α
1
2
R
HH
+
σ
1
2
I
)
w
H
=
α
1
2
wR
HH
w
H
+
σ
1
2
ww
H
;
(
13
)
the second calculation module obtains according to equations (1), (5) and (9) that:
S
u
1
=
E
k
(
Y
u
1
(
k
)
·
Y
u
1
(
k
)
*
)
=
uE
[
(
H
ideal
1
+
v
1
)
(
H
ideal
1
+
v
1
)
H
]
u
H
=
u
(
α
1
2
R
HH
+
σ
1
2
I
)
u
H
=
α
1
2
uR
HH
u
H
+
σ
1
2
uu
H
;
(
14
)
by establishing a simultaneous equation from equations (13) and (14), the second calculation module obtains that:
(
S
w
1
S
u
1
)
=
(
wR
HH
w
H
ww
H
uR
HH
u
H
uu
H
)
(
α
1
2
σ
1
2
)
;
(
15
)
by inverting the matrix
(
wR
HH
w
H
ww
H
uR
HH
u
H
uu
H
)
in equation (15), according to the biased noise power value S w 1 corresponding to the first reception antenna and the biased signal power value S u 1 corresponding to the first reception antenna, the second calculation module obtains that:
(
α
1
2
σ
1
2
)
=
(
wR
HH
w
H
ww
H
uR
HH
u
H
uu
H
)
-
1
(
S
w
1
S
u
1
)
;
(
16
)
where E(●) represents an average, (●) H represents a conjugate transpose, I represents a unit matrix, R HH represents a correlation matrix of H LS 1 and H LS 2 on the subcarriers, α 1 2 the unbiased signal power value corresponding to the first reception antenna, and σ 1 2 represents the unbiased noise power value corresponding to the first reception antenna;
the second calculation module further obtains an unbiased signal power value α 2 2 corresponding to the second reception antenna and an unbiased noise power value σ 2 2 corresponding to the second reception antenna;
the second calculation module obtains according to equations (1), (2), (3), (4) and (11) that:
P
w
12
=
E
k
(
Y
w
1
(
k
)
·
Y
w
2
(
k
)
*
)
=
wE
[
(
H
ideal
1
+
v
1
)
(
H
ideal
2
+
v
2
)
H
]
w
H
=
w
(
β
12
R
HH
+
γ
12
I
)
w
H
=
β
12
wR
HH
w
H
+
γ
12
ww
H
;
(
17
)
the second calculation module obtains according to equations (1), (2), (5), (6) and (12) that:
P
u
12
=
E
k
(
Y
u
1
(
k
)
·
Y
u
2
(
k
)
*
)
=
uE
[
(
H
ideal
1
+
v
1
)
(
H
ideal
2
+
v
2
)
H
]
u
H
=
u
(
β
12
R
HH
+
γ
12
I
)
u
H
=
β
12
uR
HH
u
H
+
γ
12
uu
H
;
(
18
)
by establishing a simultaneous equation from equations (17) and (18), the second calculation module obtains that:
(
P
w
12
P
u
12
)
=
(
wR
HH
w
H
ww
H
uR
HH
u
H
uu
H
)
(
β
12
γ
12
)
;
(
19
)
and
by inverting the matrix
(
wR
HH
w
H
ww
H
uR
HH
u
H
uu
H
)
in equation (19), according to the biased noise correlation P w 12 and the biased signal correlation P u 12 , the second calculation module obtains that:
(
β
12
γ
12
)
=
(
wR
HH
w
H
ww
H
uR
HH
u
H
uu
H
)
-
1
(
P
w
12
P
u
12
)
;
(
20
)
where β 12 represents the unbiased signal correlation between Y u 1 (k) and Y u 2 (k), and γ 12 represents the unbiased noise correlation between Y w 1 (k) and Y w 2 (k).