Method of channel estimation and a channel estimator
View Patent ↗The method includes receiving a signal comprising a symbol-carrier matrix, the symbol-carrier matrix including a predetermined pattern of pilot symbols, and determining first channel estimates at pilot symbol positions of the pilot symbols in the symbol-carrier matrix. The method further includes correcting the first channel estimates at the pilot symbol positions using Bayesian estimation, and determining second channel estimates at symbol positions other than the pilot symbol positions using interpolation based on the correct first channel estimates.
1. A method of channel estimation for a multiple carrier radio transmission system, comprising:
receiving a signal comprising a symbol-carrier matrix, the symbol-carrier matrix comprising a predetermined pattern of pilot symbols;
determining first channel estimates at pilot symbol positions of the pilot symbols in the symbol-carrier matrix;
correcting the first channel estimates at the pilot symbol positions using Bayesian estimation; and
determining second channel estimates at symbol positions other than the pilot symbol positions by means of interpolation based on the corrected first channel estimates.
2. The method according to claim 1 , wherein the Bayesian estimation comprises minimum mean square estimation.
3. The method according to claim 1 , wherein the transmission system comprises at least two transmission antennas or at least two reception antennas, the method further comprising:
selecting a spatial correlation matrix out of a set of predefined spatial correlation matrices, and
employing the selected spatial correlation matrix in the Bayesian estimation.
4. The method according to claim 3 , further comprising:
calculating correlation coefficients representing spatial correlations between the at least two transmission antennas or the at least two reception antennas, or both.
5. The method according to claim 4 , further comprising:
selecting the spatial correlation matrix based on the calculated correlation coefficients.
6. The method according to claim 5 , further comprising:
selecting the spatial correlation matrix based on whether the calculated correlation coefficients are above or below one or more predetermined threshold values.
7. The method according to claim 4 , further comprising:
determining a spatial correlation matrix based on the calculated correlation coefficients, and
selecting a smaller spatial correlation matrix according to a desired complexity of the Bayesian estimation.
8. The method according to claim 3 , wherein the spatial correlation matrix comprises the form
C
MIMO
=
diag
(
R
T
,
0
⊗
R
R
,
0
,
…
,
R
T
,
L
-
1
⊗
R
R
,
L
-
1
)
=
[
R
T
⊗
R
R
⋱
R
T
⊗
R
R
]
wherein
R
T
=
[
ρ
11
T
ρ
12
T
ρ
21
T
ρ
22
T
]
,
R
R
=
[
ρ
11
R
ρ
12
R
ρ
21
R
ρ
22
R
]
wherein ρ ij T is the correlation coefficient between the
i-th and j-th transmission antenna, and ρ ij R is the correlation coefficient between i-th and j-th reception antenna.
9. The method according to claim 8 , further comprising:
correcting the first channel estimates according to the following equations:
ĥ t =Q MIMO z=C MIMO D H ( DC MIMO D H +σ 2 I ) −1 z
wherein ĥ t is the time domain channel estimate, σ 2 is the noise power, I is the unity matrix, and
z
=
[
z
(
k
0
,
k
0
+
3
)
⋮
z
(
k
N
w
-
1
,
k
N
w
-
1
+
3
)
]
=
Dh
t
+
n
and
D
=
[
D
(
k
0
,
k
0
+
3
)
⋮
D
(
k
N
w
-
1
,
k
N
w
-
1
+
3
)
]
,
n
=
[
n
(
k
0
,
k
0
+
3
)
⋮
D
(
k
N
w
-
1
,
k
N
w
-
1
+
3
)
]
wherein k i indicate the pilot sub-carriers of a first transmission antenna and k i+3 indicates the pilot sub-carriers of a second transmission antenna.
10. The method according to claim 3 , further comprising:
predefining the set of spatial correlation matrices and storing the spatial correlation matrices and their inverse matrices, respectively, in a memory.
11. The method according to claim 1 , wherein the transmission system comprises only one transmission antenna and only one reception antenna, the method further comprising:
selecting a frequency correlation matrix out of a set of predefined frequency correlation matrices, and
employing the selected frequency correlation matrix in the Bayesian estimation.
12. The method according to claim 11 , further comprising:
calculating a signal-to-noise ratio of the received signal.
13. The method according to claim 12 , further comprising:
selecting the frequency correlation matrix based on the determined signal-to-noise ratio.
14. The method according to claim 13 , further comprising:
selecting the frequency correlation matrix based on whether the calculated signal-to-noise ratio is above or below one or more predetermined threshold values.
15. The method according to claim 11 , wherein the frequency correlation matrix comprises the form
C SISO ( i,j )= si (2 πτ max ΔF ( i−j )), i= 1, . . . , N W
wherein si is the sinc function, ΔF is the sub-carrier spacing, τ max is the delay spread of the channel, i and j are pilot indices, and N W is a predefined maximum pilot index.
16. The method according to claim 15 , further comprising:
correcting the first channel estimates according to the following:
ĥ f =D·Q MIMO z=C SISO ( C SISO +σ 2 I ) −1 z
wherein ĥ f is the frequency domain channel estimate, σ 2 is the noise power, I is the unity matrix, and
z
=
[
h
(
k
0
)
h
(
k
1
)
⋮
h
(
k
N
w
-
1
)
]
+
[
n
(
k
0
)
n
(
k
1
)
⋮
n
(
k
N
w
-
1
)
]
=
h
+
n
wherein h is the factor of the frequency domain samples at the pilot sub-carriers k i .
17. The method according to claim 1 , further comprising:
determining the first channel estimates using least squares estimation.
18. The method according to claim 1 , further comprising:
determining the second channel estimates using Bayesian estimation.
19. The method according to claim 18 , wherein Bayesian estimation comprises minimum mean square estimation.
20. A channel estimator for a multiple carrier radio transmission system, comprising:
a first channel estimation stage configured to determine first channel estimates at pilot symbol positions of pilot symbols in a symbol-carrier matrix of a received signal;
a correction unit configured to correct the first channel estimates at the pilot symbol positions using Bayesian estimation; and
a second channel estimation stage configured to determine second channel estimates at symbol positions other than the pilot symbol positions by means of interpolation based on the corrected first channel estimates.
21. The channel estimator according to claim 20 , wherein the transmission system comprises at least two transmission antennas or at least two reception antennas, the channel estimator further comprising:
a selection unit configured to select a spatial correlation matrix out of a set of predefined spatial correlation matrices and supply the selected spatial correlation matrix to the correction unit.
22. The channel estimator according to claim 21 , further comprising:
a calculation unit configured to calculate correlation coefficients representing spatial correlations between the at least two transmission antennas or the at least two reception antennas, or both, and supply the calculated correlation coefficients to the selection unit.
23. The channel estimator according to claim 21 , further comprising:
a storing unit coupled to the selection unit and configured to store the set of predefined frequency correlation matrices.
24. The channel estimator according to claim 20 , wherein the transmission system comprises only one transmission antenna and only one reception antenna, the channel estimator further comprising:
a selection unit configured to select a frequency correlation matrix out of a set of predefined frequency correlation matrices and supply the selected frequency correlation matrix to the correction unit.
25. The channel estimator according to claim 24 , further comprising:
a calculation unit configured to calculate a signal-to-noise ratio of the received signal and supply the calculated signal-to-noise ratio to the selection unit.
26. The channel estimator according to claim 24 , further comprising:
a storing unit coupled to the selection unit and configured to store the set of predefined frequency correlation matrices.
27. A method of channel estimation for a multiple carrier radio transmission system, wherein the transmission system comprises at least two transmission antennas or at least two reception antennas, comprising:
receiving a signal comprising a symbol-carrier matrix, the symbol-carrier matrix comprising a predetermined pattern of pilot symbols;
determining first channel estimates at pilot symbol positions of the pilot symbols in the symbol-carrier matrix;
correcting the first channel estimates at the pilot symbol positions using Bayesian estimation;
determining second channel estimates at symbol positions other than the pilot symbol positions by means of interpolation based on the corrected first channel estimates;
selecting a spatial correlation matrix out of a set of predefined spatial correlation matrices, and
employing the selected spatial correlation matrix in the Bayesian estimation.
28. A channel estimator for a multiple carrier radio transmission system, wherein the transmission system comprises at least two transmission antennas or at least two reception antennas, comprising:
a first channel estimation stage configured to determine first channel estimates at pilot symbol positions of pilot symbols in a symbol-carrier matrix of a received signal;
a correction unit configured to correct the first channel estimates at the pilot symbol positions using Bayesian estimation;
a second channel estimation stage configured to determine second channel estimates at symbol positions other than the pilot symbol positions by means of interpolation based on the corrected first channel estimates; and
a selection unit configured to select a spatial correlation matrix out of a set of predefined spatial correlation matrices and supply the selected spatial correlation matrix to the correction unit.