Method and device for determining smooth window length in channel estimation
A method and a device for determining a smooth window length in channel estimation is provided. The method includes: obtaining autocorrelation values of pilot signals, where the autocorrelation values include first autocorrelation values and second autocorrelation values, at least two first autocorrelation values exist, and at least one second autocorrelation value exists; determining a speed range according to symbols of the first autocorrelation values or the symbols of the first autocorrelation values and a ratio value between the second autocorrelation value and one of the first autocorrelation values; and determining the smooth window length according to the speed range. Through the method provided in embodiments of the present invention, the problems caused by a fixed smooth window length may be avoided, and this improves system performance.
1. A method for determining a smooth window length in channel estimation the method comprising:
obtaining autocorrelation values of pilot signals, wherein the autocorrelation values comprise at least two first autocorrelation values and at least one second autocorrelation value wherein the autocorrelation values of the pilot signals are autocorrelation values in non-zero correlation intervals;
determining a speed range according to symbols of the first autocorrelation values or according to the symbols of the first autocorrelation values and a ratio value between the second autocorrelation value and one of the first autocorrelation values; and
determining the smooth window length according to the speed range;
wherein a calculation formula for obtaining the autocorrelation values of the pilot signals is as follows:
R
(
k
)
=
∑
f
=
1
M
R
f
(
k
)
wherein R(k) is the autocorrelation value of the pilot signal, k is a correlation interval and not equal to 0, M is the preset number of multi-path search periods, and R f (k) is an autocorrelation value of a pilot signal within a multi-path search period of an index f , and the calculation formula of R f (k) is as follows:
R
f
(
k
)
=
∑
t
=
1
L
mpsf
∑
j
=
1
N
f
{
Re
[
A
j
(
t
-
k
)
×
A
j
*
(
t
)
]
}
wherein L mpsf is the number of pilot channel time slots contained in the multi-path search period of the index f, N f is the number of effective paths in the multi-path search period of the index f, Re( )is an operation for taking a real part, A* j is a conjugate function of A j (t), A j (t) is an accumulated value of the antenna separation values of the pilot signals within each time slot, and a calculation formula of A j (t) is as follows:
A
j
(
t
)
=
∑
i
=
1
N
/
2
A
i
,
j
(
t
)
wherein N is the number of pilot signal symbols within each time slot, A i,j (t) is an antenna separation value of a pilot signal of an index i corresponding to a path of an index j within a time slot of an index t.
2. The method according to claim 1 , wherein the first autocorrelation values are R(k 1 ) and R(k 2 ), and the second autocorrelation value is R(k 3 ), wherein |k 1 |<|k 2 |<|k 3 |.
3. The method according to claim 2 , wherein the determining the speed range according to the symbols of the first autocorrelation values or the symbols of the first autocorrelation values and the ratio value between the second autocorrelation value and one of the first autocorrelation values and the determining the smooth window length according to the speed range comprise:
when either of R(k 1 ) and R(k 2 ) is a negative value, determining a first speed range, and determining the smooth window length corresponding to the first speed range as a first length value; and
when R(k 1 ) and R(k 2 ) are both positive values, determining a second speed range, determining a fine range within the second speed range according to a ratio value between R(k 3 ) and R(k 1 ) and a preset threshold when in the second speed range, and determining the smooth window length according to the fine range,
wherein the first length value is smaller than or equal to the smooth window length within the second speed range.
4. The method according to claim 3 , wherein the determining the fine range within the second speed range according to the ratio value between R(k 3 ) and R(k 1 ) and the preset threshold and determining the smooth window length according to the fine range comprise:
determining the smooth window length as a second length value when the ratio value between R(k 3 ) and R(k 1 ) is greater than a first preset threshold;
determining the smooth window length as a third length value when the ratio value between R(k 3 ) and R(k 1 ) is between the first preset threshold and a second preset threshold;
determining the smooth window length as a fourth length value when the ratio value between R(k 3 ) and R(k 1 ) is smaller than the second preset threshold,
wherein the second length value is greater than the third length value, the third length value is greater than the fourth length value, and the fourth length value is greater than or equal to the first length value.
5. The method according to claim 4 , wherein the ratio value between R(k 3 ) and R(k 1 ) is an α-filtered ratio value.
6. The method according to claim 2 , wherein k 1 =1, k 2 =2, and k 3 =4.
7. A device for determining a smooth window length in channel estimation, comprising:
an obtaining module, configured to obtain autocorrelation values of pilot signals, wherein the autocorrelation values comprise at least two first autocorrelation values and at least one second autocorrelation value; and
a determining module, configured to:
determine a speed range according to symbols of the first autocorrelation values or according to the symbols of the first autocorrelation values and a ratio value between the second autocorrelation value and one of the first autocorrelation values, and
determine the smooth window length according to the speed range;
wherein the autocorrelation values of the pilot signals are autocorrelation values in non-zero correlation intervals, and the autocorrelation value of the pilot signals is as follows:
R
(
k
)
=
∑
f
=
1
M
R
f
(
k
)
wherein R(k) is the autocorrelation value of the pilot signal, k is a correlation interval and not equal to 0, M is the preset number of multi-path search periods, and R f (k) is an autocorrelation value of a pilot signal within a multi-path search period of an index f, and the calculation formula of R f (k) is as follows:
R
f
(
k
)
=
∑
t
=
1
L
mpsf
∑
j
=
1
N
f
{
Re
[
A
j
(
t
-
k
)
×
A
j
*
(
t
)
]
}
wherein L mpsf is the number of pilot channel time slots contained in the multi-path search period of the index f, N f is the number of effective paths in the multi-path search period of the index f, Re( ) is an operation for taking a real part, A* j (t) is a conjugate function of A j (t), A j (t) is an accumulated value of the antenna separation values of the pilot signals within each time slot, and a calculation formula of A j (t) is as follows:
A
j
(
t
)
=
∑
i
=
1
N
/
2
A
i
,
j
(
t
)
wherein N is the number of pilot signal symbols within each time slot, A i,j (t) is an antenna separation value of a pilot signal of an index i corresponding to a path of an index j within a time slot of an index t.
8. The device according to claim 7 , wherein the first autocorrelation values are R(k 1 ) and R(k 2 ) , and the second autocorrelation value is R(k 3 ), and the determining module is further configured to determine a first speed range and determine the smooth window length corresponding to the first speed range as a first length value when either of R(k 1 ) and R(k 2 ) is a negative value; or determine a second speed range when R(k 1 ) and R(k 2 ) are both positive values, determine a fine range within the second speed range according to a ratio value between R(k 3 ) and R(k 1 ) and a preset threshold when in the second speed range, and determine the smooth window length according to the fine range, wherein the first length value is smaller than or equal to the smooth window length within the second speed range, and |k 1 |<|k 2 |<|k 3 |.
9. The device according to claim 8 , wherein
the determining module is further configured to determine the smooth window length as a second length value when the ratio value between R(k 3 ) and R(k 1 ) is greater than a first preset threshold, determine the smooth window length as a third length value when the ratio value between R(k 3 ) and R(k 1 ) is between the first preset threshold and a second preset threshold, or determine the smooth window length as a fourth length value when the ratio value between R(k 3 ) and R(k 1 ) is smaller than the second preset threshold, wherein the second length value is greater than the third length value, the third length value is greater than the fourth length value, and the fourth length value is greater than or equal to the first length value.
10. The method according to claim 8 , wherein k 1 =1, k 2 =2, and k 3 =4.