Cyclic prefix-based enhanced data recovery method
View Patent ↗The cyclic prefix-based enhanced data recovery method retains the cyclic prefix (CP) upon reception and routes the CP to a data detection module to enhance the operation of the orthogonal frequency division multiplexed (OFDM) receiver whether operating in the blind, semi-blind, training, or perfectly known channel modes. Processing of the OFDM symbol and the CP is performed in the data detector and obtains data recovery by computing a maximum likelihood estimation based on the CP and the OFDM symbols.
1. A cyclic prefix-based enhanced data recovery method, comprising the steps of:
receiving data including orthogonal frequency division multiplexed (OFDM) symbols transmitted through a wireless linear channel of length L+1, wherein L is an integer, wherein each symbol has a length N, where N is an integer;
transforming the linear channel into a circular sub-channel and a linear sub-channel;
in a detector circuit, separating a cyclic prefix from the received orthogonal frequency division multiplexed symbols;
establishing a variable y i , wherein y i represents a cyclic prefix of output of the linear sub-channel at a particular time i, wherein y i is given by y i = X i h i + n i , where h i represents a channel effect of the transmitted data at the time i, n i represents noise of the transmitted data, and X i is a matrix given by X i = X Li + X Ui−1 , where X Li is a matrix composed of cyclic prefixes of current unknown OFDM symbols x i and X Ui−1 is a matrix composed of cyclic prefixes of previous known OFDM symbols x i-1 , where
X
_
Li
=
(
x
_
i
(
0
)
0
…
0
x
_
i
(
1
)
x
_
i
(
0
)
…
0
⋮
⋱
⋱
⋮
x
_
i
(
L
-
1
)
…
x
_
i
(
0
)
0
)
and
X
_
Ui
-
1
=
(
0
x
_
i
-
1
(
L
-
1
)
…
x
_
i
-
1
(
0
)
0
0
…
x
_
i
-
1
(
1
)
⋮
⋱
⋱
⋮
0
…
0
x
_
i
-
1
(
L
-
1
)
)
;
establishing a variable x i , wherein x i represents a cyclic prefix of output of the circular sub-channel at the time i, wherein x i is a length-N, zero-padded version of x i ;
performing FFT of the received OFDM symbols without the cyclic prefix;
recovering the data by maximum likelihood estimation using the cyclic prefix and the received OFDM symbols, wherein a blind method for channel estimation is utilized, thereby making collective use of natural constraints of a wireless communications protocol and channel;
refining the data by a refinement method comprising the steps of:
obtaining an initial estimate using a plurality of pilots and frequency correlation;
establishing an objective function Z;
obtaining a gradient of the objective function Z subject to a constant modulus constraint on data given by φ j =|χ i (j)| 2 =E x for integer j=1, 2, 3, . . . N, wherein Z=∥ y i −Bχ i *−Cχ i *∥ 2 , where χ i is an N-point FFT of the x i , χ i * is a convolution of χ i , B=(1/E x ) X Ui−1 Q L+1 D Y and C=(1/E x ) X Li Q L+1 D Y , Q L+1 represents the first L+1 rows of an inverse FFT matrix Q, and D Y is a diagonal matrix with elements on the diagonal being equal to N-point FFT of y i ;
obtaining a Hessian of the objective function subjected to the constant modulus constraint on data;
using the gradient and the Hessian in Newton's method.
2. The cyclic prefix-based enhanced data recovery method according to claim 1 , wherein performance of the maximum likelihood estimation includes minimizing a norm over all possible combinations of the received data.
3. The cyclic prefix-based enhanced data recovery method according to claim 2 , wherein the norm depends on input and output data only.
4. The cyclic prefix-based enhanced data recovery method according to claim 1 , wherein the constraints include a finite delay spread constraint on the wireless linear channel, a finite alphabet constraint on data, and the cyclic prefix.
5. The cyclic prefix-based enhanced data recovery method according to claim 1 , further comprising the step of processing block fading channels.
6. The cyclic prefix-based enhanced data recovery method according to claim 1 , further comprising the step of detecting symbols for channels with zeros on an FFT grid.
7. The cyclic prefix-based enhanced data recovery method according to claim 1 , wherein data is detected in presence of channel nulls.
8. A cyclic prefix-based enhanced data recovery method, comprising the steps of:
receiving data including orthogonal frequency division multiplexed (OFDM) symbols transmitted through a wireless linear channel of length L+1, wherein L is an integer, wherein each symbol has a length N, where N is an integer;
transforming the linear channel into a circular sub-channel and a linear sub-channel;
in a detector circuit, separating a cyclic prefix from the received orthogonal frequency division multiplexed symbols;
establishing a variable y i , wherein y i represents a cyclic prefix of output of the linear sub-channel at a particular time i, wherein y 1 is given by y i = X i h i +n i , where h i represents a channel effect of the transmitted data at the time i, n i represents noise of the transmitted data, and X i is a matrix given by X i = X Li + X Ui−1 , where X Li is a matrix composed of cyclic prefixes of current unknown OFDM symbols x i and X Ui−1 is a matrix composed of cyclic prefixes of previous known OFDM symbols x i−1 where
X
_
Li
=
(
x
_
i
(
0
)
0
…
0
x
_
i
(
1
)
x
_
i
(
0
)
…
0
⋮
⋱
⋱
⋮
x
_
i
(
L
-
1
)
…
x
_
i
(
0
)
0
)
and
X
_
Ui
-
1
=
(
0
x
_
i
-
1
(
L
-
1
)
…
x
_
i
-
1
(
0
)
0
0
…
x
_
i
-
1
(
1
)
⋮
⋱
⋱
⋮
0
…
0
x
_
i
-
1
(
L
-
1
)
)
;
establishing a variable x i , wherein x i represents a cyclic prefix of output of the circular sub-channel at the time i, wherein x i is a length-N, zero-padded version of x i ;
performing FFT of the received OFDM symbols without the cyclic prefix;
recovering the data by a Genetic process using the cyclic prefix and the received symbols, wherein a blind method for channel estimation is utilized, thereby making collective use of natural constraints of a wireless communications protocol and channel, the Genetic process comprising the steps of:
obtaining an initial estimate using a plurality of pilots and frequency correlation;
establishing an objective function Z;
obtaining a gradient of the objective function Z subject to a constant modulus constraint on data given by φ j =|χ i (j)| 2 =E x for integer j=1, 2, 3, . . . N, wherein Z=∥ y i −Bχ i *−Cχ i *∥ 2 , where χ i is an N-point FFT of x i , χ i * is a convolution of χ i , B=(1/E x ) X Ui−1 Q L+1 D Y and C=(1/E x ) X Li Q L+1 D Y , Q L+1 represents the first L+1 rows of an inverse FFT matrix Q, and D Y is a diagonal matrix with elements on the diagonal being equal to N-point FFT of y i ;
obtaining a Hessian of the objective function subjected to the constant modulus constraint on data;
using the gradient and the Hessian in Newton's method.
9. The cyclic prefix-based enhanced data recovery method according to claim 8 , wherein data is detected in presence of channel nulls.