Weighted tone reservation for OFDM PAPR reduction
A weighted tone reservation (WTR) method and system are disclosed, for PAPR reduction. The WTR method solves the peak re-growth problem with minimum overhead. By avoiding the drawbacks of conventional tone reservation approaches, systems employing the WTR method may experience a significant PAPR reduction. The WTR method may be applied to next generation OFDMA-based wireless broadband technologies to increase system throughput and cell coverage.
1. A weighted tone reservation method, comprising:
obtaining a sequence, X, wherein X is derived from binary data input to be transformed and transmitted wirelessly over an antenna using orthogonal frequency division multiple access modulation;
calculating a second sequence, X_PAPR, using the sequence, X and the following equation:
X_PAPR
=
10
log
10
Max
(
X
2
)
E
(
X
2
)
.
wherein |X| is an amplitude profile of sequence, X, Max(|X| 2 ) is a maximum of the square of the amplitude profile of sequence, X, and E(|X| 2 ) is a mean of the square of the amplitude profile of sequence, X; and
generating a third sequence, X NEW , a linear addition of the sequence, X, and a fourth sequence generated by reserved tones on a frequency domain if the second sequence, X_PAPR, exceeds a threshold value;
wherein the third sequence, Xnew, is the same as the sequence, X, if the second sequence, X_PAPR, does not exceed the threshold value;
determining that the second sequence, X_PAPR, exceeds the threshold value;
calculating a clipping threshold, CT, based on a predetermined clipping rate, CR; and
using the clipping threshold to generate a fifth sequence, X p , wherein the fifth sequence, X p , is used to generate the new sequence, X NEW .
2. The weighted tone reservation method of claim 1 , calculating a clipping threshold, CT, based on a predetermined clipping rate, CR, further comprising:
calculating the clipping threshold, CT, based on the predetermined clipping rate, CR, wherein the clipping rate, CR, is between 0 and √2.
3. The weighted tone reservation method of claim 1 , calculating a clipping threshold, CT, based on a predetermined clipping rate, CR, further comprising:
calculating the clipping threshold, CT, using the following equation:
CT=CR× √{square root over (2)}×std( X ),
wherein std(X) is a standard deviation of the sequence, X.
4. The weighted tone reservation method of claim 1 , further comprising:
calculating the fifth sequence, X p , using the following sequence:
for
i
=
0
to
N
-
1
if
X
>
CT
,
then
Xp
(
i
)
=
X
(
i
)
-
X
(
i
)
X
(
i
)
×
CT
;
else
Xp
(
i
)
=
0
;
end
if
;
and
calculating a weighted factor sequence, D, using the following equation:
D ( i )=| X ( i )| 2 ,
wherein |X| is an amplitude profile of the sequence, X, X(i) is the ith entry of sequence, X, Xp(i) is the ith entry of sequence, Xp, and D(i) is the ith entry of the weighted factor sequence, D.
5. The weighted tone reservation method of claim 4 , further comprising:
using the fifth sequence, X p , to calculate arrays B, E, G, and H.
6. A communication method, comprising:
using an inverse fast Fourier transform to convert binary input data for transmission into a sequence, X;
applying a weighted tone reservation algorithm to find a time domain sequence, X p , to satisfy the equation, PAPR(X−X p )=PAPR 0 , wherein PAPR is a peak-to-average power ratio of a signal and PAPR 0 is a target peak-to-average power ratio of the signal;
generating a signal similar to the sequence, X p , by transmitting a second sequence, C, in reserved tones, wherein the second sequence, C, is generated using the following criteria:
C
=
arg
C
min
D
T
X
p
-
A
C
2
wherein A is an inverse fast Fourier transform matrix of the sequence, C, D is a weight function, D T is a transpose of weight function, D, and arg min is the argument minimum of C; and
transmitting a third sequence, X new , having reduced peak-to-average power ratio over the sequence, X, derived from the second sequence, C.
7. The communication method of claim 6 , further comprising:
calculating the second sequence, C, using the following formula:
C =( A H WA ) −1 A H WX p ,
wherein H is a conjugate transpose where A H =(A′)*, A′ is a transpose of matrix A, W is a weighted array, Xp is a time domain signal, and A* is a conjugate complex of matrix A.
8. The communication method of claim 6 , further comprising:
generating randomized data from the binary input data;
encoding the randomized data to produce encoded data; and
generating interleaved data from the encoded data.
9. The communication method of claim 8 , further comprising:
transmitting the third sequence, X new , to a receiver.
10. A method, comprising:
obtaining a sequence X, wherein X is derived from binary data input;
calculating a second sequence, C, using the formula, C=(A H WA) −1 A H WX p , wherein H is a conjugate transpose where A H =(A′)*, A′ is a transpose of matrix A, A* is a conjugate complex of matrix A, X p is a time domain signal, and W is a weighted array; and
transmitting the second sequence in reserved tones of the binary input data.
11. The method of claim 10 , calculating a second sequence, C, further comprising:
generating the second sequence, C, using the following criteria:
C
=
arg
C
min
D
T
X
p
-
A
C
2
,
wherein D is a weight vector and arg min is the argument minimum of C.
12. The method of claim 11 , further comprising:
using the vector, D, to obtain the weighted array, as follows:
W
=
{
D
1
0
…
…
0
0
…
0
…
…
…
0
D
i
0
…
…
…
0
…
0
0
…
…
0
D
N
}
N
×
N
wherein N is the fast Fourier transform size and i is an integer between 1 and N.
13. The method of claim 12 , further comprising:
calculating the new sequence, X new , using the sequence, X, the matrix, A, and the second sequence, C, in the following formula:
X new =X−AC.