Adaptive quantization method and apparatus for an OFDM receiver
View Patent ↗A quadrature frequency division multiplexing (“OFDM”) wireless receiver, including methods and devices for adaptive quantization of OFDM signals according to modulation and coding schemes and sub-carrier frequency responses, is provided. Efficient quantization may be utilized to reduce the large dynamic range of signals to achieve circuit simplification and chip area reduction. In one embodiment, a quantization circuit includes a quantization selector to select quantization thresholds according to modulation and coding schemes and sub-carrier frequency responses, and a non-uniform quantizer to reduce input dynamic range so that an output is represented by fewer bits than an input.
1. A method of dynamic quantization in an orthogonal frequency division multiplexing (OFDM) system, the method comprising:
receiving quadrature amplitude modulation (QAM) symbols for each of a number of “m” sub-carrier channels, m being an integer greater than 1;
computing a channel power P CH (i)=|G CH (i)| 2 for each of the m sub-carrier channels, where G ch (i) is the i -th sub-carrier channel frequency response, i being an integer from 0 to m−1;
computing
P
max
=
max
i
{
P
CH
(
i
)
}
,
where the max{ } function computes the maximum value of P CH (i);
computing
P
min
=
min
i
{
P
CH
(
i
)
}
,
where the min{ } function computes the minimum value of P CH (i);
setting a first frequency selectivity indicator (FSI) value if P max ≧T*P min and setting a second FSI value if P max <T*P min , wherein T is a predetermined threshold value;
determining a quantization parameter, K, which is a positive real number selected according to the first or second FSI value;
applying an exponential function to the received QAM symbol, x i for the i-th subcarrier, according to the following equation:
g
(
x
j
)
=
sgn
(
x
j
)
[
1
-
exp
(
-
x
j
K
)
]
,
sgn
(
x
j
)
=
{
1
,
x
j
>
0
0
,
x
j
=
0
-
1
,
x
j
<
0
,
where g(x i ) is a transform output of x i ;
determining a set of (2 q −1) quantization thresholds according to the first and second FSI values for a q-bit quantization output in the OFDM system, wherein q is a positive integer; and
quantizing the exponential function outputs, g(x i ) for all the sub-carriers.
2. The method of claim 1 further comprising:
quantizing the received QAM symbol x i into a q-bit output including 2 q levels according to the (2 q −1) quantization thresholds, θ 0 ,±θ 1 , . . . ,±θ 2 q−1 −1 .
3. The method of claim 2 , wherein
θ
0
=
0
,
θ
1
=
K
ln
(
2
q
-
1
2
q
-
1
-
1
)
,
and
θ
t
=
θ
1
×
S
t
,
where
S
t
=
θ
t
θ
1
=
(
q
-
1
)
ln
(
2
)
-
ln
(
2
q
-
1
-
t
)
(
q
-
1
)
ln
(
2
)
-
ln
(
2
q
-
1
-
1
)
,
t
=
2
,
…
,
2
q
-
1
-
1.
4. The method of claim 3 , wherein S t is approximated to an integer.
5. The method of claim 3 , wherein
q
=
3
,
θ
0
=
0
,
θ
1
=
K
ln
(
4
3
)
,
S
2
=
[
ln
2
ln
(
4
/
3
)
]
and
S
3
=
[
ln
4
ln
(
4
/
3
)
]
.
6. The method of claim 5 , wherein S 2 =2 and S 3 =4 and wherein the quantization thresholds θ 0 ,±θ 1 ,±θ 2 ,±θ 3 for a 3-bit quantized output are computed as follows:
θ 0 =0, θ 1 =K′×D, θ 2 =θ 1 ×2 and θ 3 =θ 1 ×4.
7. The method of claim 5 , wherein S 2 =2.5 and S 3 =5 and wherein the quantization thresholds θ 0 ,±θ 1 ,±θ 2 ,±θ 3 3-bit quantized output are computed as follows:
θ 0 =0, θ 1 =K′×D , θ 2 =θ 1 ×2.5 and θ 3 =θ 1 ×5.
8. The method of claim 1 further comprising:
computing the total channel power to determine a dynamic range
D
=
∑
i
=
0
m
-
1
P
CH
(
i
)
;
selecting K′ for the first or second FSI values; and
computing K by multiplying K′ by D, K=K′×D.
9. The method of claim 8 further comprising a look-up table to provide values of K′ based on the first and second FSI values from the channel frequency selectivity detector and sub-carrier modulation and coding scheme (MCS) information, wherein K=K′×D.
10. The method of claim 1 , wherein the first FSI value is set to one and the second FSI value is set to zero.
11. A receiver for dynamic quantization in an orthogonal frequency division multiplexing (OFDM) system, the receiver comprising:
a channel power calculator to compute a channel power P CH (i)=|G CH (i)| 2 for each of “m” sub-carrier channels, where G ch (i) is the i -th sub-carrier channel frequency response, i being an integer from 0 to m−1; and
a channel frequency selectivity detector to compute
P
max
=
max
i
{
P
CH
(
i
)
}
,
where the max{ } function computes the maximum value of P CH (i), compute
P
min
=
min
i
{
P
CH
(
i
)
}
,
where the min{ } function computes the minimum value of P CH (i), and set a first frequency selectivity indicator (FSI) value if P max ≧T*P min and a second FSI value if P max <T*P min , wherein T is a predetermined threshold value; and
a q-bit exponential quantizer comprising
processing the received QAM symbols, x i , for the i-th subcarrier, according to the following exponential equation:
g
(
x
j
)
=
sgn
(
x
j
)
[
1
-
exp
(
-
x
j
K
)
]
,
sgn
(
x
j
)
=
{
1
,
x
j
>
0
0
,
x
j
=
0
-
1
,
x
j
<
0
,
where g(x i ) is a transform output of x i , and the quantization parameter K is a positive real number;
where x is either one real value of the received complex QAM symbol x i , and the quantization parameter K is a positive real number;
determining a set of (2 q −1) quantization thresholds according to the first or second FSI value for a q-bit quantization output in the OFDM system, wherein q is a positive integer; and
quantizing the exponential function outputs, for all the subcarriers.
12. The receiver of claim 11 , wherein the first FSI value is set to 1 and the second FSI value is set to 0.
13. The receiver of claim 11 ,
wherein the quantization parameter, K, is selected according to the first or second FSI values.
14. The receiver of claim 11 comprising:
quantizing the received QAM symbol x i into a q-bit into a q-bit output including 2 q levels according to the (2 q −1) quantization thresholds, θ 0 ,±θ 1 , . . . ,±θ 2 q−1 −1 .
15. The receiver of claim 14 , wherein
θ
0
=
0
,
θ
1
=
K
ln
(
2
q
-
1
2
q
-
1
-
1
)
,
and
θ
t
=
θ
1
×
S
t
,
where
S
t
=
θ
t
θ
1
=
(
q
-
1
)
ln
(
2
)
-
ln
(
2
q
-
1
-
t
)
(
q
-
1
)
ln
(
2
)
-
ln
(
2
q
-
1
-
1
)
,
t
=
2
,
…
,
2
q
-
1
-
1.
16. The receiver of claim 15 , wherein
q
=
3
,
θ
0
=
0
,
θ
1
=
K
ln
(
4
3
)
,
S
2
=
[
ln
2
ln
(
4
/
3
)
]
and
S
3
=
[
ln
4
ln
(
4
/
3
)
]
.
17. The receiver of claim 16 , wherein S 2 =2 and S 3 =4 and wherein the quantization thresholds θ 0 ,±θ 1 ,±θ 2 ,±θ 3 for a 3-bit quantized output are computed as follows:
θ 0 =0, θ 1 =K′×D, θ 2 =θ 1 ×2 and θ 3 =θ 1 ×4.
18. The receiver of claim 16 , wherein S 2 =2.5 and S 3 =5 and wherein the quantization thresholds θ 0 ,±θ 1 ,±θ 2 ,±θ 3 for a 3-bit quantized output are computed as follows:
θ 0 =0, θ 1 =K′×D , θ 2 =θ 1 ×2.5 and θ 3 =θ 1 ×5.
19. The receiver of claim 15 , wherein S t is approximated to an integer.
20. The receiver of claim 11 further comprising:
computing the total channel power to determine a dynamic range
D
=
∑
i
=
0
m
-
1
P
CH
(
i
)
;
selecting K′ for the first or second FSI value; and
computing K by multiplying K′ by D, K=K×D.
21. The receiver of claim 20 further comprising a look-up table to provide values of K′ based on the first and second FSI values from the channel frequency selectivity detector and sub-carrier modulation and coding scheme (MCS) information, wherein K=K′×D.