Power amplifying apparatus and method using pre-distortion and radio communication system including the same apparatus
A new concept of power amplifying apparatus and method is developed for adopting pre-distortion and applying to a radio communication system. The power amplifying apparatus includes a pre-distortion module for pre-distorting an input signal x(n) according to a pre-distortion function, and for outputting a pre-distorted signal z(n), a power amplifier for amplifying a signal obtained by digital-to-analog conversion of the pre-distorted signal z(n), a pre-distortion function producing module for receiving the pre-distorted signal z(n) and a signal y(n) obtained by analog-to-digital conversion of the fed-back output signal of the power amplifier, and for obtaining model coefficients c q,k,s that minimize the power of a modeling error signal e(n)=z(n)−y(n). The pre-distortion function used by the pre-distortion module is represented as z ( n ) = ∑ q = 0 Q ∑ k = 1 K ∑ s = - S min ( S , q ) c q , k , s y ( n - q + s ) y ( n - q ) k - 1 .
1. A power amplifying apparatus using pre-distortion, comprising:
a pre-distortion module for pre-distorting an input signal x(n) according to a pre-distortion function, and for outputting a predistorted signal z(n),
a power amplifier for amplifying a signal obtained by digital-to-analog conversion of the pre-distorted signal z(n), and
a pre-distortion function producing module for receiving the pre-distorted signal z(n) and a signal y(n) obtained by analog-to-digital conversion of the fed-back output signal of the power amplifier, and for obtaining model coefficents c q,k,s that minimize the power of a modeling error signal e(n)=z(n)−y(n), said pre-distortion function used by the pre-distortion module represented as,
z
(
n
)
=
∑
q
=
0
Q
∑
k
=
1
K
∑
s
=
-
S
min
(
S
,
q
)
c
q
,
k
,
s
y
(
n
-
q
+
s
)
y
(
n
-
q
)
k
-
1
wherein, Q is a maximum number of memory depth in an actual implementation, K is a maximum order of nonlinear polynomial to be implemented in the pre-distortion system, S is a maximum memory depth of cross-memory terms to be considered in the pre-distortion implementation, q is an indexing variable for summation operation over memory from 0 up to Q, k is an indexing variable for summation operation over polynomial order from 1 up to K, s is an indexing variable for summation operation over cross-memory depth from −S up to min(S, q) and c q,k,s is a model coefficient corresponding to the specific values of q, k and s.
2. The power amplifying apparatus as claimed in claim 1 , wherein the pre-distortion module pre-distorts the input signal according to the pre-distortion function in such a manner that values of
∑
k
=
1
K
c
q
,
k
,
s
x
(
n
-
q
)
k
-
1
are previously calculated for all |x(n−q)| for q and s to form look-up tables (LUT), wherein delay chains store current and previous signal values upon the input of the input signal x(n), and other delay chains store the absolute values of the signal values, wherein LUT values corresponding to x(n−q+s) for q, k and s are found and multiplied by the signal values and absolute values, and wherein the multiplied values are summed by an adder that canies out a calculation corresponding to
∑
q
=
0
Q
∑
s
=
-
S
min
(
S
,
q
)
.
3. A radio communication system including a power amplifying apparatus using pre-distortion, the power amplifying apparatus comprises:
a pre-distortion module for pre-distorting an input signal x(n) according to a pre-distortion function, and for outputting a pre-distorted signal z(n),
a power amplifier for amplifying a signal obtained by digital-to-analog conversion of the pre-distorted signal z(n), and
a pre-distortion function producing module for receiving the pre-distorted signal z(n) and a signal y(n) obtained by analog-to-digital conversion of the fed-back output signal of the power amplifier, and for obtaining model coefficients c q,k,s that minimize the power of a modeling error signal e(n)=z(n)−y(n), said pre-distortion function used by the pre-distortion module represented as,
z
(
n
)
=
∑
q
=
0
Q
∑
k
=
1
K
∑
s
=
-
S
min
(
S
,
q
)
c
q
,
k
,
s
y
(
n
-
q
+
s
)
y
(
n
-
q
)
k
-
1
wherein, Q is a maximum number of memory depth in an actual implementation, K is a maximum order of nonlinear polynomial to be implemented in the pre-distortion system, S is a maximum memory depth of cross-memory terms to be considered in the pre-distortion implementation. q is an indexing variable for summation operation over memory from 0 up to Q, k is an indexing variable for summation operation over polynomial order from 1 up to K, s is an indexing variable for summation operation over cross-memory depth from −S up to min(S, q) and c q,k,s is a model coefficient corresponding to the specific values of q, k and s.
4. The radio communication system as claimed in claim 3 , wherein the pre-distortion module pre-distorts the input signal according to the pre-distortion function in such a manner that values of
∑
k
=
1
K
c
q
,
k
,
s
x
(
n
-
q
)
k
-
1
are previously calculated for all |x(n−q)| to form look-up tables (LUT), wherein delay chlains store current and previous signal values upon the input of the input signal x(n), and other delay chains store the absolute values of the signal values, wherein LUT values corresponding to x(n−q+s) for q, k and s are found and multiplied by the signal values and absolute values, and wherein the multiplied values are summed by an adder that carries out a calculation corresponding
∑
q
=
0
Q
∑
s
=
-
S
min
(
S
,
q
)
.
5. A power amplifying method using pre-distortion, comprising:
a pre-distortion step of pre-distorting an input signal x(n) according to a pre-distortion function, and of outputting a pre-distorted signal z(n),
a power amplification step of amplifying a signal obtained by digital-to-analog conversion of the pre-distorted signal z(n), and
a pre-distortion function producing step of receiving the pre-distorted signal z(n) and a signal y(n) obtained by analog-to-digital conversion of the fed-back output signal of the power amplifier, and of obtaining model coefficients c q,k,s that minimize the power of a modeling error signal e(n)=z(n)−y(n), said pre-distortion function used in the pre-distortion step represented as,
z
(
n
)
=
∑
q
=
0
Q
∑
k
=
1
K
∑
s
=
-
S
min
(
S
,
q
)
c
q
,
k
,
s
y
(
n
-
q
+
s
)
y
(
n
-
q
)
k
-
1
wherein, Q is a maximum number of memory depth in an actual implementation, K is a maximum order of nonlinear polynomial to be implemented in the pre-distortion system, S is a maximum memory depth of cross-memory terms to be considered in the pre-distortion implementation, q is an indexing variable for summation operation over memory from 0 up to Q, k is an indexing variable for summation operation over polynomial order from 1 up to K, s is an indexing variable for summation operation over cross-memory depth from −S up to min(S, q) and c q,k,s is a model coefficient corresponding to the specific values of q, k and s.
6. The power amplifying method as claimed in claim 5 , wherein the input signal is pre-distorted according to the pre-distortion function in such a manner that values of
∑
k
=
1
K
c
q
,
k
,
s
x
(
n
-
q
)
k
-
1
are previously calculated for all |x(n−q)| for q and s to form look-up tables (LUT), wherein delay chains store current and previous signal values upon the input of the input signal x(n), and other delay chains store the absolute values of the signal values, wherein LUT values corresponding to x(n−q+s) for q, k and s are found and multiplied by the signal values and absolute values, and wherein the multiplied values are summed by an adder that carries out a calculation corresponding to
∑
q
=
0
Q
∑
s
=
-
S
min
(
S
,
q
)
.