Linearization of a non-linear electronic device
There is provided mechanisms for enabling linearization of a non-linear electronic device. A method is performed by a linearizer device. The method comprises receiving an input signal destined to be input to the non-linear electronic device. Input-output characteristics of the non-linear electronic device is in the linearizer device represented by a linearization function defined by a LUT based model of base functions. The linearizer device is configured to in a greedy pursuit framework select the base functions according to a signal reconstruction criterion. The method comprises obtaining an output signal by subjecting the input signal to the linearization function. The method comprises providing the output signal, instead of the input signal, as input to the non-linear electronic device, thereby enabling linearization of the non-linear electronic device.
1 . A method for enabling linearization of a non-linear electronic device, the method being performed by a linearizer device, the method comprising:
receiving an input signal (x) destined to be input to the non-linear electronic device;
wherein input-output characteristics of the non-linear electronic device is in the linearizer device represented by a linearization function defined by a LUT based model of base functions, and wherein the linearizer device is configured to in a greedy pursuit framework select the base functions according to a signal reconstruction criterion,
obtaining an output signal (y) by subjecting the input signal (x) to the linearization function; and
providing the output signal (y), instead of the input signal (x), as input to the non-linear electronic device, thereby enabling linearization of the non-linear electronic device.
2 . The method according to claim 1 , wherein each base function represents a memory tap pair.
3 . The method according to claim 2 , wherein each memory tap pair consists of one data delay value m and one address delay value l.
4 . The method according to claim 1 , wherein the signal reconstruction criterion pertains to selecting base functions that minimize a mean squared error or maximize a bin value based function.
5 . The method according to claim 1 , wherein the base functions form a set of base functions S (i) , and wherein according to the greedy pursuit framework the set of base functions S (i) is iteratively determined.
6 . The method according to claim 5 , wherein the set of base functions S (i) is determined from a candidate set {tilde over (S)} (i) of base functions.
7 . The method according to claim 6 , wherein per each iteration i, the candidate set {tilde over (S)} (i) of base functions is formed by adding an intermediate set T to the set of base functions S (i-1) selected at iteration i−1.
8 . The method according to claim 7 , wherein the intermediate set T comprises T base functions corresponding to smallest differences between a residual error r (i-1) and an estimate of the residual error for iteration i.
9 . The method according to claim 8 , wherein the candidate set {tilde over (S)} (i) for iteration i is formed as:
S
˜
(
i
)
=
S
(
i
-
1
)
⋃
{
arg
min
(
m
,
l
)
T
(
r
(
i
-
1
)
-
x
(
n
-
m
)
∘
f
m
,
l
(
❘
"\[LeftBracketingBar]"
x
(
n
-
l
)
❘
"\[RightBracketingBar]"
)
2
)
}
,
where ∘ denotes Hadamard product, |⋅| 2 denotes 2 vector norm, ∪ denotes set unit operator, and f m,l denotes the LUT with data delay m and address delay l.
10 . The method according to claim 6 , wherein per each iteration i, the set of base functions S (i) corresponding to smallest difference between a desired output signal y and an estimate of the desired output signal is selected from the candidate set {tilde over (S)} (i) .
11 . The method according to claim 10 , wherein one estimate of the desired output signal is formed for each of all possible combinations of the base functions in the candidate set {tilde over (S)} (i) per each iteration i.
12 . The method according to claim 11 , wherein the set of base functions S (i) , for all S-element subsets L⊆{tilde over (S)} (i) , for iteration i is selected as:
S
(
i
)
=
arg
min
L
⊆
S
~
(
i
)
1
(
y
-
∑
(
m
,
l
)
∈
L
x
(
n
-
m
)
∘
f
m
,
l
(
❘
"\[LeftBracketingBar]"
x
(
n
-
l
)
❘
"\[RightBracketingBar]"
)
2
)
,
where ∘ denotes Hadamard product, |⋅| 2 denotes 2 vector norm, and f m,l denotes the LUT with data delay m and address delay l.
13 . The method according to claim 10 , wherein per each iteration i, the residual error r (i) is updated as the difference between the desired output signal y and the estimate ŷ (i) of the desired output signal corresponding to the selected set of base functions S (i) at iteration i.
14 . The method according to claim 8 , wherein the residual error r (i) for iteration i is updated as:
r
(
i
)
=
y
-
y
^
(
i
)
.
15 . The method according to claim 13 , wherein the estimate ŷ (i) of the desired output signal for iteration i is determined as:
y
^
(
i
)
=
∑
(
m
,
l
)
∈
S
(
i
)
x
(
n
-
m
)
∘
f
m
,
l
(
❘
"\[LeftBracketingBar]"
x
(
n
-
l
)
❘
"\[RightBracketingBar]"
)
.
16 . A linearizer device for enabling linearization of a non-linear electronic device, the linearizer device comprising processing circuitry, the processing circuitry being configured to cause the linearizer device to:
receive an input signal (x) destined to be input to the non-linear electronic device;
wherein input-output characteristics of the non-linear electronic device is in the linearizer device represented by a linearization function defined by a LUT based model of base functions, and wherein the linearizer device is configured to in a greedy pursuit framework select the base functions according to a signal reconstruction criterion,
obtain an output signal (y) by subjecting the input signal (x) to the linearization function; and
provide the output signal (y), instead of the input signal (x), as input to the non-linear electronic device, thereby enabling linearization of the non-linear electronic device.
17 . A linearizer device for enabling linearization of a non-linear electronic device, the linearizer device comprising:
a receive module configured to receive an input signal (x) destined to be input to the non-linear electronic device;
wherein input-output characteristics of the non-linear electronic device is in the linearizer device represented by a linearization function defined by a LUT based model of base functions, and wherein the linearizer device is configured to in a greedy pursuit framework select the base functions according to a signal reconstruction criterion,
an obtain module configured to obtain an output signal (y) by subjecting the input signal (x) to the linearization function; and
a provide module configured to provide the output signal (y), instead of the input signal (x), as input to the non-linear electronic device, thereby enabling linearization of the non-linear electronic device.
18 . A computer program for enabling linearization of a non-linear electronic device, the computer program comprising computer code which, when run on processing circuitry of a linearizer device, causes the linearizer device to:
receive an input signal (x) destined to be input to the non-linear electronic device;
wherein input-output characteristics of the non-linear electronic device is in the linearizer device represented by a linearization function defined by a LUT based model of base functions, and wherein the linearizer device is configured to in a greedy pursuit framework select the base functions according to a signal reconstruction criterion,
obtain an output signal (y) by subjecting the input signal (x) to the linearization function; and
provide the output signal (y), instead of the input signal (x), as input to the non-linear electronic device, thereby enabling linearization of the non-linear electronic device.