KR100562666B1 - Apparatus and Method for Linearization of Nonlinear Transmitting System - Google Patents

Abstract

The present invention relates to a linearization apparatus of a nonlinear transmission system, and in particular, when a system operating point is located between active operating points of multiple LUTs provided for each system operating point, an actual system is interpolated between compensation values indexed in the multiple LUTs. By calculating the compensation value corresponding to the operating point to pre-distort the system input signal, the nonlinear transmission system can be efficiently linearized using a small number of LUTs.

Furthermore, by filtering the input signal of the predistorted system through the frequency correction filter, it is possible to linearize the nonlinear characteristics of the nonlinear transmission system according to the frequency change.

Description

Apparatus and Method for Linearization of Nonlinear Transmitting System             

1 is a block diagram showing a linearization device of a conventional nonlinear transmission system.

2 is a view showing the configuration of a linearization device according to an embodiment of the present invention.

3 is a diagram showing the configuration of interpolation means using a spline technique.

4 is a conceptual diagram of interpolation of the spline technique.

5 is a diagram illustrating a configuration of interpolation means using linear interpolation.

FIG. 6 is a diagram showing the structure of a weight applicator in FIG. 5; FIG.

7 is a flowchart illustrating a linearization procedure of a nonlinear transmission system according to an embodiment of the present invention.

8 is a flowchart illustrating a compensation value interpolation procedure using a spline technique.

9 is a flowchart illustrating a compensation value interpolation procedure using the linear interpolation method.

10 is a flowchart illustrating a LUT update procedure.

Explanation of symbols on the main parts of the drawings

100: linearization device 110: OPC (Operation Point Calculator)

120: LUT selector 130: Multi-Look Up Table

140: Indexer 150: Interpolating Means

151: Scaler 152: Interpolater

153: weight calculator 154: weight applicator

155: first multiplier 156: second multiplier

157: adder 161: complex multiplier

162: Spectrum Flattening Filter (SFF)

163: Digital to Analog Converter (DAC)

164: UC (Up Converter) 165: DC (Down Converter)

166: ADC (Analog to Digital Converter)

167: SCM (Signal Capture Module) 168: DSP (Digital Signal Processor)

200: nonlinear transmission system

The present invention relates to a linearization apparatus of a nonlinear transmission system, and in particular, when a system operating point is located between active operating points of multiple LUTs provided for each system operating point, an actual system is interpolated between compensation values indexed in the multiple LUTs. A method of pre-distorting a system input signal by calculating a compensation value corresponding to an operating point.

Conventional linearization apparatus using a multiple LUT, as shown in Figure 1 attached to the operation point calculator (OPC) 11, LUT selector (12), multi-LUT (Multi-LUT) 13, indexer ( Indexer (14), Complex Multiplier (15), Digital to Analog Converter (DAC) (16), Up Converter (UC) (17), Down Converter (18), Analog to Digital Converter (ADC) (19 ), And a signal capture module (SCM) 20 and a digital signal processor (DSP) 21 to compensate and linearize the nonlinear characteristics of the nonlinear transmission system 25.

The OPC 11 receives a weighted moving average by receiving an operation point information signal S [n] including information on an operation point for identifying an operation area of the nonlinear transmission system 25. The calculated operating point L [n] is calculated using the method.

The operating point information signal S [n] can be calculated from the output of the nonlinear transmission system 25 through an operating point measuring sensor (not shown in the figure), and is predistorted in consideration of the influence on noise. It may also calculate from the signal V _d [n].

The LUT selector 12 selects the LUT to be activated in the multiple LUTs 13 using the operating point L [n] output from the OPC 11. That is, the corresponding LUT is selected and activated by comparing the OPC 11 output operating point L [n] with the active operating points L ₁ , L ₂ ... L _M-1 of the previously stored LUTs. Let's do it.

The active operating points (L ₁ , L ₂ ... L _M-1 ) of each LUT are values that are set and stored in the first generation of each LUT of the multiple LUTs 13, and are used when generating LUTs. By using the averaging with the adjacent operating point can be set to the active operating point of the LUT, these active operating points are the boundary value between the LUT.

For example, when the operating point L [n] of the nonlinear transmission system 25 is the operating power P ₁ , P ₂ , P ₃ ... P _M-1 , P _M , the active operating point of each LUT ( L ₁ , L ₂ ... L _M-1 ) is (P ₁ + P ₂ ) / 2, (P ₂ + P ₃ ) / 2, (P ₃ + P ₄ ) / 2, ... (P _{M -1} + P _M ) / 2

The multiple LUT 13 is composed of a number of LUTs corresponding to the operating area of the nonlinear transmission system 25, each LUT is composed of N entries, and each entry has an input signal (V _i [n]). The compensation value is stored for each size.

Indexer 14 indexes the compensation value in the operating region from the entries of a specific LUT in the LUT activated by selector 12 in accordance with the size (Magnitude) to the input signal (V _i [n]).

The complex multiplier 15 multiplies the input signal _Vi [n] by the compensation value output from the specific LUT to generate a predistorted compensation signal V _d [n].

The DAC 16 converts the compensation signal V _d [n] into an analog signal, the UC 17 pulls up the input analog signal to a predetermined frequency band, and the DC 18 is a nonlinear transmission system 25. The output of the feedback is fed back down to the original frequency band, ADC 19 converts the input analog signal into a digital signal.

The SCM 20 stores the compensation signal V _d [n] generated by the complex multiplier 15 and the feedback signal V _f [n] converted into a digital signal through the ADC 19 and the DSP 21 ) Is a signal processing process of the compensation signal (V _d [n]) and feedback signal (V _f [n]) stored in the SCM 20 for a predetermined time, that is, from the input and output data of the non-linear transmission system 25 from the system 25 Update the corresponding LUT by estimating the nonlinear characteristic of

The updated LUT is an LUT that was activated at the time of data storage in the SCM 20 among the multiple LUTs 13. In this case, it is preferable not to update the LUT when the active LUT is changed while storing data in the SCM 20.

As described above, the conventional linearization apparatus linearizes the nonlinear characteristics of each non-linear transmission system by using the LUTs corresponding to the corresponding operating regions by compensating the nonlinear characteristics.

However, since multiple LUTs are generated by stepwise dividing successive operating regions of the nonlinear transmission system, operating points between the active operating points of the LUTs representing the divided operating regions are not compensated for.

Accordingly, in order to compensate for the operating points between the active operating points of each LUT, LUTs representing the operating points should be generated, which requires a large memory.

 In addition, although the nonlinear characteristics of the nonlinear transmission system may be caused by a frequency change in addition to the change in the operating region of the system, the conventional linearization apparatus has a problem that does not consider the nonlinear characteristics according to the frequency change.

SUMMARY OF THE INVENTION The present invention has been made to solve the above-described problems, and its object is to compensate for the case where the operating point of the nonlinear transmission system is the value between the active operating points of each LUT by using interpolation, thereby providing a compensation value of the multi-phased LUT. It is to reduce the error that occurs during indexing.

Furthermore, another object of the present invention is to compensate for the nonlinear characteristic of each frequency of the nonlinear transmission system by using the frequency correction filter.

A feature of the present invention for achieving the above object is composed of look up tables (LUTs) having an active operating point of a nonlinear transmission system, each of which consists of a plurality of entries and an input signal to each entry. A linearization apparatus of a non-linear transmission system having a multi-LUT for storing the compensation values of the nonlinear transmission system, by linearly distorting the nonlinear transmission system input signal by indexing the compensation value in the multiple LUT for each operating point of the nonlinear transmission system. From among the active operating points, the first active operating point of the first LUT smaller than the nonlinear transmission system operating point with the nonlinear transmission system operating point as a value between and the second active operating point of the second LUT larger than the nonlinear transmission system operating point are selected. A LUT selector configured to select and activate a predetermined number of LUTs including the first LUT and the second LUT; Interpolating means for calculating a compensation value corresponding to the nonlinear transmission system operating point by interpolating between compensation values indexed in the activated LUT using the first active operating point information and the second active operating point information. It is to provide a linearization device of a nonlinear transmission system characterized by.

Advantageously, said interpolation means comprises: a scaler for calculating an interpolation position of said nonlinear transmission system operating point using said first active operating point and second active operating point information; And an interpolator configured to calculate a compensation value corresponding to the interpolation position by interpolating between the indexed compensation values.

Alternatively, the interpolation means includes a first weight corresponding to the distance between the nonlinear transmission system operating point and the first active operating point and a second weighting value corresponding to the distance between the nonlinear transmission system operating point and the second active operating point. A weight calculator for calculating; And a weight applicator configured to calculate the compensation value by applying the first weight value and the second weight value to compensation values indexed in the first and second LUTs.

Advantageously, the weighting unit comprises: a first multiplier that multiplies the first weight value by a second compensation value indexed in the second LUT; A second multiplier for multiplying the second weight value with a first compensation value indexed by the first LUT; And an adder configured to calculate the compensation value by adding a second compensation value multiplied by the first weight value and a first compensation value multiplied by the second weight value.

More preferably, the linearization apparatus of the nonlinear transmission system according to the present invention uses the inverse function of the frequency-dependent gain change of the output of the nonlinear transmission system to compensate for the gain variation of the nonlinear transmission system input signal. And filtering means for filtering the input signal pre-distorted by.

Another feature of the present invention is a multiple LUT consisting of look up tables (LUTs) having an active operating point of a nonlinear transmission system, each consisting of a plurality of entries and storing the compensation values of the input signal in each entry. A linearization method of a nonlinear transmission system for linearly distorting the nonlinear transmission system input signal by indexing a compensation value in the multiple LUT for each operating point of a nonlinear transmission system, the nonlinear transmission among each active operating points of the multiple LUT. Selecting an active operating point with the system operating point as an interstitial value; Activating a predetermined number of LUTs required for interpolation of a compensation value corresponding to the nonlinear transmission system operating point; Calculating a compensation value corresponding to the nonlinear transmission system operating point by interpolating between compensation values indexed in the activated LUT using the selected active operating point; A method for linearizing a nonlinear transmission system comprising predistorting the nonlinear transmission system input signal using the calculated compensation value is provided.

Advantageously, calculating said compensation value comprises: calculating an interpolation position of said nonlinear transmission system operating point according to a distance between said nonlinear transmission system operating point and said selected active operating point; And interpolating between the indexed compensation values to calculate a compensation value corresponding to the interpolated position.

Also preferably, the calculating of the compensation value may include calculating a first weight value corresponding to a distance between the nonlinear transmission system operating point and the selected active operating point and a second weight value greater than the first weight value; And calculating the compensation value by applying the first weight value and the second weight value to compensation values indexed from the LUT having the selected active operating point.

Advantageously, applying said compensation value comprises: multiplying said first weight by a compensation value indexed at an LUT having an active operating point greater than said nonlinear transmission system operating point; Multiplying the second weight by a compensation value indexed at an LUT having an active operating point less than the nonlinear transmission system operating point; And adding a compensation value multiplied by the first weight value and a compensation value multiplied by the second weight value.

More preferably, the linearization method of the nonlinear transmission system according to the present invention uses an inverse Fourier transform (IFFT) on an inverse function spectrum for a spectrum representing a change in gain of each nonlinear transmission system as an initial filter coefficient and a new nonlinear transmission. And filtering the predistorted input signal to compensate for the frequency-specific gain change of the nonlinear transmission system input signal using filtering means for updating the coefficient for each system input signal.

delete

Also preferably, using the linearly distorted input signal and the output signal of the non-linear transmission system to determine the non-linear characteristics of the non-linear transmission system; And updating an LUT having an active operating point closer to the nonlinear transmission system operating point among the selected active operating points to indicate an inverse characteristic of the nonlinear characteristic.

Advantageously, updating the LUT comprises: comparing a distance between the non-linear transmission system operating point and the selected active operating points; Selecting a LUT with a shorter distance between the operating points; And updating the LUT when the operating point is a value within an update area between a minimum threshold value and a maximum threshold value preset to an active operating point of the selected LUT.

Hereinafter, exemplary embodiments of the present invention will be described in detail with reference to the accompanying drawings.

2 is a diagram illustrating a configuration of a linearization apparatus according to an embodiment of the present invention.

Referring to FIG. 2, the linearization apparatus of the present invention includes an operation point calculator (OPC) 110, an LUT selector 120, a multi-LUT 130, an indexer 140, Interpolating Means (150), Complex Multiplier (161), Spectrum Flattening Filter (SFF) (162), Digital to Analog Converter (DAC) 163, Up Converter (UC) (164), Down Converter (DC) (165), Analog to Digital Converter (ADC) 166, Signal Capture Module (SCM) 167, and Digital Signal Processor (DSP) 168, which are referred to as nonlinear transmission systems (hereinafter referred to only as 'systems'). The nonlinear nature of 200 is compensated for and linearized.

The OPC 110 receives the operating point information signal S [n] and calculates the operating point L [n] of the system 200 of the corresponding sampling period by using a weighted moving average method.

The LUT selector 120 stores the system 200 operating point L [n] calculated by the OPC 110 in advance and stores the active operating points L ₁ ,..., L _P , L _{P + 1,.} .. _{L M} ) and transmits the active operating point (L _P , L _{P + 1} ) information of each LUT whose corresponding operating point (L [n]) is located between the interpolation means 150 and the corresponding A predetermined number of LUTs required for interpolation of the compensation value g 'corresponding to the operating point L [n] are selected and activated.

That is, in the case of interpolation of the spline technique, from (Pa) th LUT to (P + b) th LUT satisfying L _Pa ≤ L _P ≤ L [n] ≤ L _{P + 1} ≤ L _{P + b} 2 LUTs (LUT P, L) having active operating points L _P and L _{P + 1} , activating + b + 1) LUTs and having a corresponding operating point L [n] for linear interpolation. LUT P + 1) is activated.

On the other hand, the LUT selector 120 is the first and second LUT (LUT 1, when the operating point (L [n]) of the system 200 is smaller than the active operating point (L ₁ ) of the first LUT (LUT 1). LUT 2) is activated and the corresponding active operating point (L ₁ , L ₂ ) information is transmitted to the interpolation means 150, and the M-th LUT (LUT) whose operating point (L [n]) of the system 200 is the last LUT. If M is greater than the active operating point L _M , the M-1th and Mth LUTs (LUT M-1, LUT M) are activated and the corresponding active operating points L _M-1 and L _M are interpolated. Forward to 150.

The multiple LUT 130 has a corresponding operating point for each operating point of the nonlinear transmission system 200 as an active operating point L ₁ ,... L _P , L _{P + 1} , ... L _M LUTs (LUT 1, ... LUT P, LUT P + 1, ... LUT M), each LUT consists of N entries, and each entry has an input signal (V _i [n ]) The compensation value is stored for each size.

Indexer 140 may be distorted line an input signal (V _i [n]), the size (Magnitude) an input signal (V _i [n]) from the particular entry of the predetermined number of LUT activated by the LUT selector 120 in accordance with the Index rewards.

The interpolation means 150 is activated at the LUT activated through a predetermined interpolation method using the operating point L [n] of the system 200 and the active operating point L _P , L _{P + 1} transmitted from the LUT selector 120. The compensation value g 'corresponding to the operating point L [n] of the system 200 is calculated by interpolating between the compensation values indexed.

The complex multiplier 161 multiplies the input signal _Vi [n] by the interpolated compensation value g ′ to generate a predistorted compensation signal V _d [n].

The SFF 162 is a filter that maintains a constant frequency change of the system 200. The SFF 162 receives a tone signal for each frequency in a band from the system 200 and represents a nonlinear characteristic of the complex gain value for each frequency. The spectrum of the filter is obtained by taking an inverse function of the spectrum of the system 200 so as to obtain a spectrum and exhibit the inverse of the nonlinear characteristic. Then, the filter spectrum is inverse Fourier transformed (IFFT) and used as an initial filter coefficient, and each time the compensation signal V _d [n] is inputted, the filter coefficient is updated in an adaptive manner for each frequency of the system 200. Compensates for nonlinear characteristics.

The DAC 163 converts the filtered compensation signal V ' _d [n] into an analog signal, the UC 164 pulls the input analog signal up to a predetermined frequency band, and the DC 165 a nonlinear transmission system. The output of 200 is fed back and down to the original frequency band, and the ADC 166 converts the input analog signal into a digital signal.

The SCM 167 stores the filtered compensation signal V ' _d [n] and the feedback signal V _f [n] converted into a digital signal, and the DSP 168 processes the SCM for a predetermined time through a signal processing process. The nonlinear characteristic of the system 200 is estimated from the compensation signal V ′ _d [n] and the feedback signal V _f [n] stored in 167, that is, the input / output data of the system 200, and the corresponding LUT is updated. .

The updated LUT updates the LUT having an active operating point closest to the operating point L [n] of the system 200 at the time of signal storage in the SCM 167 among the multiple LUTs 130. However, it is not desirable to update the LUT when the system 200 operating point L [n] is at the boundary between the active operating points, so the minimum threshold value L _P -G _low and the highest threshold for the active operating point are not desirable. Active operation only if the operating point L [n] is within the update area (L _P -G _low , L _P + G _high ) by setting the guard band between the values (L _P + G _high ). Update the LUT (LUT P) with the point L _P.

3 is a diagram illustrating an internal configuration of an interpolation means 150 using a spline technique, and FIG. 4 is a conceptual diagram of interpolation of a spline technique.

Referring to FIG. 3, the interpolation means 150 includes a scaler 151 and an interpolator 152.

The scaler 151 uses the operating point L [n] of the system 200 and the active operating point L _P and L _{P + 1} received from the LUT selector 120, through the following equation. As shown in (a) of the above, the interpolation position P 'of the operating point L [n] between the sampling point P and P + 1 as the corresponding LUT (LUT P, LUT P + 1) indication value is determined. Calculate.

P '= P + (L [n] -L _P ) / (L _{P + 1} -L _P )

Through the above equation, the interpolation position P 'of the operating point L [n] is represented by the number of decimal units between P and P + 1, which are sampling points.

The interpolator 152 constantly interpolates between the compensated compensation values g _Pa , ... g _P , g _{P + 1} , ... g _{P + b} as shown in (b) of FIG. 4. To calculate the compensation value g 'corresponding to the interpolation position P' of the operating point L [n].

5 is a diagram showing an internal configuration of interpolation means 150 using linear interpolation.

Referring to FIG. 5, the interpolation means 150 is composed of a weight calculator 153 and a weighted adder 154.

Weight calculating unit 153 is the system 200 the operating point (L [n]), and LUT selector that receives the active working point the operation by using the (L _P, L _{P + 1)} information transmitted from the (120) (L [ n]) the first compensation value, which is the compensation value g _P indexed in the LUT (LUT P) having an active operating point L _P smaller than and the active operating point L larger than the operating point L [n]. _{P + 1)} is the first of the first weighting the inner product is a weight on the second compensation value (w ₁₎ and a second weight (w ₂₎ a compensation value (g _{P + 1)} to be indexed in LUT (LUT P + 1) with It is calculated through the following equation.

w ₁ = (L [n] -L _P ) / (L _{P + 1} -L _P )

w ₂ = (L _{P + 1} -L [n]) / (L _{P + 1} -L _P )

The weight applicator 154 compensates the first weight w ₁ for a second weight to apply a larger weight to the compensation value indexed from the LUT having an active operating point that is closer to the operating point L [n]. The value g _{P + 1} is applied and the second weight value w ₂ is applied to the first compensation value g _P.

6 shows the structure of the weight applicator 154.

Referring to FIG. 6, the weight applicator 154 is composed of a first multiplier 155, a second multiplier 156, and an adder 157 and a first weight value w ₁ through a first multiplier 155. The second compensation value g _{P + 1} is multiplied, multiplied by the second weight value w ₂ and the first compensation value g _P through the second multiplier 156, and then multiplied by the adder 157. Two values are added and output as the interpolated compensation value g '.

Meanwhile, the interpolation means 150 may be implemented as a digital filter executed by a superposition operation using a sinc function, which is a function of inverse Fourier transform of a delta function of Dirac, in addition to the above-described spline technique or linear interpolation technique. have.

7 is a flowchart illustrating a linearization procedure of a nonlinear transmission system according to an embodiment of the present invention.

Referring to FIG. 7, the OPC 110 receives operating point information S [n] of the system 200 and calculates an operating point L [n] (S710).

Then, the LUT selector 120 stores the calculated operating point L [n] in each of the active operating points L ₁ ,... L _P , L _{P + 1} ,... In comparison with .L _M ), two active operating points L _P and L _{P + 1} having the operating point L [n] as a value between them are selected and transmitted to the interpolation means 150 (S720). Then, the multi-LUT 130 selects and activates a predetermined number of LUTs necessary for interpolation of the compensation value g 'corresponding to the operating point L [n] (S730).

Accordingly, the interpolation means 150 activates the activation through a predetermined interpolation method using the operating point L [n] and two active operating points L _P and L _{P + 1} transmitted from the LUT selector 120. The compensation value g 'corresponding to the operating point L [n] is interpolated from the compensation values indexed in the LUTs (S740).

Then, the complex multiplier 161 multiplies the calculated compensation value g 'by the system 200 input signal _Vi [n] to generate a predistorted compensation signal V _d [n] (S750). In operation S760, the SFF 162 filters the compensation signal V _d [n] to maintain a constant gain change of the system 200 according to the frequency.

The DAC 163 converts the filtered compensation signal V ′ _d [n] into an analog signal, and the UC 164 pulls the analog signal up to a predetermined frequency band and inputs it to the system 200. (S770).

Thereafter, the output of the system 200 is fed back to the linearizer 100 (S780), DC 165 down the system 200 output to the original frequency band, ADC 166 is the frequency band The down-converted analog signal is converted into a digital signal, and the SCM 167 stores the feedback signal V _f [n] converted into the digital signal together with the compensation signal V ′ _d [n] (S780). .

Then, the DSP 168 processes the signals stored in the SCM 167 for a predetermined time to grasp the input / output characteristics of the system 200 so as to display an inverse characteristic with respect to the non-linear characteristics of the system 200. The LUT having the active operating point closest to the operating point L [n] of the system 200 at the time when the signals are stored in 167 is updated (S790).

8 is a flowchart illustrating the compensation value interpolation procedure S740 using the spline technique. Referring to FIG. 8, the scaler 151 in the interpolation means 150 is the operating point L [n] of the system 200 calculated by the OPC 110 and the active operating point L transmitted from the LUT selector 120. _P , L _{P + 1} ) to calculate the interpolation position P 'of the operating point L [n] through Equation 1 (S741).

Then, the interpolator 152 is activated by the LUT selector 120 according to the interpolation position P 'and compensated for by the indexer 140 (g _Pa , ... g _P , g _{P +). 1} , ... g _{P + b} ) is constantly interpolated (S742), and a compensation value g ′ corresponding to the interpolation position P ′ is calculated (S743).

9 is a flowchart illustrating the compensation value interpolation procedure S740 using linear interpolation. Referring to FIG. 9, the weight calculator 153 in the interpolation means 150 is an active operating point delivered from the system 200 operating point L [n] calculated by the OPC 110 and the LUT selector 120. using (L _P, L _{P + 1)} and calculates the information system 200, between the operating point (L [n]) and each active operating point (L _P, L _{P + 1)} the distance (S745).

Then, the weight (w ₁ , w ₂ ) corresponding to the calculated distance is calculated through Equation 2 (S746), and the calculated weight (w ₁ , w ₂ ) by the LUT selector 120 The compensation value g 'corresponding to the operating point L [n] is calculated by applying the two compensation values g _P and g _{P + 1} that are activated and indexed by the indexer 140 (S747). .

10 is a flowchart illustrating the LUT update procedure (S790).

Referring to FIG. 10, the DSP 168 pre-distorts the compensation signal V ′ _d [n] among the active operating points L _P and L _{P + 1} of two LUTs activated by the LUT selector 120. ) And an active operating point closer to the system 200 operating point L [n] calculated by the OPC 110 at the time when the feedback signal V _f [n] is stored in the SCM 167 (S791). ).

That is, the distance between the system operating point L [n], the active operating point L _P , and the active operating point L _{P + 1} is obtained, respectively, and the active operating point (eg, L [n) having a relatively closer distance is obtained. ] -L _P <L _{P + 1-} If L [n], select L _P ).

In operation S792, it is determined whether the operating point L [n] of the system 200 is a value within the update area L _P -G _low and L _P + G _high set to the selected active operating point.

Thus, when the operating point L [n] of the system 200 is out of the update area, the operating point L [n] is located at a boundary area between the active operating points L _P and L _{P + 1} . In this case, the LUT update operation is terminated, and when the value is within the update region, the LUT (LUT P) having the corresponding active operating point L _P is updated (S793).

In addition, the embodiment according to the present invention is not limited to the above description, and various alternatives, modifications, and changes can be made within the scope apparent to those skilled in the art.

As described above, the present invention calculates a compensation value corresponding to the actual system operating point by interpolating between compensation values indexed in the multiple LUT when the nonlinear transmission system operating point is located between each active operating points of the multiple LUTs. By pre-distorting the input signal, the nonlinear transmission system can be efficiently linearized using a small number of LUTs.

Furthermore, by filtering the input signal of the predistorted system through the frequency correction filter, it is possible to linearize the nonlinear characteristics of the nonlinear transmission system according to the frequency change.

Claims (14)

Non-linear transmission system, consisting of LUTs (Look Up Tables) having an active operating point of the non-linear transmission system, each of the LUT consists of a plurality of entries, each having a multiple LUT for storing the compensation values of the input signal A linearization apparatus of a nonlinear transmission system for linearly distorting an input signal of a nonlinear transmission system by indexing a compensation value in the multiple LUT for each operating point, A first active operating point of the first LUT smaller than the nonlinear transmission system operating point having the nonlinear transmission system operating point as an interstitial value among the respective active operating points of the multiple LUTs, and a second activity of the second LUT larger than the nonlinear transmission system operating point. A LUT selector for selecting an operating point and selecting and activating a predetermined number of LUTs including the first LUT and the second LUT; Interpolating means for calculating a compensation value corresponding to the nonlinear transmission system operating point by interpolating between compensation values indexed in the activated LUT using the first active operating point information and the second active operating point information. A linearization device for a nonlinear transmission system. The method of claim 1, The interpolation means includes: a scaler for calculating an interpolation position of the nonlinear transmission system operating point using the first active operating point information and the second active operating point information; And an interpolator configured to interpolate between the indexed compensation values to calculate a compensation value corresponding to the interpolated position. The method of claim 1, The interpolation means calculates a first weight corresponding to the distance between the nonlinear transmission system operating point and the first active operating point and a weighting value for calculating a second weighting value corresponding to the distance between the nonlinear transmission system operating point and the second active operating point. A calculator; And a weight applicator configured to calculate the compensation value by applying the first weight value and the second weight value to compensation values indexed in the first and second LUTs. The method of claim 3, wherein The weight applicator includes: a first multiplier that multiplies the first weight value by a second compensation value indexed by the second LUT; A second multiplier for multiplying the second weight value with a first compensation value indexed by the first LUT; And an adder configured to add the second compensation value multiplied by the first weight value and the first compensation value multiplied by the second weight value to calculate the compensation value. The method of claim 1, And filtering means for filtering the input signal pre-distorted by the compensation value using an inverse function of the gain variation per frequency of the nonlinear transmission system output to compensate for the frequency variation of the gain of the nonlinear transmission system input signal. Linearization apparatus of a non-linear transmission system, characterized in that. The method of claim 5, The filtering means uses an inverse Fourier transform (IFFT) on an inverse function value for a spectrum representing a change in gain of each nonlinear transmission system as an initial filter coefficient, and updates the coefficient for each new nonlinear transmission system input signal. Linearization apparatus of nonlinear transmission system. Non-linear transmission system, consisting of LUTs (Look Up Tables) having an active operating point of the non-linear transmission system, each of the LUT consists of a plurality of entries, each having a multiple LUT for storing the compensation values of the input signal In the linearization method of the nonlinear transmission system for linearly distorting the nonlinear transmission system input signal by indexing the compensation value in the multiple LUT for each operating point, Selecting an active operating point having a nonlinear transmission system operating point as a value among respective active operating points of the multiple LUTs; Activating a predetermined number of LUTs required for interpolation of a compensation value corresponding to the nonlinear transmission system operating point; Calculating a compensation value corresponding to the nonlinear transmission system operating point by interpolating between compensation values indexed in the activated LUT using the selected active operating point; And linearly distorting the nonlinear transmission system input signal using the calculated compensation value. The method of claim 7, wherein The calculating of the compensation value may include calculating an interpolation position of the nonlinear transmission system operating point according to a distance between the nonlinear transmission system operating point and the selected active operating point; And interpolating between the indexed compensation values to calculate a compensation value corresponding to the interpolated position. The method of claim 7, wherein The calculating of the compensation value may include: calculating a first weight value corresponding to a distance between the nonlinear transmission system operating point and the selected active operating point and a second weight value greater than the first weight value; Calculating the compensation value by applying the first weight and the second weight value to compensation values indexed from the LUT having the selected active operating point. The method of claim 9, The applying of the compensation value comprises: multiplying the first weight by a compensation value indexed at an LUT having an active operating point greater than the nonlinear transmission system operating point; Multiplying the second weight by a compensation value indexed at an LUT having an active operating point less than the nonlinear transmission system operating point; And adding a compensation value multiplied by the first weight value and a compensation value multiplied by the second weight value. The method of claim 7, wherein Using inverse Fourier transform (IFFT) on the inverse function spectrum for the spectrum representing the gain variation for each frequency of the nonlinear transmission system output as an initial filter coefficient and using filtering means for updating the coefficient for each new nonlinear transmission system input signal, And filtering the linearly distorted input signal to compensate for a change in gain per frequency of a nonlinear transmission system input signal. delete The method of claim 7, wherein Determining nonlinear characteristics of the nonlinear transmission system using the linearly distorted input signal and an output signal of the nonlinear transmission system; And updating a LUT having an active operating point closer to the nonlinear transmission system operating point among the selected active operating points to indicate an inverse characteristic of the nonlinear characteristic. The method of claim 13, The updating of the LUT may comprise comparing a distance between the nonlinear transmission system operating point and the selected active operating points; Selecting a LUT with a shorter distance between the operating points; And updating the LUT if the operating point is a value within an update region between a minimum threshold value and a maximum threshold value preset to an active operating point of the selected LUT.

Images