JP2010157938A - Distortion compensating circuit and radio base station - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To obtain a distortion compensating circuit capable of attaining high accuracy of distortion compensation, by performing model updating, even in a state where the frequency of occurrence of the maximum value input signal is low. <P>SOLUTION: A DPD (digital predistortion) processing unit 2 is provided with: an inverse model estimating unit 22, which estimates an inverse model for a model expressing the input/output characteristics of a HPA (high-power amplifier) 6, based on an input signal S1 to the HPA 6 and an output signal S10 from the HPA 6; a distortion compensating unit 26, which compensates distortion of the input/output characteristics by adding the inverse model to the input signal S1; and a sampling circuit 20, which carries out sampling signals S2, S10 in the latest predetermined time, and input them to the inverse model estimating unit 22. The inverse model estimating unit 22 updates the inverse model, based on the S2, S10 input from the sampling circuit 20, regardless of whether the maximum value which the input signal S1 is able to assume, is included in the sampling range by the sampling circuit 20. <P>COPYRIGHT: (C)2010,JPO&INPIT

Description

  The present invention relates to a distortion compensation circuit and a radio base station including the same.

  In a radio base station included in a communication system using a mobile phone or the like, a high power amplifier (HPA) for amplifying and outputting a transmission signal is mounted in the transmission processing unit. Generally, HPA gives priority to amplification efficiency, and therefore has low linearity of input / output characteristics. That is, in HPA, the input / output characteristics between the input signal and the output signal exhibit nonlinear distortion characteristics. Therefore, when an input signal is amplified using an HPA having such input / output characteristics, a desired output signal may not be obtained due to the distortion. Therefore, as one of the distortion compensation methods for compensating for such distortion, a model representing the input / output characteristics of the amplifier is estimated, and an inverse model exhibiting characteristics opposite to the model is generated by digital signal processing. A method (so-called DPD: Digital Pre-Distortion) that compensates for distortion in the input / output characteristics of the amplifier by adding its inverse model to the input signal to the amplifier (digital signal before being converted into an analog signal) The following non-patent document 1 proposes. In the following Non-Patent Documents 2 and 3, HPA high-efficiency amplification techniques are proposed.

Lei Ding, "Digital predistortion of Power Amplifiers for Wireless Applications", Georgia Institute of Technology, March 2004. Donald F. Kimball, et al., "High-Efficiency Envelope-Tracking W-CDMA Base-Station Amplifier Using GaN HFETs", IEEE Transactions on Microwave Theory and Techniques, Vol.54, NO.11, November 2006. Feipeng Wang, et al., "Design of Wide-Bandwidth Envelope-Tracking Power Amplifiers for OFDM Applications", IEEE Transactions on Microwave Theory and Techniques, Vol.53, NO.4, April 2005.

  Since the input / output characteristics of the HPA fluctuate due to temperature changes or the like, it is necessary to always update the inverse model to the latest one in order to realize highly accurate distortion compensation. Here, in order to accurately create an inverse model that covers the entire range of the input signal, an input signal including a maximum value (peak value) and an output signal corresponding to the input signal are required. However, the input signal with the maximum value is not always included in the communication signal, and the appearance frequency of the input signal with the maximum value decreases in a situation where the amount of communication data is small (such as a midnight time zone). Therefore, when updating the inverse model always waiting for the arrival of the maximum input signal, the frequency of updating the inverse model decreases in a situation where the amount of communication data is small, so high-precision distortion compensation can be performed. It becomes difficult.

  The present invention has been made in view of such circumstances, and it is possible to realize highly accurate distortion compensation by updating the inverse model even in a situation where the appearance frequency of the maximum input signal is low. An object of the present invention is to obtain a distortion compensation circuit and a radio base station including the distortion compensation circuit.

  The distortion compensation circuit according to the first aspect of the present invention includes: an estimation unit that estimates an inverse model for a model representing input / output characteristics of the amplifier based on an input signal to the amplifier and an output signal from the amplifier; By adding the inverse model to the input signal, a distortion compensator that compensates for distortion of the input / output characteristics, and the input signal and the output signal within the latest predetermined time are sampled and input to the estimation unit A sampling unit, and the estimation unit receives the input from the sampling unit regardless of whether a maximum value that the input signal can take is included in a range sampled by the sampling unit. The inverse model is updated based on the signal and the output signal.

  According to the distortion compensation circuit according to the first aspect, the estimation unit is input from the sampling unit regardless of whether or not the maximum value that the input signal can take is included in the range sampled by the sampling unit. The inverse model is updated based on the input signal and the output signal. Accordingly, since the estimation unit can update the inverse model even in a situation where the appearance frequency of the maximum value input signal is low, it is possible to realize highly accurate distortion compensation in the distortion compensation unit.

  The distortion compensation circuit according to a second aspect of the present invention is the distortion compensation circuit according to the first aspect, in particular, the estimation unit has a predetermined maximum value of the input signal within a range sampled by the sampling unit. The inverse model is updated on condition that the threshold value is equal to or greater than the threshold value.

  According to the distortion compensation circuit according to the second aspect, the estimation unit updates the inverse model on the condition that the maximum value of the input signal within the range sampled by the sampling unit is equal to or greater than a predetermined threshold value. To do. Therefore, when the signal level of the input signal within the sampled range is less than the threshold value, the inverse model is not updated. As a result, it can be avoided that the range of the input signal covered by the inverse model becomes extremely small.

  The distortion compensation circuit according to the third aspect of the present invention is the distortion compensation circuit according to the second aspect, in particular, the predetermined threshold value is set in accordance with the upper limit value of the input signal in the current inverse model. It is characterized by being set.

  According to the distortion compensation circuit according to the third aspect, a plurality of threshold values are set according to the upper limit value of the input signal in the current inverse model. Therefore, it is possible to avoid a sudden decrease in the range of the input signal covered by the inverse model.

  The distortion compensation circuit according to the fourth aspect of the present invention is the distortion compensation circuit according to any one of the first to third aspects, in particular, based on the average value of the input signal within the most recent predetermined period. A setting unit configured to set a predicted maximum value of the input signal; and after the sampling unit starts sampling, when the input signal that is equal to or greater than the predicted maximum value is detected, without waiting for the predetermined time to elapse The sampling unit ends sampling.

  According to the distortion compensation circuit according to the fourth aspect, the setting unit sets the predicted maximum value of the input signal, and after the sampling unit starts sampling, when the input signal equal to or larger than the predicted maximum value is detected The sampling unit ends the sampling. Accordingly, the sampling period can be shortened and the input signal of the predicted maximum value can be acquired with certainty, so that it is possible to reliably create an inverse model that covers the range below the predicted maximum value.

  A radio base station according to a fifth aspect of the present invention includes an amplifier and a distortion compensation circuit according to any one of the first to fourth aspects.

  According to the radio base station according to the fifth aspect, it is possible to transmit a desired transmission signal from the radio base station by appropriately compensating the distortion of the input / output characteristics of the amplifier by the distortion compensation circuit.

  According to the present invention, since the inverse model can be updated even in a situation where the appearance frequency of the maximum value input signal is low, highly accurate distortion compensation can be realized.

  Hereinafter, embodiments of the present invention will be described in detail with reference to the drawings. In addition, the element which attached | subjected the same code | symbol in different drawing shall show the same or corresponding element.

  FIG. 1 is a block diagram showing a part of the configuration of radio base station 1 according to the embodiment of the present invention. As shown in the connection relationship of FIG. 1, the radio base station 1 includes a DPD (Digital Pre-Distortion) processing unit 2, a DAC (Digital-to-Analog Converter) 3, an LPF (Low Pass Filter) 4, and a frequency conversion unit 5. , HPA (High Power Amplifier) 6, coupler 7, antenna 8, frequency converter 9, LPF 10, and ADC (Analog-to-Digital Converter) 11.

  The DPD processing unit 2 outputs the signal S2 by correcting the input signal S1, which is a digital signal. The details of correction by the DPD processing unit 2 will be described later. The DAC 3 converts the signal S2 that is a digital signal into a signal S3 that is an analog signal and outputs the signal S3. The LPF 4 performs a low-pass filter process on the signal S3 and outputs a signal S4. The frequency converter 5 converts the baseband signal S4 into a high-frequency signal S5 and outputs it. The HPA 6 outputs a signal S6 by amplifying the signal S5. The signal S6 is transmitted from the antenna 8.

  A part of the signal S6 from the HPA 6 toward the antenna 8 is extracted by the coupler 7 as a signal S7. The frequency converter 9 converts the high-frequency signal S7 into a baseband signal S8 and outputs it. The LPF 10 performs low-pass filter processing on the signal S8 and outputs a signal S9. The ADC 11 converts the signal S9, which is an analog signal, into a signal S10, which is a digital signal, and outputs it. The signal S10 is input to the DPD processing unit 2.

  FIG. 2 is a block diagram illustrating a first configuration example of the DPD processing unit 2. 2, the DPD processing unit 2 includes a sampling circuit 20, a detection unit 21, an inverse model estimation unit 22, an upper limit value storage unit 23, a coefficient storage unit 24, a determination unit 25, and a distortion compensation unit 26. It is configured with.

  The sampling circuit 20 receives the signal S2 from the distortion compensator 26 and the signal S10 from the ADC 11. The sampling circuit 20 samples the signals S2 and S10 within the latest predetermined time (hereinafter referred to as “sampling time”), and inputs these signals S2 and S10 to the inverse model estimation unit 22. The inverse model estimation unit 22 estimates a model representing the input / output characteristics of the HPA 6 based on the signals S2 and S10, and expresses the inverse model corresponding to the model in the form of a polynomial of an nth power series (n is a natural number). Each order coefficient (that is, a coefficient set of the inverse model) for obtaining is calculated. Here, the inverse model is a model that exhibits characteristics opposite to the distortion characteristics of the model for compensating for nonlinear distortion in the model.

  The detection unit 21 detects the maximum value (for example, the maximum power value) of the level of the input signal S1 within the sampling time, and inputs data S21 related to the maximum power value to the sampling circuit 20. The data S21 is input from the sampling circuit 20 to the inverse model estimation unit 22, and is associated with the inverse model coefficient set obtained above. The maximum power value given by the data S21 represents the upper limit value of the cover area covered by the corresponding inverse model.

  Data S22 regarding the coefficient set of the inverse model is input from the inverse model estimation unit 22 to the coefficient storage unit 24 and stored in the coefficient storage unit 24. In other words, the inverse model is stored in the coefficient storage unit 24. Further, the data S21 relating to the maximum power value corresponding to the inverse model is input from the inverse model estimation unit 22 to the upper limit value storage unit 23 and stored in the upper limit value storage unit 23.

  The distortion compensator 26 receives the input signal S1 and data S24 related to the coefficient set from the coefficient storage 24. The distortion compensation unit 26 corrects the input signal S1 based on a coefficient set (inverse model) given by the data S24. As a result, a signal S2 in which appropriate distortion compensation has been performed on the input signal S1 is output from the distortion compensation unit 26.

  By the way, since the input / output characteristics of the HPA 6 fluctuate due to a temperature change or the like, in order to realize highly accurate distortion compensation by the distortion compensator 26, the inverse model estimation unit 22 periodically updates the model and the inverse model. Need to be updated.

  FIG. 3 is a diagram illustrating an example of a plurality of models K0 to K2 updated in order. The horizontal axis represents the signal level (for example, power value) of the input signal S1, and the vertical axis represents the signal level (for example, power value) of the output signal (signal S10). The model K0 is a model estimated by the inverse model estimation unit 22 in a situation where the amount of communication data is large, and the upper limit value Pm0 of the coverage area of the model K0 is the maximum value (peak value) that the signal level of the input signal S1 can take. It matches. Therefore, by using the model K0, the inverse model estimation unit 22 can create an inverse model that covers the entire range of the input signal S1. Data S21 indicating the upper limit value Pm0 is stored in the upper limit value storage unit 23 in association with the inverse model of the model K0.

  The model K1 updated from the model K0 is a model estimated by the inverse model estimation unit 22 in a situation where the amount of communication data is slightly small. The upper limit value Pm1 of the coverage area of the model K1 is higher than the upper limit value Pm0 of the model K0. It is getting smaller. Therefore, by using an inverse model created based on the model K1, the distortion compensator 26 cannot perform distortion compensation on the input signal S1 having a signal level exceeding the upper limit value Pm1. However, with respect to the range below the upper limit value Pm1, input / output characteristics newer than the model K0 can be expressed by the model K1, so that highly accurate distortion compensation can be realized. The data S21 indicating the upper limit value Pm1 is stored in the upper limit value storage unit 23 in association with the inverse model of the model K1.

  The model K2 updated from the model K1 is a model estimated by the inverse model estimation unit 22 in a situation where the amount of communication data is still smaller, and the upper limit value Pm2 of the model K2 is larger than the upper limit value Pm1 of the model K1. It is getting smaller. Therefore, by using the inverse model created based on the model K2, the distortion compensator 26 cannot perform distortion compensation on the input signal S1 having a signal level exceeding the upper limit value Pm2. However, with respect to the range below the upper limit value Pm2, the input / output characteristics newer than the models K0 and K1 can be expressed by the model K2, so that highly accurate distortion compensation can be realized. Data S21 indicating the upper limit value Pm2 is stored in the upper limit value storage unit 23 in association with the inverse model of the model K2.

  In the DPD processing unit 2 according to the present embodiment, coefficient sets of all inverse models corresponding to the models K0 to K2 are stored in the coefficient storage unit 24. The distortion compensation unit 26 performs distortion compensation on the input signal S1 by selecting an optimal inverse model from among the three inverse models corresponding to the models K0 to K2. Specifically, it is as follows.

  FIG. 4 is a diagram for explaining selection of an inverse model by the distortion compensation unit 26. The input signal S1 is input to the distortion compensator 26 and also input to the determination unit 25. In addition, data S <b> 21 related to the upper limit value of the cover area of each inverse model is input from the upper limit value storage unit 23 to the determination unit 25. When the signal level of the input signal S1 is equal to or lower than the upper limit value Pm2, the determination unit 25 selects an inverse model corresponding to the model K2. That is, when the signal level of the input signal S1 is equal to or lower than the upper limit value Pm2, there are three inverse models corresponding to the models K0 to K2 as the inverse models that cover the signal level. In this case, the determination unit 25 selects the latest inverse model (the inverse model corresponding to the model K2) from these three inverse models. Data S23 related to selection of the inverse model is input from the determination unit 25 to the coefficient storage unit 24. Then, data S24 regarding the coefficient set of the selected inverse model is input from the coefficient storage unit 24 to the distortion compensation unit 26. The distortion compensator 26 corrects the input signal S1 based on the inverse model of the coefficient set given by the data S24.

  Similarly, when the signal level of the input signal S1 is greater than the upper limit value Pm2 and less than or equal to the upper limit value Pm1, the determination unit 25 selects an inverse model corresponding to the model K1. That is, when the signal level of the input signal S1 is greater than the upper limit value Pm2 and less than or equal to the upper limit value Pm1, there are two inverse models corresponding to the models K0 and K1 as the inverse models that cover the signal level. . In this case, the determination unit 25 selects the latest inverse model (the inverse model corresponding to the model K1) from the two inverse models, and the distortion compensation unit 26 is based on the selected inverse model. The input signal S1 is corrected.

  Further, when the signal level of the input signal S1 exceeds the upper limit value Pm1, only an inverse model corresponding to the model K0 exists as an inverse model that covers the signal level. In this case, the determination unit 25 selects an inverse model corresponding to the model K0, and the distortion compensation unit 26 corrects the input signal S1 based on the selected inverse model.

  In the above, the example in which the model is updated in the order in which the upper limit value gradually decreases (in the order of model K0 → K1 → K2) has been described. Unlike this, for example, when the models are updated in the order of model K 1 → K 0 → K 2, the inverse model corresponding to model K 1 is stored when the inverse model corresponding to model K 0 is stored in coefficient storage unit 24. You may delete from the coefficient memory | storage part 24. FIG. In this case, the determination unit 25 selects an inverse model corresponding to the model K2 when the signal level of the input signal S1 is equal to or lower than the upper limit value Pm2, and when the signal level of the input signal S1 exceeds the upper limit value Pm2. Will select the inverse model corresponding to model K0.

  FIG. 5 is a block diagram illustrating a second configuration example of the DPD processing unit 2. The comparison unit 30 is added to the configuration shown in FIG. 2, and the other configurations are the same as those in FIG.

  FIG. 6 is a diagram illustrating an example of a plurality of models K0 to K3 corresponding to FIG. A predetermined threshold value H is set for the signal level of the input signal S1. The threshold value H is set to, for example, half the maximum value (peak value) that the input signal S1 can take. However, the threshold value H may be set to other values. The set value of the threshold value H is taught to the comparison unit 30 in advance.

  When the inverse model estimation unit 22 estimates the inverse model, the comparison unit 30 includes a maximum value of the input signal S1 within the range sampled by the sampling circuit 20 (that is, an upper limit value given by the data S21), a threshold value, and the like. Compare with H. Then, the result of the comparison is input to the inverse model estimation unit 22 as data S30. Based on the data S30, the inverse model estimation unit 22 updates the inverse model when the maximum value of the input signal S1 within the sampling range is greater than or equal to the threshold value H. On the other hand, when the maximum value of the input signal S1 within the sampling range is less than the threshold value H, the inverse model is not updated. In the example shown in FIG. 6, although the inverse model corresponding to the model K0 → K1 → K2 is updated, the upper limit value of the cover area of the model K3 (that is, the input signal within the sampling range for estimating the model K3) Since the maximum value of S1) Pm3 is less than the threshold value H, the inverse model corresponding to the model K2 is not updated to the inverse model corresponding to the model K3.

  FIG. 7 is a diagram illustrating another example of the plurality of models K0 to K2 corresponding to FIG. A plurality of threshold values H0 and H1 are set for the signal level of the input signal S1. The threshold value H1 is set to, for example, a half value of the maximum value (peak value) that the input signal S1 can take. The threshold value H0 is set to an intermediate value between the peak value and the threshold value H1, for example. However, the threshold values H0 and H1 may be set to other values. Further, the number of threshold values is not limited to two, and may be three or more (for example, four). The set values of the threshold values H0 and H1 are taught in advance to the comparison unit 30.

  When the inverse model estimation unit 22 estimates the inverse model, the comparison unit 30 compares the maximum value of the input signal S1 within the sampling range with the threshold values H0 and H1. Then, the result of the comparison is input to the inverse model estimation unit 22 as data S30.

  When the inverse model estimation unit 22 updates the inverse model corresponding to the model K0 including the peak value to an inverse model having a smaller upper limit value of the coverage area, the updated upper limit value of the inverse model is the threshold value. Update is performed on condition that the value is equal to or greater than H0. That is, based on the data S30, updating is performed when the updated upper limit value of the inverse model is equal to or greater than the threshold value H0, and updating is not performed when it is less than the threshold value H0. In the example shown in FIG. 7, since the upper limit value Pm1 of the model K1 is equal to or higher than the threshold value H0, the inverse model corresponding to the model K0 is updated to the inverse model corresponding to the model K1. However, since the upper limit value Pm2 of the model K2 is less than the threshold value H0, the direct update from the inverse model corresponding to the model K0 to the inverse model corresponding to the model K2 is not performed. In the example shown in FIG. 7, instead of the data S21 indicating the upper limit value Pm0, flag information regarding that the signals S2 and S10 cover the peak value is sampled in association with the signals S2 and S10. Input from the circuit 20 to the inverse model estimation unit 22. Similarly, in the example shown in FIG. 7, instead of the data S21 indicating the upper limit value Pm1, flag information regarding that the signals S2 and S10 cover the threshold value H0 is associated with the signals S2 and S10. And input to the inverse model estimation unit 22 from the sampling circuit 20.

  In addition, when the inverse model estimation unit 22 updates the inverse model corresponding to the model K1 whose upper limit value is equal to or greater than the threshold value H0, to the inverse model having the lower upper limit value of the coverage area, the inverse model estimation unit 22 Update is performed on condition that the upper limit value of the inverse model is equal to or greater than the threshold value H1. That is, based on the data S30, updating is performed when the updated upper limit value of the inverse model is greater than or equal to the threshold value H1, and updating is not performed when it is less than the threshold value H1. In the example shown in FIG. 7, since the upper limit value Pm2 of the model K2 is equal to or higher than the threshold value H1, the inverse model corresponding to the model K1 is updated to the inverse model corresponding to the model K2. Similarly to the above, in the example shown in FIG. 7, instead of the data S21 indicating the upper limit value Pm2, flag information regarding that the signals S2 and S10 cover the threshold value H1 is the signal S2, Corresponding to S10, it is input from the sampling circuit 20 to the inverse model estimation unit 22.

  FIG. 8 is a diagram for explaining selection of an inverse model by the distortion compensator 26 corresponding to FIG. When the signal level of the input signal S1 is equal to or lower than the threshold value H1, the distortion compensation unit 26 is based on an inverse model corresponding to a portion equal to or lower than the threshold value H1 in the latest model K2 in this range. The input signal S1 is corrected. When the signal level of the input signal S1 is greater than the threshold value H1 and less than or equal to the threshold value H0, the distortion compensation unit 26 exceeds the threshold value H1 among the latest models K1 in this range. The input signal S1 is corrected based on the inverse model corresponding to the portion below the value H0. When the signal level of the input signal S1 is greater than the threshold value H0, the distortion compensator 26 is based on the inverse model corresponding to the portion of the latest model K0 in this range that exceeds the threshold value H0. Thus, the input signal S1 is corrected.

  FIG. 9 is a block diagram illustrating a third configuration example of the DPD processing unit 2. A setting unit 40 is added to the configuration shown in FIG. 2, and the other configurations are the same as those in FIG.

  The setting unit 40 obtains an average value of the signal level (for example, power value) of the input signal S1 within a predetermined period immediately before the sampling circuit 20 starts sampling. Then, a value higher than the average value by a predetermined level (for example, 10 to 11 dBm) is set as the predicted maximum value of the input signal S1. The upper limit of the predicted maximum value is the maximum value (peak value) that the input signal S1 can take. Data S40 related to the predicted maximum value is input from the setting unit 40 to the detection unit 21.

  When the sampling circuit 20 starts sampling, the detection unit 21 sequentially monitors the signal level of the input signal S1 to be sampled. And if the detection part 21 detects the input signal S1 more than the prediction maximum value given by data S40, the data S41 which shows that will be input into the sampling circuit 20. FIG. Receiving the data S41, the sampling circuit 20 ends the sampling at that time. On the other hand, when the detection unit 21 does not detect the input signal S1 that is equal to or greater than the predicted maximum value given by the data S40, the sampling circuit 20 ends the sampling after a predetermined time has elapsed from the start of sampling.

  As described above, according to the DPD processing unit 2 (distortion compensation circuit) according to the present embodiment, the inverse model estimation unit 22 samples the maximum value (peak value) that can be taken by the input signal S1 by the sampling circuit 20. The inverse model is updated based on the signals S2 and S10 input from the sampling circuit 20 regardless of whether or not they are included in the input signal S1 within the range. Accordingly, the inverse model estimation unit 22 can update the model even in a situation where the appearance frequency of the input signal S1 having the maximum value (peak value) is low, and the inverse model is also updated accordingly. High-precision distortion compensation can be realized.

  Also, according to the DPD processing unit 2 shown in FIGS. 5 and 6, the inverse model estimation unit 22 has a maximum value of the input signal S1 within a range sampled by the sampling circuit 20 being a predetermined threshold value H or more. The inverse model is updated on the condition that it exists. Therefore, when the signal level of the input signal S1 within the sampled range is less than the threshold value H, the inverse model is not updated. As a result, it can be avoided that the range of the input signal S1 covered by the inverse model becomes extremely small.

  Further, according to the DPD processing unit 2 shown in FIG. 7, a plurality of threshold values H0 and H1 are set according to the upper limit value of the input signal S1 in the current inverse model. Therefore, it is possible to avoid a sudden decrease in the range of the input signal S1 covered by the inverse model. For example, it can be avoided that the inverse model corresponding to the model K0 is directly updated to the inverse model corresponding to the model K2.

  Further, according to the DPD processing unit 2 shown in FIG. 9, the setting unit 40 sets the predicted maximum value of the input signal S1, and after the sampling circuit 20 starts sampling, the input signal S1 that is equal to or larger than the predicted maximum value. Is detected, the sampling circuit 20 ends the sampling. Therefore, the sampling period can be shortened, and the input signal S1 having the predicted maximum value can be acquired with certainty, so that it is possible to reliably create an inverse model that covers the range below the predicted maximum value.

  Further, according to the radio base station 1 according to the present embodiment, a desired transmission signal is transmitted from the radio base station 1 by appropriately compensating for distortion of the input / output characteristics of the HPA 6 by the DPD processing unit 2. Is possible.

  The embodiment disclosed this time should be considered as illustrative in all points and not restrictive. The scope of the present invention is defined not by the above-mentioned meaning but by the scope of claims for patent, and is intended to include all modifications within the scope and meaning equivalent to the scope of claims for patent.

It is a block diagram which shows a part of structure of the wireless base station which concerns on embodiment of this invention. It is a block diagram which shows the 1st structural example of a DPD process part. It is a figure which shows an example of the some model updated in order. It is a figure for demonstrating selection of the inverse model by a distortion compensation part. It is a block diagram which shows the 2nd structural example of a DPD process part. It is a figure which shows an example of a some model corresponding to FIG. It is a figure which shows the other example of a some model corresponding to FIG. FIG. 8 is a diagram for explaining selection of an inverse model by a distortion compensator corresponding to FIG. 7. It is a block diagram which shows the 3rd structural example of a DPD process part.

Explanation of symbols

1 Wireless base station 2 DPD processor 6 HPA
DESCRIPTION OF SYMBOLS 20 Sampling circuit 21 Detection part 22 Inverse model estimation part 23 Upper limit value memory | storage part 24 Coefficient memory | storage part 25 Judgment part 26 Distortion compensation part 30 Comparison part 40 Setting part H, H0, H1 Threshold value K0-K2 Model Pm0-Pm2 Upper limit value

Claims (5)

An estimation unit for estimating an inverse model for a model representing input / output characteristics of the amplifier based on an input signal to the amplifier and an output signal from the amplifier;
A distortion compensator that compensates for distortion of the input / output characteristics by adding the inverse model to the input signal;
A sampling unit that samples the input signal and the output signal within the most recent predetermined time and inputs them to the estimation unit;
The estimation unit determines whether the maximum value that the input signal can take is included in the range sampled by the sampling unit, whether the input signal or the output signal is input from the sampling unit. A distortion compensation circuit for updating the inverse model based on the distortion compensation circuit.   The distortion according to claim 1, wherein the estimation unit updates the inverse model on condition that a maximum value of the input signal within a range sampled by the sampling unit is equal to or greater than a predetermined threshold value. Compensation circuit.   The distortion compensation circuit according to claim 2, wherein a plurality of the predetermined threshold values are set according to an upper limit value of the input signal in the current inverse model. Based on the average value of the input signal within the most recent predetermined period, further comprising a setting unit for setting the predicted maximum value of the input signal,
The sampling unit finishes sampling without waiting for the elapse of the predetermined time when the input signal equal to or greater than the predicted maximum value is detected after sampling by the sampling unit is started. The distortion compensation circuit according to any one of the above. An amplifier;
A radio base station comprising the distortion compensation circuit according to claim 1.

Images