Distortion compensation circuit, wireless device

Abstract

To perform stable and precise distortion compensation in a case an operation signal is a burst signal.SOLUTION: A system control part 18 sets whether a coefficient selection part 152 selects an update distortion compensation coefficient or a fixed distortion compensation coefficient. The system control part 18 compares length relationship between a burst length of an input signal and a burst length threshold value stored in a storage part 20, if the burst length is the burst length threshold value or more, it causes the coefficient selection part to select the update distortion compensation coefficient, and if the burst length is less than the burst length threshold value, it cases the coefficient selection part to select the fixed distortion compensation coefficient. The fixed distortion compensation coefficient is defined not only by bandwidth, frequency, and modulation method, but also a duty ratio of the burst signal and a temperature of a power amplifier 100.SELECTED DRAWING: Figure 1

Description

[Technical Field]

[0001] The present invention relates to a distortion compensation circuit that compensates for nonlinearity that occurs during amplification in a power amplifier circuit used in a radio device or the like, and to a radio device using the same. [Background technology]

[0002] Wireless devices in mobile communication systems and the like use power amplifiers to increase the power of transmitted signals. Generally, the amplification characteristics of a power amplifier have high linearity when the power of the input signal is small (within an appropriate range), but have saturation characteristics such that the power of the output signal saturates when the power of the input signal is large (outside the appropriate range). As a result, particularly when the power of the input signal is large, a waveform change (nonlinear distortion) occurs in the output signal from the input signal. For this reason, generally, when the power of the output signal is increased, nonlinear distortion also increases, and there is a trade-off between increasing the power of the output signal and suppressing nonlinear distortion.

[0003] In order to compensate for such nonlinear distortion in the output signal, a distortion compensation circuit is connected to the power amplifier. A known distortion compensation method is a predistortion method in which such nonlinear distortion is predicted in advance and a distortion is introduced into the input signal before it is input to the power amplifier, which distortion cancels the nonlinear distortion in the output signal, thereby reducing the nonlinear distortion in the output signal.

[0004] In such a predistortion scheme, to stably compensate for nonlinear distortion in an output signal, it is necessary to optimize the distortion introduced into the input signal so that the nonlinear distortion in the output signal is always minimized. In a predistortion distortion compensation circuit, the distortion introduced into the input signal is set by a certain parameter (distortion compensation coefficient). However, the state of such nonlinear distortion varies depending on the type of input signal (frequency, modulation method, bandwidth, etc.) and the environment of the power amplifier (temperature, etc.). For this reason, in a predistortion distortion compensation circuit, optimal values ​​for the distortion compensation coefficient corresponding to the type of input signal, the environment of the power amplifier, etc. are pre-determined and stored for each condition. In this case, an appropriate distortion compensation coefficient can be selected and used after recognizing which condition the current input signal and state of the power amplifier correspond to (fixed coefficient scheme). The distortion compensation coefficient for each condition is calculated using a test signal, which is a dummy input signal used only for calculating the distortion compensation coefficient, rather than an operational signal that serves as the input signal during normal operation of the device (e.g., wireless device) in which the power amplifier is used. The test signal is generated inside the distortion compensation circuit in a form that allows the distortion compensation coefficient to be calculated with high precision, for example, as a continuous wave.

[0005] On the other hand, the state of nonlinear distortion also fluctuates with changes over time in the power amplifier (radio device, etc.). For this reason, it is preferable to calculate the latest distortion compensation coefficients and update the distortion compensation coefficients stored as described above (update method). In this case, if a test signal is used as the input signal to the power amplifier device, the radio device, etc., cannot be used as intended. Therefore, although the accuracy is lower than when a test signal is used, an actual operating signal can also be used to calculate the distortion compensation coefficients. This allows the distortion compensation coefficients to be updated more frequently.

[0006] However, there are cases where the operating signal is not suitable for calculating the distortion compensation coefficient. For example, the operating signal may be a burst signal whose intensity is maintained only for a certain short period of time (burst length). When the burst length is short, the number of samples used when comparing the waveform of the input signal with the waveform of the output signal is insufficient, making it difficult to accurately calculate the distortion compensation coefficient. In the technology described in Patent Document 1, when the input signal (operating signal) is such a burst signal in the update method, the technology properly recognizes this fact and does not adopt the distortion compensation coefficient calculated in that case (does not update it as the latest distortion compensation coefficient), thereby achieving more appropriate distortion compensation.

[0007] Patent Document 2 describes an apparatus in which an operating signal and a test signal are input to a power amplifier, and a distortion compensation coefficient is calculated from the output signal of the power amplifier corresponding to the continuous wave test signal. This apparatus basically uses a fixed coefficient method, and the calculation of the distortion compensation coefficient is performed for each temperature by changing the temperature of the power amplifier. Even if the temperature of the power amplifier is the same, different distortion compensation coefficients are applied to the operating signal when the signal is a continuous wave signal and when the signal is a burst signal. However, it is also possible to input a test signal to the power amplifier while the operating signal is not being input to the power amplifier, and calculate and update the distortion compensation coefficient. The test signal can also be a burst signal, and a distortion compensation coefficient corresponding to the burst signal can be calculated.

[0008] Furthermore, as mentioned above, the accuracy of calculation of the optimal distortion compensation coefficient is lower when the burst length is short compared to when the burst length is long. For this reason, Patent Document 3 describes a technique for switching between a distortion compensation coefficient that is appropriately updated as described above (an updated distortion compensation coefficient) and a distortion compensation coefficient that is preset without such updating (a fixed distortion compensation coefficient) according to the recognized burst length. In this case, a threshold value for the burst length for this switching is also appropriately updated, similarly to the updated distortion compensation coefficient, according to the accuracy of the appropriately calculated updated distortion compensation coefficient. As a result, the updated distortion compensation coefficient is used when its accuracy is high, and the fixed distortion compensation coefficient is used when the burst length is short and the accuracy of the distortion compensation coefficient calculated as described above is low. Therefore, by setting the fixed distortion compensation coefficient to an appropriate value, distortion compensation can be performed with high accuracy regardless of the burst length. That is, the technique described in Patent Document 3 enables distortion compensation to be performed with higher accuracy for signals of various burst lengths compared to the techniques described in Patent Documents 1 and 2. [Prior art documents] [Patent documents]

[0009] [Patent Document 1] International Publication No. 2007 / 049474 [Patent Document 2] Japanese Patent Application Laid-Open No. 2005-33535 [Patent Document 3] International Publication No. 2021 / 039256 DISCLOSURE OF THE INVENTION [Problem to be solved by the invention]

[0010] For example, the temperature of the power amplifier described above is not simply determined by the burst length, but changes in a complex manner due to various factors. For this reason, even with the technology described in Patent Document 3, there are cases where distortion compensation cannot be performed with high precision.

[0011] For example, when a burst signal is repeatedly transmitted intermittently, a state in which a burst signal is transmitted (transmission interval) and a non-transmission interval (non-transmission interval) alternate may be periodically repeated for a long period of time. In this case, even if the burst length is the same, if the non-transmission interval is long (when the duty ratio is small) and if the non-transmission interval is short (when the duty ratio is large), the latter case essentially becomes closer to a continuous wave (corresponding to an extremely long burst length), and the temperature rise is more severe than the former case. In other words, even with the same burst length, the temperature state of the power amplifier varies greatly depending on the repetition situation. Furthermore, if the latter operation is repeated for a long period of time, the temperature gradually rises over time. For this reason, it is not appropriate to set the distortion compensation conditions based solely on the burst length, and even the technology described in Patent Document 3 has sometimes failed to perform appropriate distortion compensation.

[0012] For this reason, it has been desired to perform stable and highly accurate distortion compensation, particularly when the operating signal is a burst signal.

[0013] The present invention has been made in view of the above circumstances, and has as its object to solve the above problems. [Means for solving the problem]

[0014] The present invention provides a pre-distortion distortion compensation circuit that performs distortion compensation to impart to an input signal a distortion that compensates for nonlinear distortion, which is a difference between the waveform of an output signal obtained by amplifying an input signal by a power amplifier and the waveform of the input signal, and the type of distortion is set by a distortion compensation coefficient selected in accordance with the type of the input signal, wherein the input signal is constituted by an intermittently repeated burst signal, and the circuit is equipped with: a memory unit that stores a plurality of distortion compensation coefficients set in accordance with the type of the input signal; a temperature notification unit that detects the temperature of the power amplifier; a control unit that recognizes the type of the input signal and selects one of the plurality of distortion compensation coefficients stored in the memory unit; and a distortion imparting unit that imparts to the input signal the distortion corresponding to the distortion compensation coefficient selected by the control unit,The setting conditions are bandwidth, frequency, modulation method, burst length, duty ratio, and temperature. and two types of distortion compensation coefficients, an updated distortion compensation coefficient and a fixed distortion compensation coefficient, are set as the distortion compensation coefficient according to a burst length of the input signal, and the storage unit stores a plurality of the updated distortion compensation coefficients, a plurality of the fixed distortion compensation coefficients, and a burst length threshold value which is the burst length and determines whether the updated distortion compensation coefficient or the fixed distortion compensation coefficient is to be selected as the distortion compensation coefficient. For each of the above setting conditions and storing the Setting conditions The fixed distortion compensation coefficient is set in accordance with the burst length of the input signal. corresponding to the set conditions If the burst length is equal to or greater than the burst length threshold, the updated distortion compensation coefficient corresponding to the burst length is selected as the distortion compensation coefficient to be used in the distortion adding unit, and corresponding to the set conditions If the burst length is less than the burst length threshold, the fixed distortion compensation coefficient corresponding to the duty ratio and the temperature of the input signal is selected as the distortion compensation coefficient to be used in the distortion imparting unit, and the frame structure of the burst signal is provided with a data section in which data to be transmitted by the input signal and the output signal is transferred, and a non-data section in which information other than the data accompanying the data transmission is transferred, and when the time when the temperature of the power amplifier is recognized corresponds to the non-data section in the burst signal, the control section corrects the temperature to a value in the data section and uses the corrected temperature when selecting the fixed distortion compensation coefficient. The present invention includes a coefficient generation unit that recognizes the burst length in the input signal, and calculates, by statistical processing, a distortion compensation coefficient to be used for the distortion compensation and the accuracy of calculation of the distortion compensation coefficient by comparing the input signal with the corresponding output signal, and the control unit, when it is recognized that the accuracy is high, may cause the calculated distortion compensation coefficient to be updated in a storage unit as the updated distortion compensation coefficient in association with the burst length, and may update the burst length threshold stored in the storage unit in accordance with the accuracy when the distortion compensation coefficient is calculated in the coefficient generation unit in accordance with the input signal and the burst length of the input signal. A radio device of the present invention includes the distortion compensation circuit and the power amplifier. [Effects of the Invention]

[0015] According to the present invention, when the operation signal is a burst signal, stable and highly accurate distortion compensation can be performed. [Brief explanation of the drawings]

[0016] [Figure 1] 1 is a diagram illustrating a configuration of a wireless device in which a distortion compensation circuit according to an embodiment is used. [Figure 2] FIG. 2 is a diagram showing a frame structure of an operational signal and a test signal used. [Figure 3] 3A and 3B are diagrams illustrating formats of an updated distortion compensation coefficient and a burst length threshold value stored in a storage unit in the distortion compensation circuit according to the embodiment. [Figure 4] 3 is a diagram illustrating a format of a fixed distortion compensation coefficient stored in a storage unit in the distortion compensation circuit according to the embodiment. FIG. [Figure 5] 4 is a flowchart showing a distortion compensation operation in the distortion compensation circuit according to the embodiment. [Figure 6] 10 is a flowchart showing an operation of updating a distortion compensation coefficient and a burst length threshold using an operation signal in the distortion compensation circuit according to the embodiment. [Figure 7] 10 is a flowchart showing an operation of updating an updated distortion compensation coefficient using a test signal in the distortion compensation circuit according to the embodiment. [Figure 8] 5 is a flowchart showing an operation of setting a fixed distortion compensation coefficient in the distortion compensation circuit according to the embodiment. [Figure 9] 10A and 10B are diagrams showing a schematic diagram of a change in temperature over time of a power amplifier when the duty ratio of an input signal is different from that of the power amplifier when the duty ratio is different from that of the input signal; [Figure 10] FIG. 10 is a diagram showing a schematic diagram of a change in temperature of a power amplifier over time when a single burst signal is input; DETAILED DESCRIPTION OF THE INVENTION

[0017] Next, an embodiment of the present invention will be described in detail with reference to the drawings. FIG. 1 is a block diagram showing the configuration of a radio device using a distortion compensation circuit 1 according to an embodiment of the present invention. The distortion compensation circuit 1 is connected to a power amplifier 100 and is used to compensate for nonlinear distortion in the power amplifier 100. In practice, the distortion compensation circuit 1 and the power amplifier 100 are combined and used in a radio device. The radio device uses the power amplifier 100 to amplify an operational signal and transmit it to the outside. FIG. 1 shows only parts of the radio device related to the power amplifier 100 and the distortion compensation circuit 1, and other parts are omitted. Furthermore, as will be described later, many of the components of the distortion compensation circuit 1 are common to those described in Patent Document 3. Therefore, although the description of these common components will be partially simplified below, the content is the same as that of Patent Document 3.

[0018] Here, since this distortion compensation circuit 1 operates using a predistortion method, distortion using a distortion compensation coefficient for compensating for nonlinear distortion generated in the power amplifier 100 is added to the operational signal before it is input to the power amplifier 100. Meanwhile, a test signal whose mode is controlled in advance and set solely for updating this distortion compensation coefficient is generated by a test signal generator 11. The test signal controls the frequency, bandwidth, modulation method, burst length, duty ratio, repetition rate, etc., and the distortion compensation coefficient corresponding to each of these conditions is calculated using this test signal. The operational signal and the test signal are input to a selector 12, and one of them becomes the input signal to the power amplifier 100. This selection is set by a switching operation signal issued by a user operation outside the scope of FIG. 1.

[0019] This input signal is smoothed by a peak power suppression unit 13 so that the PAPR (Peak to Average Power Ratio) is kept within an appropriate range, and then the transmission power is adjusted by a transmission power control unit 14 so that the transmission power is within a certain range, and then input to a distortion compensation calculation unit 15 which performs distortion compensation.

[0020] As with the technology described in Patent Document 3, there are two types of distortion compensation coefficients applied by the distortion compensation calculation unit 15: updated distortion compensation coefficients and fixed distortion compensation coefficients. These are stored in a storage unit 20 configured with a hard disk or nonvolatile memory according to the bandwidth, frequency, modulation method, burst length, duty ratio, etc. As will be described later, updated distortion compensation coefficients are used when the burst length is long (close to continuous wave), and fixed distortion compensation coefficients are used when the burst length is short. Distortion is applied to the input signal by a coefficient multiplication unit (distortion applying unit) 151 in the distortion compensation calculation unit 15 using one of the updated distortion compensation coefficient and the fixed distortion compensation coefficient selected by a coefficient selection unit 152. This is similar to the technology described in Patent Document 3.

[0021] The updated distortion compensation coefficient is calculated by the coefficient generation unit 153 using the output signal when an input signal is input to the power amplifier 100 in accordance with the operation of the distortion compensation circuit 1 (wireless device). In this process, the FF (feedforward) signal, which is the input signal to the distortion compensation calculation unit 15, is compared with the FB (feedback) signal, which is the corresponding output signal from the power amplifier 100, and the distortion compensation coefficient is calculated by extracting representative points in the waveforms of both signals, performing data interpolation, and various statistical processing. In this process, the coefficient generation unit 153 can also quantify and recognize the accuracy (estimation error) of the calculated distortion compensation coefficient. When a burst signal is repeatedly transmitted, transmission intervals during which burst signals are transmitted and non-transmission intervals during which no (burst) signal is transmitted alternate with each other. When the coefficient generation unit 153 performs the above process, the offset component correction unit 16 calculates the offset component in the FB signal using data from the non-transmission interval when the input signal is a burst signal, and corrects this in advance.

[0022] Furthermore, in this case, in order to cope with the case where the input signal is a burst signal, a data capture control unit 17 is provided which recognizes the burst section of the burst signal. In the data capture control unit 17, the timing of the start and end of the burst of the FF signal is recognized by an FF signal buffer unit 171, and the timing of the start and end of the burst of the FB signal is recognized by an FB signal buffer unit 172, and the coefficient generation unit 153 in the distortion compensation calculation unit 15 recognizes these timings during the above processing. With this configuration, the calculated distortion compensation coefficients (updated distortion compensation coefficients) are stored in a storage unit 20 according to the bandwidth, frequency, modulation method, burst length, duty ratio, etc.

[0023] However, when the coefficient generation unit 153 determines that the accuracy of the distortion compensation coefficients calculated as described above is not high, it does not update the updated distortion compensation coefficients in the storage unit 20, but only updates the updated distortion compensation coefficients when the accuracy is high. Such low accuracy corresponds to a short burst length, and high accuracy corresponds to a long burst length (close to a continuous wave). The technique for updating the distortion compensation coefficients is the same as that described in Patent Document 3.

[0024] The fixed distortion compensation coefficients are stored in the storage unit 20 according to the bandwidth, frequency, modulation method, burst length, duty ratio, temperature of the power amplifier 100, etc. However, unlike the updated distortion compensation coefficients, the fixed distortion compensation coefficients are not updated, at least according to the operation of the distortion compensation circuit 1 (wireless device), once they are set in advance. This point is also similar to the technology described in Patent Document 3. Therefore, when the accuracy of the updated distortion compensation coefficients is low as described above, a coefficient with higher accuracy than the updated distortion compensation coefficient can be calculated instead of the updated distortion compensation coefficient, particularly when the burst length is short, and this coefficient can be used as the fixed distortion compensation coefficient. However, as will be described later, the fixed distortion compensation coefficients can be calculated and stored by the distortion compensation circuit 1 in the same way as the updated distortion compensation coefficients, or an updated distortion compensation coefficient calculated under specific conditions can be stored as the fixed distortion compensation coefficient.

[0025] In general, the state of nonlinear distortion (or the distortion compensation coefficient corresponding thereto) fluctuates depending on the environment (temperature, etc.) of the power amplifier and changes over time. From this perspective, it is preferable to use updated distortion compensation coefficients for distortion compensation. Therefore, for example, when the input signal (operation signal) is a continuous wave, distortion compensation using the latest updated distortion compensation coefficients is performed by the distortion compensation calculation unit 15, thereby making it possible to properly compensate for nonlinear distortion.

[0026] However, when the burst length of an input signal is short, the accuracy of calculating distortion compensation coefficients using the input signal as described above is not high. That is, when the burst length of an input signal is short, the updated distortion compensation coefficients calculated using the input signal are often inappropriate. For this reason, when the burst length of an input signal is short, it is preferable to separately calculate in advance fixed distortion compensation coefficients that are particularly suitable for short burst lengths and use these. Furthermore, it is preferable not to use low-accuracy distortion compensation coefficients calculated using an input signal with a short burst length in subsequent distortion compensation. As an indicator for this determination, the accuracy when the coefficient generation unit 153 calculates the distortion compensation coefficients can be used. That is, this accuracy (estimation error) can be quantified, and if the estimation error exceeds a certain value, the distortion compensation coefficient is not used (it is not updated as an updated distortion compensation coefficient).

[0027] On the other hand, the fixed distortion compensation coefficients used when the burst length is short are not updated according to the operation as described above, and the fixed distortion compensation coefficients are not updated according to the operation of the distortion compensation circuit 1 (radio device) at least like the updated distortion compensation coefficients. However, as will be described later, they can be updated as appropriate by user operation or the like.

[0028] As described above, whether the coefficient selector 152 selects an updated distortion compensation coefficient or a fixed distortion compensation coefficient is determined by the scheme controller (controller) 18. The scheme controller 18 compares the burst length of the input signal with the magnitude relationship of the burst length threshold (FIG. 3) corresponding to the mode of this input signal, which is stored in the memory 20, and selects an updated distortion compensation coefficient if the burst length is equal to or greater than the burst length threshold, or selects a fixed distortion compensation coefficient if the burst length is less than the burst length threshold. Here, like the updated distortion compensation coefficient, the value of this burst length threshold is also updated in accordance with the operation of the distortion compensation circuit 1 (wireless device). In particular, when updating, the accuracy with which the updated distortion compensation coefficient is calculated by the coefficient generator 153 can be referenced. This point is also similar to the technology described in Patent Document 3.

[0029] However, in the technology described in Patent Document 3, both the updated distortion compensation coefficients and the fixed distortion compensation coefficients used are determined for each of the conditions, using the modulation method, bandwidth, frequency, and burst length of the operation signal as parameters, whereas the fixed distortion compensation coefficients used in this distortion compensation circuit 1 use the temperature and duty ratio of the power amplifier 100 as parameters in addition to these. For this reason, Fig. 1 includes a temperature notification unit 19 that measures the temperature of the power amplifier 100. The temperature notification unit 19 includes a temperature detection unit 191 that actually measures the temperature of the power amplifier 100, and a temperature calculation unit 192 that performs filtering processing, such as averaging, on the measured temperature value so that switching in response to temperature changes can be performed appropriately.

[0030] If there were a one-to-one correspondence between the temperature recognized by the temperature notification unit 19 and the state of nonlinear distortion occurring in the power amplifier 100, ideal distortion compensation would be possible by selecting a distortion compensation coefficient based only on this temperature. However, in reality, the temperature actually measured by the temperature notification unit 19 (temperature detection unit 191) or the temperature output from the temperature calculation unit 192 corresponds to the temperature of only a part of the power amplifier 100, and this temperature does not necessarily correspond one-to-one to the state of nonlinear distortion occurring in the power amplifier 100. However, since there is clearly a correlation between this temperature and the state of nonlinear distortion, in this distortion compensation circuit 1, the temperature notified by the temperature notification unit 19 is used, along with other parameters, to select a distortion compensation coefficient.

[0031] Similar to the technology described in Patent Document 3, the parameters other than the temperature include the modulation method, bandwidth, frequency, and burst length. Furthermore, as will be described later, when a burst signal is generated intermittently, its duty cycle is also included in these parameters. The values ​​of these parameters (or information regarding the modulation method) can be included as a header in the input signal (operation signal, test signal). Figure 2 is a diagram schematically illustrating the frame structure of the operation signal and test signal used here. Here, a header H containing information regarding the modulation method, bandwidth, frequency, burst length, and duty cycle can be provided after a preamble P used for synchronization and before a frame of data D that transmits the information that should originally be transmitted. The test signal generator 11 appropriately sets the header H in the test signal.

[0032] The method control unit 18 in Fig. 1 recognizes various types of information in such input signals (operation signal, test signal). First, when a currently input signal is recognized, the frame information control unit 181 recognizes whether the currently input signal is a preamble P, a header H, or data D in Fig. 2. Here, if it is recognized as a preamble P, a header H, or data D, for example, the transmission frame type signal is set to 1, 2, or 3. Furthermore, when a burst signal is transmitted intermittently, the transmission frame type signal is set to 0 in non-transmission sections.

[0033] Furthermore, by recognizing the header H of the input signal, the frame information control unit 181 can recognize the modulation method, bandwidth, frequency, burst length, and duty ratio specified in the header H in Fig. 2. The method selection unit 182 uses the information recognized from the header H, as well as the temperature recognized by the temperature notification unit 19, to control the coefficient selection unit 152. This operation will be described later.

[0034] The coefficient update determination unit 183 recognizes and notifies the timing of updating (storing in the storage unit 20) the updated distortion compensation coefficients as described above through user operation. This allows the user to recognize the timing when the latest updated distortion compensation coefficients were obtained. In this case, the updated distortion compensation coefficients are obtained for each condition of bandwidth, frequency, modulation method, and burst length as described above, and this timing can also be recognized for each of these conditions.

[0035] The update impossibility detection unit 184 recognizes whether or not the updated distortion compensation coefficient could not be updated for the input signal because the calculation accuracy in the coefficient generation unit 153 was not high (the estimation error was large). As described in Patent Document 3, such estimation error becomes large especially when the burst length is short. The scheme selection threshold generation unit 185 can recognize the burst length of the input signal for which updating was not possible using the frame information control unit 181. The scheme selection threshold generation unit 185 can update the burst length threshold stored in the storage unit 20, taking into account the burst length when updating was not possible.

[0036] For example, if 20 μsec is stored as the burst length threshold value in the storage unit 20, and the scheme selection threshold generation unit 185 recognizes that the updated distortion compensation coefficient could not be updated when the burst length is newly set to 21 μsec, the burst length threshold value can be updated from 20 μsec to 22 μsec (21 μsec + α). As a result, when the burst length is 21 μsec, the updated distortion compensation coefficient is used before this, but the fixed distortion compensation coefficient is used after this. Similar to the coefficient update determination unit 183, the threshold update determination unit 186 recognizes and notifies the user of the timing for updating the burst length threshold (storing it in the storage unit 20) based on a user operation.

[0037] As described above, the fixed distortion compensation coefficient applied when the burst length is short is not updated according to the operation. On the other hand, instead, the burst length threshold is updated according to the operation. Therefore, even if there is a change over time, the range of burst lengths to which the updated distortion compensation coefficient is applied is kept appropriate, and the distortion compensation response becomes more appropriate. This point is also similar to the technology described in Patent Document 3.

[0038] FIG. 3 shows a schematic representation of the format of the updated distortion compensation coefficients and burst length thresholds stored and updated in the memory unit 20. Here, the bandwidth is A1, the frequencies within it are B1 and B2, and the modulation methods for each frequency are C1, C2, and C3. The burst length thresholds and updated distortion compensation coefficients are set corresponding to each condition, and each value is updated for each item. Here, aaa, bbb, etc. are numerical values ​​that become the actual distortion compensation coefficients (updated distortion compensation coefficients), and these may be not single numerical values ​​but also numerical strings or matrices. Although not shown, the same applies to bandwidth A2 and other bandwidths. The scheme selection unit 182 recognizes the burst length thresholds corresponding to each condition recognized in the input signal in FIG. 3. If the burst length of the input signal is equal to or greater than the burst length thresholds, the scheme selection unit 182 sets the scheme (updating scheme) for using the updated distortion compensation coefficients as the distortion compensation coefficients used in the coefficient multiplication unit 151. In this case, the coefficient selection unit 152 selects the updated distortion compensation coefficient corresponding to this condition and outputs it to the coefficient multiplication unit 151. The content of the updated distortion compensation coefficients and the operation thereof are the same as those described in Patent Document 3.

[0039] In the technology described in Patent Document 3, the fixed distortion compensation coefficients used when the burst length is shorter than the burst length threshold are determined for each bandwidth, frequency, and modulation method, similar to the updated distortion compensation coefficients shown in Fig. 3. In contrast, the fixed distortion compensation coefficients used by the method control unit 18 are determined not only according to the bandwidth, frequency, and modulation method, but also according to the duty ratio of the burst signal and the temperature of the power amplifier 100. Fig. 4 shows the format of such fixed distortion compensation coefficients, similar to Fig. 3. Here, only the case of bandwidth A1 is shown, but the same applies to other bandwidths.

[0040] Here, modulation methods C1, C2, and C3 are set for each of frequencies B1 and B2, and further, duty ratios D1, D2, D3... (D1 < D2 < D3 <...) of the burst signal are set therein. Also, temperatures T1, T2, T3... (T1 < T2 < T3 <...) are set as the above temperatures. In this case, the item of duty ratio D1 indicates the case where 0 < duty ratio ≤ D1, and the item of D2 indicates the case where D1 < duty ratio ≤ D2 (the same applies to D3 and later). Similarly, the item of temperature T1 indicates the case where room temperature < temperature ≤ T1, and the item of temperature T2 indicates the case where T1 < temperature ≤ T2 (the same applies to T3 and later). A, AA, A1, etc. are actually numerical values that become distortion compensation coefficients (fixed distortion compensation coefficients), and like the updated distortion compensation coefficients, they may be numerical sequences or matrices. Therefore, in FIG. 4, for example, the fixed distortion compensation coefficient in the case of frequency B1, modulation method C2, D1 < duty ratio ≤ D2, and T1 < temperature ≤ T2 is BB1.

[0041] Each numerical value in FIG. 4 can be set in advance by experiments or the like, but as described later, it is also possible to use the value calculated by the coefficient generation unit 153 as the updated distortion compensation coefficient under specific conditions. Also, the bandwidth, frequency, modulation method, duty ratio, and current temperature of the power amplifier 100 of the input signal can be recognized as described above. Therefore, when the burst length is short, the coefficient selection unit 152 can specify the fixed distortion compensation coefficient to be used from among those in FIG. 4 using this recognized information.

[0042] An example of the operation of this distortion compensation device 1 will be illustrated below. The operation of this distortion compensation device 1 can be roughly classified into three types: (1) an operation of selecting and applying a distortion for compensating non-linear distortion to the operation signal by selecting a distortion compensation coefficient, (2) an operation of updating the updated distortion compensation coefficient and the burst length threshold value, and (3) an operation of setting and storing the fixed distortion compensation coefficient.

[0043] [[ID=十一]] First, (1) the operation of selecting and applying a distortion for compensating non-linear distortion to the operation signal by selecting a distortion compensation coefficient will be described. FIG. 5 is a flowchart showing this operation.

[0044] In FIG. 5, when an operating signal is input, the method selection unit 182 recognizes the input of the operating signal via the frame information control unit 181 (S11), and if this operating signal is a burst signal, recognizes the contents of the header H in FIG. 2 (band, frequency, modulation method, burst length, duty ratio) (S12), and in FIG. 3, recognizes the latest burst length threshold for the corresponding band, frequency, and modulation method, and determines whether this burst length threshold is larger or smaller than the recognized burst length (S13).

[0045] If the burst length is less than the burst length threshold (S13: Yes), the scheme selection unit 182 selects a scheme (fixed scheme) for performing distortion compensation using a fixed distortion compensation coefficient (S14). In this case, the coefficient selection unit 152 recognizes the temperature of the power amplifier 100 using the temperature notification unit 19 (S15), and reads out the duty ratio recognized (S12) from the header H in Fig. 4 and the fixed distortion compensation coefficient corresponding to this temperature (S16), and uses this for distortion compensation.

[0046] If the burst length is equal to or greater than the burst length threshold (S13: No), the scheme selection unit 182 selects a scheme (update scheme) for performing distortion compensation using an updated distortion compensation coefficient (S17). In this case, the coefficient selection unit 152 reads out the current updated distortion compensation coefficient (FIG. 3) corresponding to the previously used burst length threshold in FIG. 3 (S18) and uses this for distortion compensation.

[0047] Thereafter, the coefficient multiplier 151 applies distortion to this operation signal (S19) using either the read fixed distortion compensation coefficient (S16) or the read updated distortion compensation coefficient (S18). In this way, distortion compensation using the predistortion method is performed.

[0048] Next, (2) the operation of updating the distortion compensation coefficient and the burst length threshold will be described. As mentioned above, there are two ways to do this: using an operation signal or using a test signal. Figure 6 is a flowchart of the operation for updating the distortion compensation coefficient and the burst length threshold using an operation signal. In practice, these operations are performed one after the other when a single operation signal is input, but either may be performed before or after the other, or they may be performed in parallel to the extent possible.

[0049] 6, the distortion compensation calculation unit 15 recognizes that an operation signal (FF signal) has been input and that an FB signal has been emitted from the power amplifier 100 (S21), and the coefficient generation unit 153 recognizes the distortion compensation coefficients corresponding to the FF signal and FB signal and the accuracy of their calculation (estimated error) as described above. If the coefficient generation unit 153 recognizes that this accuracy is high as described above (S22: Yes), it updates the updated distortion compensation coefficients stored in the storage unit 20 corresponding to the conditions of this operation signal to the distortion compensation coefficients calculated here (S23). On the other hand, if the accuracy is recognized to be low (S22: No), the updated distortion compensation coefficients are not updated.

[0050] Thereafter, the scheme selection threshold generation unit 185 uses the frame information control unit 181 to recognize the burst length of this operational signal (S24), and the update impossibility detection unit 184 can recognize whether or not the updated distortion compensation coefficients were updated using this operational signal (whether or not the calculation of the distortion compensation coefficients was accurate). Here, the scheme selection threshold generation unit 185 can recognize whether the magnitude relationship between the recognized burst length and the current burst length threshold is consistent with the result of whether or not the updated distortion compensation coefficients could be updated (S25). That is, for example, if the burst length is equal to or greater than the burst length threshold but updating is not possible, or if the burst length is less than the burst length threshold but updating is possible, it can be recognized that the current burst length threshold is inappropriate, and the burst length threshold is updated based on this result (S26). This operation is similar to that described in Patent Document 3.

[0051] 7 is a flowchart of the operation for updating the distortion compensation coefficient and the burst length threshold using a test signal. This operation is performed when the wireless device is in a dormant state (when the operating signal is not amplified and output). Here, as with the case of the burst length threshold, the case where the user updates the distortion compensation coefficient is also described. Therefore, first, the user is inquired as to whether or not they will update the coefficient themselves (S31). If they will update the coefficient themselves (S31: Yes), the user performs the update in the same manner as in the previous case (S32). On the other hand, if the update is to be performed automatically using a test signal (S31: No), the user is inquired as to whether or not they will perform frame control of the test signal to be used (setting the mode of the test signal corresponding to the information included in the header H in FIG. 2) (S33). If the user will perform frame control (S33: Yes), the frame information control unit 181 determines frame conditions based on the conditions specified by the user (S34), and then causes the test signal generation unit 11 to generate a test signal including the header H and conforming to these conditions (S35). On the other hand, if the user does not perform frame control (S33: No), the frame information control unit 181 automatically performs this frame control according to predetermined conditions (S36), and then similarly causes the test signal generation unit 11 to generate a test signal (S35).

[0052] Thereafter, the test signal is used as an input signal to calculate distortion compensation coefficients, and the operation of determining whether or not to update the updated distortion compensation coefficients based on the accuracy of the calculated coefficients is the same as S22 in Fig. 6, and if updating is to be performed, the updated distortion compensation coefficients are updated in the same manner as described above (S32). This operation is also the same as that in Patent Document 3.

[0053] Next, (3) the operation of setting and storing a fixed distortion compensation coefficient will be described. The user can calculate a fixed distortion compensation coefficient for each condition as shown in FIG. 4 and store it in the storage unit 20. However, similar to the updated distortion compensation coefficient, a value calculated by the distortion compensation device 1 itself under certain conditions can also be stored as the fixed distortion compensation coefficient. This operation will be described here. This operation can be performed in the same way whether the input signal is an operational signal or a test signal. As mentioned above, if the burst length is short, the sample length used to calculate the distortion compensation coefficient using this signal will be short, which reduces the accuracy of the calculated updated distortion compensation coefficient, making it undesirable to use this signal. However, if an input signal with a short burst length is repeated at a constant duty ratio, the sample length will not necessarily be short, and in this case, the accuracy of calculating the distortion compensation coefficient using this input signal will not be reduced.

[0054] Therefore, for example, it is estimated that there will be no significant error even if the distortion compensation coefficient (updated distortion compensation coefficient) calculated when the burst length is equal to the burst length threshold in Fig. 3 is used as the distortion compensation coefficient when the burst length is shorter than this. In this case, by associating the duty ratio with the temperature of the power amplifier 100 when this input signal is actually input, the appropriateness of this distortion compensation coefficient in particular can be improved.

[0055] Fig. 8 is a flowchart showing this operation. Unlike the update of the updated distortion compensation coefficients (Fig. 6), this operation is performed as appropriate by user operation, particularly at the time of initial setup. Here, similar to the operation of Fig. 5, the method selection unit 182 recognizes the input of the input signal via the frame information control unit 181 (S41), and recognizes the contents of the header H in Fig. 2 (bandwidth, frequency, modulation method, burst length, duty ratio) (S42). Thereafter, the distortion compensation calculation unit 15 updates the updated distortion compensation coefficients using the operation of Fig. 6 (S43). That is, as a result, new updated distortion compensation coefficients are calculated using this input signal, and the burst length threshold is also updated in some cases.

[0056] Next, the scheme selection unit 182 determines (S44) whether the burst length recognized from the header H (S42) is equal to the burst length threshold value corresponding to other conditions in this header H in Fig. 3. Here, "equal" does not mean that they are strictly equal, but rather that the absolute value of the difference between the burst length and the burst length threshold value is within a certain range.

[0057] If the burst length is equal to the burst length threshold (S44: Yes), the coefficient generation unit 153 recognizes the temperature of the power amplifier 100 using the temperature notification unit 19 (S45), and sets the value of the fixed distortion compensation coefficient corresponding to the band, frequency, modulation method, burst length, and duty ratio of this input signal and this temperature in Fig. 4 as the updated value of the updated distortion compensation coefficient (S43) (S46). As a result, the value of the fixed distortion compensation coefficient is input.

[0058] In this way, particularly when the burst length is short, by selecting a fixed distortion compensation coefficient using the temperature and duty ratio of the power amplifier 100, more appropriate distortion compensation becomes possible. This will be described below.

[0059] When a burst signal is input to power amplifier 100, the temperature of power amplifier 100 rises regardless of the burst signal. However, if the burst length is sufficiently long (close to continuous wave), this temperature rise saturates and becomes constant after a certain time has passed, and it is thought that the nonlinear distortion condition remains constant, at least in this state. However, as mentioned above, this condition changes over the long term, so it is preferable to use updated distortion compensation coefficients that have been fed back with the latest condition, which allows for appropriate distortion compensation.

[0060] On the other hand, when the burst length is short, the temperature rises and ends before saturation, and the temperature drops during the subsequent non-transmission period. This operation is repeated each time a burst signal is input. Figure 9 is a diagram schematically showing temperature changes in power amplifier 100 for two cases where the burst length is short and the burst length is the same but the duty ratio is different. The horizontal axis represents the progression of time, the lower part represents the state of burst signal transmission, and the upper part represents the corresponding temperature change in power amplifier 100. In the lower part of each of Figures 9(a) and 9(b), the hatched areas represent the transmission period (burst signal), and the rest of the diagram represents the non-transmission period. Figure 9(a) shows the case where the duty ratio is approximately 0.1 (transmission period length:non-transmission period length≈1:9), and Figure 9(b) shows the case where the duty ratio is approximately 0.9 (transmission period length:non-transmission period length≈9:1).

[0061] When the duty ratio is small as shown in Figure 9(a), when a burst signal is input, the temperature gradually rises from the start of one burst signal (transmission period) using the initial temperature (room temperature) as a reference, reaches a maximum (maximum) at the end of the burst signal, and then gradually drops back down to the initial temperature. The maximum temperature is lower than the maximum temperature (saturation value) in the case of a continuous wave as described above. This operation is repeated each time a burst signal is input, and the temperature changes similarly when the next burst signal is input. Therefore, the temperature of power amplifier 100 over time in this case is as shown in the upper part of Figure 9(a), and the temperature also changes periodically, just like the period of the burst signal.

[0062] On the other hand, when the duty ratio is large as shown in FIG. 9(b), the temperature rises in response to one burst signal, reaches a maximum at the end of the burst signal, and then drops, just as in FIG. 9(a). However, before the temperature has fully dropped to the initial temperature after the end of the burst signal, the next burst signal is input, causing the temperature to rise again. This cycle is repeated. Therefore, the temperature at the start of the next burst signal is higher than the temperature at the start of the previous burst signal, and the temperature thereafter is also higher than the corresponding point in the previous burst signal. In FIG. 9(b), the temperature at the end of the first cycle (the start of the second cycle) is Ta, and the temperature at the end of the last cycle is Tb, so that Tb > Ta. This cycle is repeated each time a burst signal is input. Therefore, the maximum temperature corresponding to each burst signal gradually increases with each input of a burst signal. In this case, the temperature of the power amplifier 100 over time is as shown in the upper part of FIG. 9(b). When the number of cycles of this operation becomes sufficiently large (outside the range shown on the right side of FIG. 9(b)), the temperature rise saturates, just as in the case of continuous waves.

[0063] For this reason, when the burst length is short and the temperature rise during one transmission interval is small, the temperature rise status of power amplifier 100 varies greatly depending on the duty ratio even if the burst length is the same. As mentioned above, there is no one-to-one correspondence between the temperature recognized by temperature notification unit 19 and the status of nonlinear distortion, but there is clearly a correlation between this temperature and the status of nonlinear distortion. Therefore, by combining this temperature and duty ratio and using it as a guideline for selecting a distortion compensation coefficient, more appropriate distortion compensation becomes possible.

[0064] 10 is a diagram that schematically illustrates the temperature change state shown in FIG. 9 within one cycle, enlarged over time. In FIG. 10, the lower part shows the time progression of the burst signal, as in FIG. 9, the middle part shows the corresponding frame structure (FIG. 2), and the upper part shows a schematic representation of the temperature change in this case. As mentioned above, in this frame structure, the transmission frame type signal (0 to 3) is recognized as shown in FIG. 10.

[0065] In FIG. 10, in the section of data D (transmission frame type signal = 3: data section) where information that should originally be transmitted is transferred, distortion compensation must be performed with particularly high precision compared to the section of preamble P and header H (transmission frame type signal = 1, 2: non-data section) which constitute the input signal but are different from data D. On the other hand, the temperature output from temperature notification unit 19 (temperature calculation unit 192) does not necessarily correspond to the temperature at any point in this transmission frame, particularly when filtering processing or the like is performed on the temperature measurement results in temperature calculation unit 192, and is not necessarily the temperature in the data section.

[0066] In this case, if the temperature change characteristics shown in FIG. 10 are stored in advance in the storage unit 20 for each burst length and duty ratio, the temperature calculation unit 192 can recognize these characteristics. In this case, if it is recognized that the calculated temperature is for the non-data section, it is possible to estimate the temperature for the subsequent data section (>temperature for the section of preamble P) from this temperature, and the temperature estimated in this manner can be used by the coefficient selection unit 152 in selecting the distortion compensation coefficient (S15 in FIG. 5). This makes it possible to perform distortion compensation particularly in the data section with high accuracy. In other words, by recognizing the frame structure and temperature, it is possible to perform distortion compensation particularly for the data D signal with high accuracy.

[0067] In the above example, two types of distortion compensation coefficients are stored in the storage unit 20: an updated distortion compensation coefficient used when the burst length is long, and a fixed distortion compensation coefficient used when the burst length is short. Only the fixed distortion compensation coefficients are set for each duty ratio and temperature of the power amplifier 100. However, the updated distortion compensation coefficients may also be set for each duty ratio and its temperature. In this case, the accuracy of distortion compensation can be further improved.

[0068] The present invention has been described above based on an embodiment. This embodiment is merely an example, and it will be understood by those skilled in the art that various modifications are possible in the combination of the respective components, and that such modifications are also within the scope of the present invention. [Explanation of symbols]

[0069] 1 Distortion compensation circuit 11 Test signal generation unit 12 Selectors 13 Peak power suppression section 14 Transmission power control section 15 Distortion compensation calculation section 16 Offset component correction section 17 Data capture control section 18 Method control unit (control unit) 19 Temperature notification section 20 Memory section 100 Power Amplifier 151 Coefficient multiplication unit (distortion unit) 152 Coefficient selection section 153 Coefficient Generation Unit 171 FF signal buffer section 172 FB signal buffer section 181 Frame information control section 182 Method selection section 183 Coefficient update decision unit 184 Update Unavailable Detector 185 Method selection threshold generation unit 186 Threshold update determination unit 191 Temperature detection unit 192 Temperature calculation section D Data H Header P Preamble

Claims

1. A pre-distortion distortion compensation circuit that performs distortion compensation to impart to an input signal a distortion that compensates for nonlinear distortion, which is a difference between a waveform of an output signal obtained by amplifying an input signal by a power amplifier and a waveform of the input signal, and the type of the distortion is set by a distortion compensation coefficient selected in accordance with the type of the input signal, the input signal is constituted by intermittently repeating a burst signal; a storage unit that stores a plurality of distortion compensation coefficients that are set according to the state of the input signal; a temperature notification unit that detects the temperature of the power amplifier; a control unit that recognizes a state of the input signal and selects one of the plurality of distortion compensation coefficients stored in the storage unit; and a distortion imparting unit that imparts the distortion corresponding to the distortion compensation coefficient selected by the control unit to the input signal; Equipped with the distortion compensation coefficient is set in accordance with set conditions, which are the bandwidth, frequency, modulation method, burst length, and duty ratio of the burst signal, and temperature, and two types of distortion compensation coefficients, an updated distortion compensation coefficient and a fixed distortion compensation coefficient, are set in accordance with the burst length of the input signal; the storage unit stores, for each of the setting conditions, a plurality of the updated distortion compensation coefficients, a plurality of the fixed distortion compensation coefficients, and a burst length threshold value, which is the burst length that serves as a threshold value for determining whether the updated distortion compensation coefficients or the fixed distortion compensation coefficients are to be selected as the distortion compensation coefficient; the fixed distortion compensation coefficient is set in accordance with the setting condition of the input signal; the control unit, when the burst length of the input signal is equal to or greater than the burst length threshold corresponding to the setting condition, selects the updated distortion compensation coefficient corresponding to the burst length as the distortion compensation coefficient to be used in the distortion imparting unit, and when the burst length of the input signal is less than the burst length threshold corresponding to the setting condition, selects the fixed distortion compensation coefficient according to the duty ratio and the temperature of the input signal as the distortion compensation coefficient to be used in the distortion imparting unit; a frame structure of the burst signal includes a data section in which data to be transmitted by the input signal and the output signal is transferred, and a non-data section in which information other than the data is transferred in association with the data transmission; a control unit that, when the time when the temperature of the power amplifier is recognized corresponds to the non-data section of the burst signal, corrects the temperature to a value in the data section and uses the corrected temperature when selecting the fixed distortion compensation coefficient.

2. a coefficient generation unit that recognizes the burst length of the input signal, and calculates a distortion compensation coefficient to be used for the distortion compensation and the calculation accuracy of the distortion compensation coefficient by statistical processing by comparing the input signal with the corresponding output signal; The control unit If the accuracy is recognized to be high, the calculated distortion compensation coefficient is associated with the burst length and updated as the updated distortion compensation coefficient in the storage unit; 2. The distortion compensation circuit according to claim 1, wherein the burst length threshold stored in the storage unit is updated in accordance with the accuracy when the distortion compensation coefficient is calculated in the coefficient generation unit in accordance with the input signal and the burst length of the input signal.

3. 3. A radio device comprising: the distortion compensation circuit according to claim 1; and the power amplifier.

Images