Digital predistortion method and digital predistortion system

By obtaining historical and real-time signals in mobile terminal devices, calculating model coefficients and generating predistorted signals, the linearity and efficiency problems caused by impedance mismatch in mobile terminal devices are solved, and the tracking and correction of digital predistortion system with low complexity is achieved , reducing calculation overhead.

CN115102509BActive Publication Date: 2024-11-12TSINGHUA UNIVERSITY +1
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202210200744.2
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2022-03-02
Publication Date
2024-11-12
Estimated Expiration
2042-03-02

AI Technical Summary

Technical Problem

Due to impedance mismatch problems in mobile terminal devices, the linearity and efficiency of the digital predistortion system are affected. Due to the high cost and large volume of the isolator, it is impossible to set up an isolator in mobile terminal devices, and the prior art is difficult to effectively solve this problem. question.

Method used

By obtaining historical signals and real-time signals, calculating fixed model coefficients and real-time reflection coefficients, generating predistorted signals, and using predistorters to correct them based on real-time model coefficients, it is suitable for real-time changes in the amplifier load in terminal equipment, reducing the system's calculation overhead.

Benefits of technology

Low-complexity tracking and correction of digital predistortion systems is realized, frequent updates are avoided, computing overhead is significantly reduced, and system adaptability and efficiency are improved.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN115102509B_ABST
    Figure CN115102509B_ABST
Patent Text Reader

Abstract

The present invention provides a digital predistortion method, comprising: obtaining at least one set of historical signals and one set of real-time signals; calculating fixed model coefficients according to at least one set of historical signals; calculating real-time reflection coefficients according to one set of real-time signals; and generating predistortion signals according to the fixed model coefficients and the real-time reflection coefficients. The present invention also provides a digital predistortion system. The digital predistortion method proposed in the present invention is applicable to terminal equipment, and can track and adjust the model coefficients according to the real-time changes of the power amplifier load in the terminal equipment, thereby avoiding frequent updates of the digital predistortion system and greatly reducing the calculation overhead of the digital predistortion system.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The invention relates to mobile communication technology, and in particular to a data predistortion method and a digital predistortion system. Background Art

[0002] Digital pre-distortion (DPD) is a mainstream power amplifier linearization technology in today's communication systems. Digital pre-distortion can improve the linearity of the power amplifier while ensuring the efficiency of the power amplifier. However, when there is an impedance mismatch between the power amplifier and the load device in the digital pre-distortion system, it will have a negative impact on the linearity of the power amplifier.

[0003] At present, the research on digital pre-distortion technology is mainly focused on the background of base stations. In base station equipment, an isolator is set between the power amplifier and the load device to stabilize the output impedance of the power amplifier, so as to achieve the effect of matching the load impedance and output impedance of the power amplifier. Under the condition of impedance matching, the power amplifier can achieve maximum power transmission.

[0004] Impedance mismatch is a common phenomenon in mobile terminal devices due to the multi-lateral nature of their working environment. However, due to the high cost and large size of isolators, isolators cannot be installed in mobile terminal devices. Therefore, research on digital pre-distortion systems suitable for mobile terminal devices is an urgent problem to be solved. Summary of the invention

[0005] In view of this, in order to overcome at least one aspect of the above-mentioned problems, the present invention provides a digital predistortion method, characterized in that it includes: acquiring at least one group of historical signals and one group of real-time signals; calculating fixed model coefficients according to the at least one group of historical signals; calculating real-time reflection coefficients according to the group of real-time signals; and generating a predistortion signal according to the fixed model coefficients and the real-time reflection coefficients; wherein the at least one group of historical signals includes at least one first input signal and at least one first output signal and at least one first reflection signal respectively corresponding to the at least one first input signal, and the group of real-time signals includes a second output signal and a second reflection signal corresponding to the second output signal.

[0006] According to an embodiment of the present invention, obtaining at least one group of historical signals and one group of real-time signals includes: obtaining the at least one group of historical signals at a first preset frequency, and obtaining the group of real-time signals at a second preset frequency; wherein the first preset frequency is less than the second preset frequency.

[0007] According to an embodiment of the present invention, the at least one group of historical signals includes multiple input signals, multiple output signals and multiple first reflection signals under multiple load conditions; the acquisition of at least one group of historical signals and one group of real-time signals includes: acquiring multiple first input signals and multiple first output signals; calculating the variation range of the load conditions based on the multiple first input signals and the multiple first output signals; and acquiring the multiple first reflection signals, the second output signal and the second reflection signal when the variation range is greater than a preset range.

[0008] According to an embodiment of the present invention, the at least one group of historical signals includes multiple input signals, multiple output signals and multiple first reflection signals under multiple load conditions; calculating the fixed model coefficients based on the at least one group of historical signals includes: calculating multiple historical reflection coefficients based on the multiple first reflection signals and the multiple first output signals, respectively; and calculating the fixed model coefficients based on the multiple first input signals, the multiple first output signals and the multiple historical reflection coefficients.

[0009] According to an embodiment of the present invention, generating a predistortion signal according to the fixed model coefficient and the real-time reflection coefficient includes: calculating a correction coefficient according to the historical reflection coefficient and the real-time reflection coefficient; correcting the fixed model coefficient according to the correction coefficient to obtain a real-time model coefficient; and generating a predistortion signal according to the real-time model coefficient through a predistorter.

[0010] According to an embodiment of the present invention, the method also includes: obtaining multiple initial input signals, multiple initial output signals and multiple initial reflection signals under multiple initial load conditions; calculating multiple initial reflection coefficients based on the multiple initial output signals and the multiple initial reflection signals; obtaining initial model coefficients based on the multiple initial input signals, the multiple initial output signals and the multiple initial reflection coefficients; and generating an initial predistortion signal based on the initial model coefficients through a predistorter.

[0011] According to an embodiment of the present invention, before correcting the fixed model coefficient according to the correction coefficient to obtain the real-time model coefficient, the method further includes: updating the initial model coefficient to the fixed model coefficient.

[0012] The present invention also provides a digital predistortion system, characterized in that it also includes: a first calculation module, used to obtain at least one group of historical signals, and calculate fixed model coefficients based on the at least one group of historical signals; a second calculation module, used to obtain a group of real-time signals, and calculate real-time reflection coefficients based on the group of real-time signals; a predistorter, used to generate a predistortion system signal based on the fixed model coefficients and the real-time reflection coefficients; wherein the at least one group of historical signals includes at least one first input signal and at least one first output signal and at least one first reflection signal respectively corresponding to the at least one first input signal, and the group of real-time signals includes a second output signal and a second reflection signal corresponding to the second output signal.

[0013] According to an embodiment of the present invention, the predistortion system also includes: an adjustable load, used to generate multiple initial loads; an initialization module, used to obtain multiple initial input signals, multiple initial output signals and multiple initial reflection signals under the multiple initial loads; calculate multiple initial reflection coefficients based on the multiple initial output signals and the multiple initial reflection signals; and obtain initial model coefficients based on the multiple initial input signals, the multiple initial output signals and the multiple initial reflection coefficients.

[0014] According to an embodiment of the present invention, the at least one group of historical signals includes multiple input signals, multiple output signals and multiple first reflection signals under multiple load conditions; the first calculation module is used to obtain multiple first input signals and multiple first output signals; based on the multiple first input signals and multiple first output signals, the change range of the load conditions is calculated; when the change range is greater than a preset range, the multiple first reflection signals are obtained; and the second calculation module is used to obtain the second output signal and the second reflection signal when the change range is greater than a preset range.

[0015] Compared with the prior art, the digital predistortion method proposed in the present invention is suitable for terminal equipment and can track and adjust the model coefficients according to the real-time changes of the power amplifier load in the terminal equipment, thereby avoiding frequent updates of the digital predistortion system and greatly reducing the calculation overhead of the digital predistortion system. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] For a more complete understanding of the present invention and its advantages, reference will now be made to the following description taken in conjunction with the accompanying drawings, in which:

[0017] Figure 1 A schematic diagram schematically shows load impedance matching and mismatching according to an embodiment of the present invention;

[0018] Figure 2 A flowchart of a digital predistortion method according to an embodiment of the present invention is schematically shown;

[0019] Figure 3 A schematic diagram of a digital predistortion initialization calibration system according to an embodiment of the present invention is schematically shown;

[0020] Figure 4 A schematic diagram of a digital predistortion system according to an embodiment of the present invention is schematically shown;

[0021] Figure 5 A schematic diagram of a model parameter extraction module according to an embodiment of the present invention is schematically shown;

[0022] Figure 6 A schematic diagram schematically shows a predistorter according to an embodiment of the present invention; and

[0023] Figure 7 A schematic diagram of a digital predistortion system according to another embodiment of the present invention is schematically shown. DETAILED DESCRIPTION

[0024] In order to more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the accompanying drawings required for use in the embodiments or the prior art descriptions are briefly introduced below. Obviously, these descriptions are exemplary only and are not intended to limit the scope of the present invention. For those skilled in the art, other accompanying drawings can also be obtained based on these accompanying drawings without paying creative work. In the following detailed description, for ease of explanation, many specific details are described to provide a comprehensive understanding of the embodiments of the present invention. Unless otherwise defined, the technical terms or scientific terms used in the present invention should be the common meanings understood by people with general skills in the field to which the present invention belongs.

[0025] The terms used herein are only for describing specific embodiments and are not intended to limit the present invention. The terms "comprise", "include", etc. used herein indicate the existence of the features, steps, operations and / or components, but do not exclude the existence or addition of one or more other features, steps, operations or components.

[0026] The power amplifier (PA) is one of the most energy-consuming devices in the transmitter. Its power consumption, efficiency and other indicators directly determine the performance of the transmitter. In order to pursue high efficiency, the working point of the PA needs to be as close to the saturation region as possible. At this time, the gain compression of the PA is more serious and the nonlinear distortion is very significant. Digital pre-distortion technology (DPD), as a mainstream PA linearization technology, can improve the linearity of the PA while ensuring its efficiency.

[0027] A digital predistortion system usually includes two parts: a transmitting channel and a feedback channel.

[0028] In the transmission channel, the predistorter receives the input signal and generates a predistortion signal based on the predistortion model coefficients. The predistortion signal is converted into an RF signal through a digital-to-analog converter and an up-converter, and after being amplified by a power amplifier, the signal is transmitted using an antenna (load). When the digital predistortion system is working, the model coefficients stored in the predistorter need to be updated at regular intervals to ensure the linearization accuracy of the predistorter.

[0029] In the feedback channel, the model parameter extraction module receives the output signal and input signal after time delay and phase alignment. The model parameter extraction module has the function of forward and reverse behavior modeling of the power amplifier, and can extract model coefficients according to the input and output signals. The extracted reverse behavior model coefficients can be used to update the model coefficients stored in the predistorter.

[0030] In actual use, the model parameter extraction module is usually started at regular intervals or when the linearization performance of the predistorter degrades. However, since the model coefficient extraction algorithm is generally complex, the model parameter extraction module consumes a lot of computing resources when extracting the model coefficients.

[0031] During the use of terminal equipment, the pre-distortion model will often have impedance mismatch problems. Under impedance matching conditions, the load antenna can absorb all the output power of the power amplifier device, and the power amplifier device does not generate reflected echoes. In the case of load impedance mismatch, the power amplifier device will generate reflected echoes. Reflected echoes will affect the efficiency, power, linearity and other characteristics of the amplifier.

[0032] Figure 1 The schematic diagram of load impedance matching and mismatching according to an embodiment of the present invention is schematically shown. Figure 1 As shown, the input signal a is input to the digital predistortion system to generate the output signal b2. When the output impedance of the power amplifier is mismatched with the load impedance, there is a reflection echo a2 at the output of the power amplifier. Due to transmission line faults, coupling effects between adjacent antennas or devices, the antenna characteristics will change, and impedance mismatch may occur.

[0033] Impedance mismatch in the PA reduces the output power and overall radiation efficiency of the loaded antenna, thus reducing the coverage range of the device. Secondly, the PA operating point will also change and may cause additional nonlinear distortion, thus reducing the throughput of the device.

[0034] In addition, impedance mismatch also has a negative impact on digital predistortion technology and affects the linearity of the power amplifier. Once a predistorter that works normally under impedance matching encounters an impedance mismatch, the performance of the power amplifier equipment will be seriously degraded. In the case of impedance mismatch, the performance of the predistorter is also sensitive to changes in the load impedance of the power amplifier. For example, a predistorter obtained through adaptive training under a certain load condition cannot be directly applied to another load condition, and the predistorter must be adaptively trained again.

[0035] However, in the actual use of terminal equipment, impedance mismatch problems are unavoidable, especially in mobile terminal equipment. The working environment of mobile terminal equipment is complex and the load conditions change frequently. If the pre-distortion model is frequently updated according to the load changes, it will generate huge computational overhead.

[0036] Therefore, the present invention provides a low-complexity digital predistortion method, which can track and correct the predistortion model coefficients according to the load changes of the terminal equipment and reduce the power consumption overhead generated during the correction process.

[0037] Figure 2 The flowchart of the correction method of the digital pre-distortion system according to the embodiment of the present invention is schematically shown.

[0038] like Figure 2 As shown, the correction method of the digital predistortion system includes at least the following steps:

[0039] Step S210, obtaining at least one set of historical signals and one set of real-time signals.

[0040] A group of historical signals includes a first input signal, a first output signal and a first reflected signal. The first input signal, the first output signal and the first reflected signal included in each group of historical signals have a corresponding relationship. For example, at a certain moment, an input signal is sent to the digital predistortion system to generate an output signal corresponding to the input signal and a reflected signal corresponding to the input signal.

[0041] A set of real-time signals includes a second output signal and a second reflected signal. The second output signal and the second reflected signal have a corresponding relationship, for example, an output signal at a current moment and a reflected signal corresponding to the output signal.

[0042] The reflected signal is the reflected echo generated at the output end of the power amplifier.

[0043] Step S220, calculating fixed model coefficients according to at least one set of historical signals.

[0044] In the embodiment of the present invention, the fixed model coefficients can be calculated based on the input, output and reflection signals of a set of historical signals, and the fixed model coefficients can also be calculated based on multiple input, output and reflection signals of multiple sets of historical signals. The fixed model coefficients calculated based on multiple sets of historical signals can adapt to more load conditions.

[0045] Step S230, calculating a real-time reflection coefficient according to a set of real-time signals.

[0046] In the embodiment of the present invention, a real-time reflection coefficient is calculated according to the real-time output signal and the real-time reflection signal. The real-time reflection coefficient is used to update the model coefficient in real time.

[0047] Step S240: generating a predistortion signal according to the fixed model coefficients and the real-time reflection coefficients.

[0048] In the embodiment of the present invention, step S240, generating a predistortion signal according to the fixed model coefficient and the real-time reflection coefficient, includes: calculating a correction coefficient according to the historical reflection coefficient and the real-time reflection coefficient; correcting the fixed model coefficient according to the correction coefficient to obtain the real-time model coefficient; and generating a predistortion signal according to the real-time model coefficient by a predistorter. The historical reflection coefficient is calculated according to the historical input and output signals.

[0049] The historical reflection coefficient is calculated based on the multiple sets of historical signals obtained, and the real-time reflection coefficient is calculated based on a set of real-time signals at the current moment. The correction coefficient can be calculated based on the ratio of the historical reflection coefficient to the real-time reflection coefficient. The change in reflection coefficient can be a manifestation of load change. The correction coefficient is determined based on the real-time reflection coefficient at the current moment and the historical reflection coefficient at the previous moment, and then the fixed model coefficient is corrected based on the correction coefficient, thereby coping with frequent changes in load conditions.

[0050] After the input signal is input into the predistortion system, a predistortion signal is generated according to the corrected model coefficients.

[0051] The calculation process of calculating the fixed model coefficients according to the input, output and reflection signals is complicated. Frequent calculation of the fixed model coefficients to update the original model coefficients will generate a large calculation overhead. Compared with calculating the fixed model coefficients, the process of calculating the reflection coefficient according to the output signal and the reflection signal is faster and simpler, so the original model coefficients can be corrected according to the fixed model coefficients and the reflection coefficients at different frequencies.

[0052] As an optional embodiment, step S210, acquiring at least one set of historical signals and one set of real-time signals, includes acquiring at least one set of historical signals at a first preset frequency, and acquiring one set of real-time signals at a second preset frequency, wherein the first preset frequency is less than the second preset frequency.

[0053] In an embodiment of the present invention, a time period can be set, multiple groups of historical signals within the time period are obtained, a fixed model coefficient is calculated based on the multiple groups of historical signals, and a predistortion signal is generated based on the fixed model coefficient. Since a fixed coefficient is obtained by calculation after a time period ends, the model coefficient will remain unchanged in the next time period. In order to achieve real-time tracking and correction of the model coefficient, a real-time reflection coefficient is calculated through the real-time output signal and the real-time reflection signal within the time period, and then the fixed model coefficient is corrected in real time based on the real-time reflection coefficient.

[0054] For example, the first preset frequency is set to 1min / time, multiple groups of historical signals in the previous minute are counted, and a fixed model coefficient is calculated every minute; the second preset frequency is set to 10s / time, the real-time output signal and reflection signal of the 10th second are obtained, and a real-time reflection coefficient is calculated every 10 seconds. The fixed model coefficient is calculated based on multiple groups of historical signals in the previous minute. In the current minute, a real-time reflection coefficient is calculated every 10 seconds. The fixed model coefficient is corrected in real time at a frequency of 10s / time.

[0055] After the current minute is over, new fixed model coefficients can be calculated based on the input, output and reflection signals within the current minute, and the new fixed model coefficients replace the original fixed model coefficients. In the next minute, the new fixed model coefficients are tracked and corrected based on the real-time reflection coefficients.

[0056] The above preset frequencies are all for exemplary purposes only and the present invention does not limit the preset frequencies.

[0057] Since the process of calculating the fixed model coefficients is complicated, the fixed model coefficients are tracked and corrected through the real-time calibration coefficients obtained in real time to ensure the impedance matching of the digital pre-distortion system. In addition, the model coefficients are updated through multiple sets of historical signals at a relatively low frequency to ensure that the model coefficients have a certain adaptability even when the load changes greatly.

[0058] Illustratively, the present invention also provides another digital predistortion method.

[0059] In the case where at least one group of historical signals includes multiple input signals, multiple output signals and multiple first reflection signals under multiple load conditions, step S110 of acquiring at least one group of historical signals and one group of real-time signals includes: acquiring multiple first input signals and multiple first output signals; calculating the variation range of the load conditions based on the multiple first input signals and the multiple first output signals; and in the case where the variation range is greater than a preset range, acquiring multiple first reflection signals, second output signals and second reflection signals.

[0060] Under the premise of ensuring the impedance matching of the digital predistortion system, the power consumption generated during the correction of the model coefficients can be further reduced. The correction method of the model coefficients can be selected according to the degree of load change to reduce the calculation overhead while ensuring the linearization of the predistortion signal.

[0061] Based on obtaining multiple input and output signals under multiple load conditions, the real-time load change is calculated. When the load change range is less than the preset range, the model coefficients may not be corrected. Exemplarily, the model coefficients of the predistortion system are fixed model coefficients updated in the previous time period. Since the fixed model coefficients are calculated from multiple groups of historical signals, the fixed model coefficients have a certain adaptability to load changes within a certain range. In this case, the predistortion signal generated can still remain linear. Therefore, when the load change range is less than the preset range, there is no need to correct the historical model coefficients.

[0062] When the load variation range is greater than the preset range, the model coefficient needs to be corrected. When the load variation range is greater than the preset range, serious impedance mismatch may occur, and the original fixed model coefficient cannot meet the current load variation, which will seriously affect the efficiency of the power amplifier. Therefore, it is necessary to obtain multiple sets of historical signals and real-time signals, calculate the fixed model coefficients and real-time reflection coefficients respectively, so as to correct the original model coefficients and ensure the linearization of the pre-distorted signal.

[0063] It should be noted that the input, output and reflection signals at each moment are stored in the memory. In different situations, the required data can be obtained for calculation. Therefore, when the load variation range is calculated based on the input and output signals under multiple load conditions, the corresponding reflection signals are also stored in the memory. When the model coefficients need to be updated, the corresponding reflection signals can be obtained for calculating the fixed model coefficients and the historical reflection coefficients.

[0064] The present invention also discloses another embodiment of the digital predistortion method. The digital predistortion method also includes an initialization calibration method. The initialization calibration method includes: obtaining multiple initial input signals, multiple initial output signals, and multiple initial reflection signals under multiple initial load conditions; calculating multiple initial reflection coefficients according to the multiple initial output signals and the multiple initial reflection signals; obtaining initial model coefficients according to the multiple initial input signals, the multiple initial output signals, and the multiple initial reflection coefficients; generating an initial predistortion signal according to the initial model coefficients through a predistorter. Subsequently, a real-time reflection coefficient is calculated according to the real-time output signal and the reflection signal. Then, the initial model coefficient can be tracked and corrected according to the real-time reflection coefficient.

[0065] In an embodiment of the present invention, initialization calibration is required before the digital predistortion system starts working. During the initial calibration, multiple power amplifier states under multiple power amplifier loads are obtained. First, multiple initial reflection coefficients under multiple power amplifier states are calculated based on multiple output signals and multiple reflection signals under multiple power amplifier states; then, initial model coefficients are obtained based on multiple initial input signals, multiple initial output signals and multiple initial reflection coefficients. After obtaining the initial model coefficients, the digital predistortion system is turned on to work in real time.

[0066] In the case where the load variation range is too large, the initial model coefficient needs to be replaced by a new fixed model coefficient, and then the new model coefficient is tracked and corrected. Step S240, generating a predistortion signal according to the fixed model coefficient and the real-time reflection coefficient, including: updating the initial model coefficient to the fixed model coefficient; calculating the correction coefficient according to the historical reflection coefficient and the real-time reflection coefficient; correcting the fixed model coefficient according to the correction coefficient to obtain the real-time model coefficient; and generating a predistortion signal according to the real-time model coefficient through a predistorter.

[0067] The initial model coefficients calculated from the signals in multiple power amplifier states are adaptable to load changes within a certain range. When the digital pre-distortion system is working, the initial model coefficients can be tracked and corrected according to the reflection coefficients obtained from the real-time signal. The model coefficients are updated according to the reflection coefficients obtained from the real-time signal, and the tracking and correction of the model coefficients are achieved while ensuring the reduction of the power consumption required for the update.

[0068] Figure 3 A schematic diagram of a digital pre-distortion initialization calibration system according to an embodiment of the present invention is schematically shown.

[0069] like Figure 3 As shown, the present invention provides a digital predistortion initialization calibration system 300, including a transmission channel and a feedback channel.

[0070] The transmission channel includes a digital-to-analog converter, an up-converter, a power amplifier, and an adjustable load. The feedback channel includes a down-converter, a filter, an analog-to-digital converter, and an initialization module.

[0071] Adjustable load, used to generate multiple initial loads. The load generated by the adjustable load is controlled by the system.

[0072] An initialization module is used to obtain multiple initial input signals, multiple initial output signals and multiple initial reflection signals under multiple initial load conditions; calculate multiple initial reflection coefficients based on the multiple initial output signals and the multiple initial reflection signals; and obtain initial model coefficients based on the multiple initial input signals, the multiple initial output signals and the multiple initial reflection coefficients.

[0073] It should be noted that the digital predistortion initialization calibration system in the embodiment of the present invention corresponds to the initialization calibration method of the digital predistortion system in the embodiment of the present invention. The description of the digital predistortion initialization calibration system specifically refers to the initialization calibration method of the digital predistortion system, which will not be repeated here.

[0074] Figure 4 The figure schematically shows a digital predistortion system according to an embodiment of the present invention.

[0075] like Figure 4 As shown, the present invention provides a digital predistortion system 400, including a transmitting channel, a feedback channel and a reflection coefficient detection module.

[0076] The transmission channel includes a predistorter, a digital-to-analog converter, an upconverter, a power amplifier, a coupler, and a load antenna. The predistorter receives the input signal and generates a predistorted signal. The predistorted signal is converted into an RF signal through the digital-to-analog converter and the upconverter. The power amplifier amplifies the RF signal, and the amplified RF signal is transmitted through the load antenna.

[0077] The feedback channel includes a down converter, a filter, an analog-to-digital converter and a model parameter extraction module. The model parameter extraction module receives the output signal through the down converter, the filter and the analog-to-digital converter.

[0078] The reflection coefficient detection module receives the output signal and the reflection signal, calculates the reflection coefficient according to the output signal and the reflection signal, and sends the reflection coefficient to the predistorter and the model parameter extraction module. The calculated reflection coefficient includes amplitude and phase.

[0079] For example, the reflection coefficient can be calculated by voltage standing wave ratio and radio frequency return loss. The reflection coefficient can also be directly detected by a specific hardware device.

[0080] The reflection coefficient is related to the load impedance. When the load impedance changes, the reflection coefficient changes accordingly. By tracking and correcting the model coefficient according to the reflection coefficient, the pre-distortion model coefficient can be tracked and corrected according to the load impedance change, thereby reducing the update frequency of the pre-distortion model coefficient through adaptive training.

[0081] The principle of correcting the predistortion model coefficients according to load impedance changes is as follows.

[0082] In the case of load impedance mismatch, the signal flows as follows Figure 4As shown, there is a reverse echo signal (reflected signal) a2 at the output end of the power amplifier. The input signal a1 and the reverse echo signal (reflected signal) a2 of the input predistorter and the model parameter extraction module will affect the nonlinear distortion of the power amplifier. The output signal b2 under load mismatch can be recorded as a function represented by the input signal a1 and the echo signal a2, b2(a1, a2).

[0083] The influence of the echo signal a2 needs to be considered when modeling the power amplifier behavior and the predistorter. According to the PHD (Poly-harmonic distortion) model, the output signal b2 is converted into a time domain form, and the power amplifier output signal is expressed as formula (1):

[0084]

[0085] in, is the conjugate of the echo signal a2; | a1| is the envelope of the input signal; is the model basis function; c1(k), c2(k) and c3(k) are the model coefficients; k is the index value of the model order, k = -1, 2, ..., K, K is a positive integer.

[0086] The echo signal a2 can be simplified as a function of the input signal a1 and can be obtained from the original input signal a1.

[0087] Normally, the amplitude of the input signal a1 is much larger than the amplitude of the echo signal a2. According to the superposition of harmonics, the total deviation of the output signal b2 can be considered to be the independent superposition of the deviations caused by all components. Assuming that the input signal a1 remains unchanged, when the load of the power amplifier changes, the echo signal a2 changes, which causes the output signal b2 of the power amplifier to change. Therefore, it can be considered that when the load changes, only the coefficients c2(k) and c3(k) in the power amplifier behavior model (1) change, and the coefficient c1(k) remains unchanged.

[0088] Assume that equation (1) represents the output signal of the power amplifier when the load impedance is Z1, and the corresponding reflection coefficient is Γ1. When the load impedance changes to Z2, the corresponding reflection coefficient is Γ γ , the amplifier output signal changes to:

[0089]

[0090]

[0091] Among them, the mark γ is used to indicate the parameters when the load impedance is Z2, and γ has no practical meaning; Represents the output signal of the power amplifier with a load impedance of Z2; Represents the echo signal with load impedance Z2; Echo signal The conjugation of.

[0092] consider as well as Substituting equations (1) and (2) into Γ1 and Γ γ The expression of and simplifying it, the echo signal It can be expressed as the echo signal a2 and the reflection coefficients Γ1, Γ γ The function represented by:

[0093]

[0094] Among them, l() represents a2, Γ1 and Γ γ The mapping relationship between them.

[0095] Substituting equation (3) into equation (2), the output signal of the power amplifier when the load impedance is Z2 is Expressed as

[0096]

[0097] in, is the conjugate of Γ1, is Γ γ The conjugation of.

[0098] Formula (4) can be approximated as

[0099]

[0100] in,

[0101]

[0102] in, and represents the recombinant model coefficient when the load impedance is Z2. According to the harmonic superposition assumption, since the change caused by the change of a2 is small, it can be assumed that as well as Model coefficients for load impedance Z2 The model coefficients c1(k), c2(k), and c3(k) with load impedance Z1 are related as follows:

[0103]

[0104] Therefore, when the power amplifier load changes from Z1 to Z2, it is only necessary to multiply the existing model coefficients c1(k), c2(k), and c3(k) by the corresponding adjustment factor g(Γ1, Γ γ )or In order to improve the accuracy of the model, the input and output signals of the power amplifier under different load conditions can be used to solve the better model coefficients c1(k), c2(k) and c3(k).

[0105] Figure 5 The figure schematically shows a model parameter extraction module according to an embodiment of the present invention.

[0106] like Figure 5 As shown, since the model parameter extraction module needs to use a complex model coefficient algorithm to calculate the model coefficients according to the input and output signals each time, in order to reduce the extraction frequency of the fixed model coefficients (c1(k), c2(k) and c3(k)) as much as possible and improve the accuracy of the fixed model coefficients, the model parameter extraction module can calculate the fixed model coefficients (c1(k), c2(k) and c3(k)) according to the power amplifier input, output signals and corresponding reflection coefficients under different load impedance mismatch conditions within a certain range.

[0107] The at least one set of historical signals includes a plurality of input signals, a plurality of output signals, and a plurality of first reflection signals under a plurality of load conditions. Step S220, calculating fixed model coefficients according to the at least one set of historical signals, includes: calculating a plurality of historical reflection coefficients according to the plurality of first reflection signals and the plurality of first output signals respectively; and calculating fixed model coefficients according to the plurality of first input signals, the plurality of first output signals, and the plurality of historical reflection coefficients.

[0108] like Figure 5 As shown, the corresponding Smith original image is shown on the right side of each output signal, and Smith represents the corresponding power amplifier load impedance. The reflection coefficient detection module can observe the load change in real time. The reflection coefficient detection module obtains the output signal and reflection signal under different load conditions and calculates the corresponding reflection coefficient. The model parameter extraction module obtains the corresponding input, output signal and reflection coefficient. The model parameter extraction module uses multiple sets of power amplifier state information (input signal, output signal and reflection coefficient) to calculate the power amplifier model or predistorter model coefficient.

[0109] Figure 6 The figure schematically shows a schematic diagram of a predistorter according to an embodiment of the present invention.

[0110] like Figure 6 As shown, the predistorter can generate a real-time predistortion signal according to the fixed model coefficients c1(k), c2(k), c3(k) and the original input signal a1. First, the predistorter retrieves the fixed model coefficients c1(k), c2(k) and c3(k) calculated by the model parameter extraction module. The fixed model coefficients are suitable for the predistorter model with load variation within a certain range.

[0111] In order to improve the linearization correction effect of the predistorter, it is also necessary to adjust the fixed model coefficients c1(k), c2(k) and c3(k) according to the real-time reflection coefficient. Therefore, the predistorter also needs to obtain the real-time reflection coefficient calculated by the reflection coefficient detection module, and adjust the fixed model coefficients c1(k), c2(k) and c3(k) according to equations (8) to (10). The adjusted model coefficients and It is directly used in the predistorter to make the predistorter generate a predistortion signal.

[0112] Figure 7 A schematic diagram of a digital predistortion system according to another embodiment of the present invention is schematically shown.

[0113] like Figure 7 As shown, the present invention provides a digital predistortion system 700 , including a first calculation module 710 , a second calculation module 720 and a predistorter 730 .

[0114] The first calculation module 710 is used to obtain at least one set of historical signals and calculate fixed model coefficients according to the at least one set of historical signals. The second calculation module 720 is used to obtain a set of real-time signals and calculate real-time reflection coefficients according to the set of real-time signals. And the predistorter 730 is used to generate a predistorted signal according to the fixed model coefficients and the real-time reflection coefficients.

[0115] Among them, at least one group of historical signals includes at least one first input signal and at least one first output signal and at least one first reflection signal corresponding to the at least one first input signal respectively, and a group of real-time signals includes a second output signal and a second reflection signal corresponding to the second output signal.

[0116] Exemplarily, the first calculation module 710 may include a model parameter extraction module, an analog-to-digital converter, a filter, a down converter, and a first reflection coefficient calculation unit. The first reflection coefficient calculation unit obtains a first output signal and a first reflection signal, and calculates a historical reflection coefficient. The model parameter extraction module receives the output signal through the down converter, the filter, and the analog-to-digital converter, and calculates a fixed model coefficient according to the first output signal, the first input signal, and the historical reflection coefficient. The second calculation module 720 may include a digital-to-analog converter, an up converter, a power amplifier, a coupler, a load antenna, and a second reflection coefficient calculation unit. The second reflection coefficient calculation unit obtains a real-time output signal and a real-time reflection signal, and calculates a real-time reflection coefficient. The predistorter 730 receives an input signal and generates a predistortion signal according to the real-time reflection coefficient and the fixed model coefficient.

[0117] In an embodiment of the present invention, at least one group of historical signals includes a plurality of input signals, a plurality of output signals, and a plurality of first reflection signals under a plurality of load conditions.

[0118] The first calculation module 710 is used to obtain a plurality of first input signals and a plurality of first output signals; calculate the variation range of the load condition according to the plurality of first input signals and the plurality of first output signals; and obtain a plurality of first reflection signals when the variation range is greater than a preset range. The second calculation module 720 is used to obtain a second output signal and a second reflection signal when the variation range is greater than a preset range.

[0119] It should be noted that the digital predistortion system in the embodiment of the present invention corresponds to the digital predistortion method in the embodiment of the present invention. The description of the digital predistortion system specifically refers to the digital predistortion method, which will not be repeated here.

[0120] Regarding the embodiments of the present invention, it should also be noted that, in the absence of conflict, the embodiments of the present invention and the features therein may be combined with each other to obtain new embodiments.

[0121] In addition, the terms "first", "second", etc. are used for descriptive purposes only and should not be understood as indicating or implying relative importance or implicitly indicating the number of the indicated technical features. Thus, a feature defined as "first", "second", etc. may explicitly or implicitly include one or more of the features. In the description of the present invention, unless otherwise specified, "plurality" means two or more.

[0122] Finally, it should be noted that the above embodiments are only used to illustrate the technical solution of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solution of the present invention can be modified or replaced by equivalents without departing from the spirit and scope of the technical solution of the present invention.

Claims

1. A digital predistortion method, characterized in that: include: Obtain at least one set of historical signals and one set of real-time signals; Calculating fixed model coefficients based on the at least one set of historical signals; Calculating a real-time reflection coefficient according to the set of real-time signals; as well as generating a predistortion signal according to the fixed model coefficient and the real-time reflection coefficient; The at least one group of historical signals includes at least one first input signal, at least one first output signal corresponding to the at least one first input signal, and at least one first reflection signal, and the group of real-time signals includes a second output signal and a second reflection signal corresponding to the second output signal; The at least one group of historical signals includes a plurality of input signals, a plurality of output signals, and a plurality of first reflected signals under a plurality of load conditions; and the step of calculating the fixed model coefficients according to the at least one group of historical signals includes: Calculate a plurality of historical reflection coefficients respectively according to the plurality of first reflection signals and the plurality of first output signals; and calculating the fixed model coefficients according to the plurality of first input signals, the plurality of first output signals and the plurality of historical reflection coefficients, The step of generating a predistortion signal according to the fixed model coefficient and the real-time reflection coefficient includes: Calculating a correction coefficient according to the historical reflection coefficient and the real-time reflection coefficient; Correcting the fixed model coefficients according to the correction coefficients to obtain real-time model coefficients; and A predistortion signal is generated by a predistorter according to the real-time model coefficients.

2. The digital predistortion method according to claim 1, characterized in that: The obtaining of at least one set of historical signals and one set of real-time signals comprises: Acquiring the at least one set of historical signals at a first preset frequency, and acquiring the set of real-time signals at a second preset frequency; Wherein, the first preset frequency is less than the second preset frequency.

3. The digital predistortion method according to claim 1, characterized in that: The at least one set of historical signals includes a plurality of input signals, a plurality of output signals, and a plurality of first reflected signals under a plurality of load conditions; the obtaining of at least one set of historical signals and a set of real-time signals includes: Acquire a plurality of first input signals and a plurality of first output signals; Calculating a variation range of a load condition according to the plurality of first input signals and the plurality of first output signals; and When the variation range is greater than a preset range, acquiring the plurality of first reflection signals, the second output signal and the second reflection signal, A predistortion signal is generated by a predistorter according to the real-time model coefficients.

4. The digital predistortion method according to claim 1, characterized in that: The method further comprises: Acquire multiple initial input signals, multiple initial output signals, and multiple initial reflection signals under multiple initial load conditions; Calculating a plurality of initial reflection coefficients according to the plurality of initial output signals and the plurality of initial reflection signals; Obtaining initial model coefficients according to the multiple initial input signals, the multiple initial output signals and the multiple initial reflection coefficients; and An initial predistortion signal is generated by a predistorter according to the initial model coefficients.

5. The digital predistortion method according to claim 4, characterized in that: Before correcting the fixed model coefficient according to the correction coefficient to obtain the real-time model coefficient, the method further includes: The initial model coefficients are updated to the fixed model coefficients.

6. A digital predistortion system, executing the digital predistortion method according to claim 1, characterized in that: include: A first calculation module, used for acquiring at least one set of historical signals, and calculating fixed model coefficients according to the at least one set of historical signals; A second calculation module is used to obtain a set of real-time signals and calculate a real-time reflection coefficient according to the set of real-time signals; as well as A predistorter, configured to generate a predistortion signal according to the fixed model coefficients and the real-time reflection coefficients; Among them, the at least one group of historical signals includes at least one first input signal and at least one first output signal and at least one first reflection signal corresponding to the at least one first input signal respectively, and the group of real-time signals includes a second output signal and a second reflection signal corresponding to the second output signal.

7. The predistortion system according to claim 6, characterized in that: The predistortion system further comprises: an adjustable load for generating multiple initial loads; and An initialization module is used to obtain multiple initial input signals, multiple initial output signals and multiple initial reflection signals under the multiple initial load conditions; calculate multiple initial reflection coefficients according to the multiple initial output signals and the multiple initial reflection signals; and obtain initial model coefficients according to the multiple initial input signals, the multiple initial output signals and the multiple initial reflection coefficients.

8. The predistortion system according to claim 6, characterized in that: The at least one set of historical signals includes a plurality of input signals, a plurality of output signals, and a plurality of first reflected signals under a plurality of load conditions; The first calculation module is used to obtain a plurality of first input signals and a plurality of first output signals; calculate a variation range of the load condition according to the plurality of first input signals and the plurality of first output signals; and obtain the plurality of first reflection signals when the variation range is greater than a preset range; as well as The second calculation module is used to obtain the second output signal and the second reflection signal when the variation range is greater than a preset range.