Digital pre-distortion method for millimeter wave communication transmitting system of sub-array architecture

Abstract

The invention discloses a digital pre-distortion method for a millimeter wave communication transmitting system of a sub-array architecture. Analyzing the nonlinear effect of the system according to the non-ideal characteristic of each device in the hybrid beam forming millimeter wave communication transmitting system based on the sub-array architecture; a receiving antenna is arranged in the direction of a main beam, transmitted radio frequency signals are subjected to down-conversion processing and sampled into baseband digital signals through an analog-to-digital converter, and the baseband digital signals are input into a digital pre-distortion module; signals fed back by the analog-to-digital converter and signals input into a radio frequency link RF chain through digital pre-coding are subjected to differential processing; and the digital signals of the collected baseband are continuously corrected by using an adaptive algorithm, so that the digital pre-distortion in the main beam direction is realized. According to the method, the pre-distortion parameter can be dynamically adjusted according to the actual nonlinear characteristic. By continuously updating the pre-distortion model, nonlinear changes in the system can be effectively coped with, and the method has a wider application scene.

Description

Technical Field

[0001] The present invention belongs to the technical field of wireless communications, and in particular relates to an adaptive digital predistortion method for a hybrid beamforming millimeter wave communication transmission system of a subarray architecture. Background Art

[0002] Millimeter wave communication technology has become one of the key technologies for the next generation of communication systems. Millimeter waves have a large bandwidth and can support higher data transmission rates and lower latency, so their application prospects are very broad. However, millimeter wave signals attenuate greatly during propagation and are significantly affected by obstacles. In order to overcome the propagation attenuation problem of millimeter waves, beamforming technology has emerged. Beamforming adjusts the phase and amplitude of multiple antennas to concentrate energy to form a directional transmission beam, thereby improving the propagation efficiency and anti-interference ability of the signal. As the scale of antenna arrays increases, the complexity of beamforming technology and the demand for computing resources also increase. In order to solve this problem, hybrid beamforming technology has received widespread attention. Hybrid beamforming combines the advantages of digital beamforming and analog beamforming. While ensuring system performance, it can reduce hardware complexity and power consumption. It is especially suitable for large-scale antenna arrays in millimeter wave communications.

[0003] In a large-scale antenna array, the large number of antennas leads to increased hardware complexity and power consumption. In order to optimize this situation, the subarray architecture came into being. This architecture divides the large-scale antenna array into several smaller subarrays, each of which works independently, which can reduce the complexity of the system while improving flexibility and resource utilization.

[0004] In millimeter wave communication systems, power amplifiers and other devices often cause signal distortion due to their nonlinear characteristics, thus affecting communication quality. Digital pre-distortion technology can effectively compensate for the nonlinear distortion caused by power amplifiers and other devices. By pre-processing the transmitted signal, the signal after passing through the amplifier remains as linear as possible, reducing the impact of distortion on the communication system.

[0005] Existing digital pre-distortion methods only consider the nonlinear effects of the power amplifier, ignoring the nonlinear characteristics of other non-ideal devices in the system. In addition, since the millimeter wave communication environment and channel characteristics usually have strong dynamic changes, the traditional fixed digital pre-distortion method may not be able to cope with various changes. Summary of the invention

[0006] In view of the deficiencies in the prior art, the present invention provides a digital predistortion method for a millimeter wave communication transmission system with a subarray architecture. A digital predistortion module is designed on the basis of a hybrid beamforming millimeter wave communication transmission system based on a subarray architecture to compensate for nonlinear distortion in the system. A nonlinear model of the system is established based on the non-ideal characteristics of each device in the system. When the standard test signal emitted by the source enters the MIMO channel through the nonlinear system, it is received by the receiving antenna in the direction of the main beam. The signal is then down-converted and sampled into a baseband digital signal. The signal fed back by the analog-to-digital converter and the signal input to the RF chain after digital precoding are differentially processed. Under the action of the adaptive algorithm, the difference between the signal fed back by the analog-to-digital converter and the signal input to the RF chain after digital precoding is finally made close to 0. At this time, the adaptive algorithm converges and digital predistortion is achieved.

[0007] The technical solution of the present invention comprises the following steps:

[0008] Step 1: Analyze the nonlinear effects of the system based on the non-ideal characteristics of each device in the millimeter wave communication transmission system;

[0009] Step 2: Set up a receiving antenna in the direction of the main beam, down-convert the transmitted RF signal and sample it through an analog-to-digital converter into a baseband digital signal, which is then input into the digital predistortion module:

[0010] Step 3: Perform differential processing on the signal fed back by the analog-to-digital converter and the signal input to the RF chain after digital precoding, and use an adaptive algorithm to continuously correct the collected baseband digital signal to achieve digital predistortion in the main beam direction;

[0011] Step 4: Within the maximum scanning range of the phased array antenna array, continuously adjust the angle to achieve digital pre-distortion under different beam pointing directions.

[0012] Furthermore, the specific method of step 1 is as follows:

[0013] The hybrid beamforming millimeter wave communication transmission system based on subarray architecture consists of a signal source, a digital precoder, a digital-to-analog converter DAC, an upconversion, an analog beamformer, an antenna array, a MIMO channel, a downconversion, and an analog-to-digital converter ADC. On this basis, a digital predistortion module is designed to compensate for the nonlinear distortion in the system. Among them, DAC, upconversion, and analog beamformer form the RF chain. The analog beamformer consists of a phase shifter and a power amplifier. The digital predistortion module includes two parts: the predistorter DPD and the DPD parameter estimation.

[0014] In the millimeter-wave communication transmission system based on the subarray architecture, the nonlinear interference caused by the non-ideal characteristics of the DAC, up-conversion module, power amplifier, down-conversion module, and ADC is considered.

[0015] The resolution of the DAC determines the accuracy of the converted signal, and some distortion will inevitably occur during the signal conversion process. When the digital-to-analog converter discretizes the continuous analog signal into a digital signal, there is a quantization error. The conversion function of the digital-to-analog converter is not completely linear, which will introduce nonlinear distortion.

[0016] The up-conversion module is mainly composed of two modules: IQ modulator and up-converter. The incomplete balance of the I-path and Q-path transmission links will lead to imbalance in amplitude and phase.

[0017] For broadband and high-frequency applications, power amplifiers introduce nonlinear distortion. When the power amplifier is working, the relationship between its input and output is not linear, which causes nonlinear distortion of the output signal, including harmonic distortion, intermodulation distortion, etc.

[0018] The down-conversion module consists of a down-converter and an IQ demodulator. The nonlinear distortion of this module is similar to that of the up-conversion module, and there is amplitude and phase imbalance caused by IQ imbalance.

[0019] ADC also has quantization error when discretizing analog signals into digital signals. Similar to DAC, the limited number of digital quantization bits cannot fully accurately represent the continuous analog signal, resulting in a difference between the output digital signal and the original analog signal. The transfer function of the analog-to-digital converter is not completely linear and will introduce nonlinear distortion.

[0020] In the digital beamforming architecture, the digital predistortion module is implemented after the beamforming operation, that is, each power amplifier is equipped with a dedicated predistorter to compensate for the distortion generated by itself. In this case, the signal received by the user end is in the following form:

[0021] R=H T G[F D (W D X)]

[0022] in represents the transmit signal matrix, represents the received signal matrix of different user end directions, K is the sample length, Q is the number of transmitted signal streams; W D and H are P×Q digital beamformers and the channel matrix represents the number of power amplifiers; G(·)=[g1(·),g2(·),...,g P (·)]and Represents the nonlinear transfer functions of the power amplifiers and predistorters in the array.

[0023] In the digital predistortion operation, the inverse behavioral model of the amplifier is identified and used as a predistorter, i.e. p=1,2,...,P. Therefore, the cascaded module of predistorter and power amplifier is regarded as a linear system, and the user will receive a linearized signal containing only information from the corresponding RF chain.

[0024] For a hybrid beamforming millimeter wave communication transmission system with P antennas and Q independent RF chains (Q<<P), each digital stream drives multiple antennas and power amplifiers simultaneously. After analog beamforming, the received signal is described as:

[0025] R=H T G[WF A (X)]

[0026] in represents the nonlinear function of the predistorter. In the SA (subarray) architecture, each Q RF chain is independently connected to a set of P antennas and power amplifiers through N phase shifters, resulting in an overall analog beamformer W SA becomes a block diagonal matrix:

[0027]

[0028] in q=1,...,Q. So the input signal of the power amplifier can be expressed as:

[0029]

[0030] in is the Kronecker product, Represents the input signal of each power amplifier. It can be seen from the above formula that each power amplifier in the subarray hybrid beamforming system is driven only by the phase signal of the corresponding RF chain. Therefore, the received signal of the i-th user can be expressed as:

[0031]

[0032] in i=1,2,...,Q represents a set of univariate nonlinear transfer functions. Ignoring crosstalk, the i-th beam of the SA hybrid beamforming array contains only the transmit signal from the corresponding RF chain. The digital predistortion (DPD) model configured independently for each RF chain can be equivalent to the inverse model of multiple univariate nonlinear models.

[0033] Any continuously differentiable nonlinear function can always be Taylor expanded, so Taylor series are often used to approximate nonlinear functions. In particular, in the modeling of nonlinear behavior of power amplifiers, a model very similar to Taylor series, the Volterra series model, is widely used, and its discretized bandpass expression is as follows:

[0034]

[0035] in are the bandpass and real input and output signals, M is the memory depth, P is the nonlinear order, h p (m1,…,m p ) is a p-order Volterra kernel.

[0036] In DPD processing, the signal is represented by a low-pass complex signal. Therefore, in practical applications, the band-pass Volterra series model in the above formula can be converted into the following equivalent low-pass Volterra series model representation:

[0037]

[0038] The above formulas x(n) and y(n) represent low-pass, complex input and output signals respectively.

[0039] Since the Volterra series model is only applicable to weak nonlinear system modeling and the model complexity is relatively high, the memory polynomial (MP) model is used to replace the equivalent low-pass Volterra series model; the memory polynomial (MP) model is a classic simplified model derived from the Volterra series model. The expression of the commonly used MP model now contains both odd-order distortion terms and even-order distortion terms, as shown in the following formula:

[0040]

[0041] Furthermore, the specific method of step 2 is as follows:

[0042] Based on the construction of a hybrid beamforming millimeter-wave communication transmission system based on a subarray architecture and the nonlinear analysis of the system, a receiving antenna is set in the direction of the main beam to receive the RF signal generated by the transmission system. The signal source transmits standard test signals, such as standard dual-tone signals, QPSK signals, and 16QAM signals. The standard test signal is digitally precoded by a digital precoder, divided into several data streams and input into the RF chain for up-conversion and analog beamforming, and then the beam signal is transmitted through a power amplifier and a phased array antenna array. In actual experiments, the position of the receiving antenna is often fixed. The main beam direction is determined by the azimuth and elevation angles between the phased array antenna array and the receiving antenna.

[0043] The predistorter of the digital predistortion module receives the precoded digital signal as input when the system starts running. Since the DPD parameter estimation module has not received the feedback output signal at this time, the predistorter will not process the signal. The signal is input to the RF link and is affected by the nonlinearity of the DAC, up-conversion module and power amplifier. After passing through the phase shifter network, it is converted into an RF signal with nonlinear distortion. The RF signal is transmitted in the MIMO channel, and after being received by the receiving antenna, it is successively processed by down-conversion and ADC sampling, and converted into a baseband digital signal and fed back to the digital predistortion module. In this process, nonlinear interference of the down-converter and ADC is introduced.

[0044] Furthermore, the specific method of step 3 is as follows:

[0045] In step 2, the system's transmit signal has been down-converted and sampled into a baseband digital signal by an analog-to-digital converter. The baseband digital signal is input into the DPD parameter estimation module in the digital predistortion module to extract the nonlinear model parameters. After inversion, the parameter matrix of the predistorter is obtained. The specific operations are as follows: First, the nonlinear model parameters are extracted by differentially processing the signal fed back by the analog-to-digital converter and the signal input to the RF chain after digital precoding. The parameter matrix of the predistorter is continuously learned and adjusted under the action of the adaptive LMS algorithm, and finally the difference between the standard test signal and the signal fed back by the output signal of the power amplifier is close to 0. At this time, the system response corresponding to the parameter matrix of the predistorter is the inverse effect of the nonlinear response of the entire system, and a digital predistortion model suitable for non-ideal hardware conditions can be obtained.

[0046] The LMS algorithm calculates the output error at each time step and then adjusts the coefficients of the predistorter based on this error. This update process is gradual, that is, the LMS algorithm can gradually approach the ideal linearization effect by adjusting the coefficients in small steps.

[0047] Specifically, for a given input signal x(n) and a desired output signal d(n), the LMS algorithm outputs the actual signal y(n) based on the current model parameter w(n), and then updates the next model parameter w(n+1) by comparing the difference between the actual output signal and the desired output signal. The update rule of the LMS algorithm is as follows:

[0048] w(n+1)=w(n)+μ·e(n)x(n)

[0049] e(n)=d(n)-y(n)

[0050] Among them, w(n) represents the model parameters at the nth moment, μ is the learning rate (also called step size), which determines the update amount of the model parameters in each iteration and affects the convergence speed of the algorithm, e(n) represents the degree to which the actual output signal y(n) deviates from the expected output signal d(n), and x(n) represents the current input signal. According to this update rule, the LMS algorithm will continuously adjust the model parameters so that the predicted output signal gradually approaches the expected output signal, thereby achieving the purpose of signal analysis and prediction.

[0051] When the system transmits a standard test signal under the action of beamforming, the digital predistortion model in the current main beam direction can be obtained using the LMS algorithm.

[0052] Furthermore, the specific method of step 4 is as follows:

[0053] Digital predistortion in all main beam directions is implemented by using the operations of step 2 and step 3. Since the phased array antenna array can change the angle of the main beam pointing by modifying the feeding phase of the phase shifter, it is also necessary to obtain digital predistortion models for all different angles by continuously adjusting the angle of the main beam pointing within the maximum scanning range of the phased array antenna array.

[0054] The beam scanning range of the phased array antenna array is determined by the azimuth angle φ and the scanning angle θ, and the scanning angle is complementary to the elevation angle. The value range of the azimuth angle φ is 0-360°, and the value range of the scanning angle θ is 0-60°. Each set of azimuth angles and scanning angles corresponds to a beam pointing, and each beam pointing corresponds to a digital pre-distortion model. While meeting the beam pointing accuracy, all possible combinations of azimuth angles and scanning angles are traversed to obtain digital pre-distortion models for different angles.

[0055] The digital pre-distortion models trained at different angles are used to process the standard test signals at the transmitter. For each digital pre-distortion model, a set of baseband digital signals are obtained at the receiver through down-conversion and ADC sampling by continuously changing the beam pointing. The error between this set of signals and the standard test signal is calculated, and under the criterion of minimizing the error of the output signal, an optimal digital pre-distortion model is selected as the final adaptive digital pre-distortion model.

[0056] The beneficial effects of the present invention are as follows:

[0057] The present invention proposes an adaptive digital pre-distortion method for a hybrid beamforming millimeter wave communication transmission system based on a subarray architecture, which can compensate for the nonlinear distortion of the system transmission signal caused by non-ideal devices such as power amplifiers, IQ modulators, and up-converters. Since the digital pre-distortion method has an adaptive characteristic, the pre-distortion parameters can be dynamically adjusted according to the actual nonlinear characteristics. By continuously updating the pre-distortion model, it can effectively cope with nonlinear changes in the system and has a wider range of application scenarios. BRIEF DESCRIPTION OF THE DRAWINGS

[0058] Figure 1 Hybrid millimeter wave communication transmission system based on subarray architecture;

[0059] Figure 2 RF chain structure of sub-array architecture;

[0060] Figure 3 The structure of the digital pre-distortion module;

[0061] Figure 4 Principle of LMS algorithm;

[0062] Figure 5 LMS algorithm flow chart;

[0063] Figure 6 Beam pointing of phased array antenna arrays;

[0064] Figure 7 RF chain structure in the embodiment;

[0065] Figure 8 The DAC module in the embodiment;

[0066] Fig. 9 The up-conversion module in the embodiment;

[0067] Fig.10 The analog beamformer of the embodiment;

[0068] Fig.11 A digital predistortion module in an embodiment. DETAILED DESCRIPTION

[0069] The specific implementation method of the present invention is further described below in conjunction with the accompanying drawings.

[0070] Step 1: Build a hybrid beamforming millimeter wave communication transmission system based on subarray architecture:

[0071] like Figure 1 As shown in the figure, the hybrid beamforming millimeter wave communication transmission system based on the subarray architecture consists of a signal source, a digital precoder, a digital-to-analog converter DAC, an upconversion, an analog beamformer, an antenna array, a MIMO channel, a downconversion, an analog-to-digital converter ADC, and a digital predistortion module. Among them, the DAC, the upconversion, and the analog beamformer form the RF chain. The analog beamformer consists of a phase shifter and a power amplifier. The digital predistortion module includes two parts: the predistorter DPD and the DPD parameter estimation.

[0072] The main function of the signal source module is to generate standard test signals, such as standard dual-tone signals, QPSK signals and 16QAM signals, etc. The original signal is modulated by the corresponding modulator and then transmitted backward through the root raised cosine transmit filter to eliminate inter-symbol interference during signal transmission.

[0073] The function of the digital precoder is to precode the baseband signal input by the signal source. ZF precoding, MMSE precoding, SVD precoding and other methods can be used. In a multi-user MIMO system, the precoder uses the channel state information to perform linear transformation on the signals of different users, thereby effectively separating the user signals and reducing the mutual interference between users. The digital precoder uses an algorithm to eliminate the interference of different users to the maximum extent during transmission, ensuring that each user can receive the signal with the least interference. At the same time, the digital precoder can also realize spatial multiplexing, that is, to transmit multiple independent data streams to multiple users or through a multi-antenna system under the same time and frequency resources. This spatial multiplexing adjusts the directionality of the transmitted signal so that multiple data streams can be transmitted through different spatial paths, thereby improving the spectrum efficiency and transmission capacity of the system.

[0074] The DAC converts the pre-coded digital signal into an analog signal for subsequent signal processing. The up-conversion module is mainly composed of two modules: the IQ modulator and the modulator. The IQ modulator first modulates the baseband signal to the intermediate frequency, and then the modulator modulates the intermediate frequency signal to a high frequency signal.

[0075] The analog beamformer mainly consists of a phase shifter and a power amplifier. The main function of the phase shifter in hybrid beamforming is to adjust the phase of the signal so that the signals emitted by multiple antennas form interference superposition in a specific direction, thereby achieving beamforming. Specifically, the phase shifter changes the signal phase of each antenna unit so that the signals of different antenna units have the same phase in the target direction, forming a coherent superposition, thereby maximizing the signal strength in that direction. At the same time, in the non-target direction, the signal phases are inconsistent and cancel each other out, thereby suppressing interference. This method of adjusting the signal phase through a phase shifter can effectively achieve analog beamforming and control the directionality of the transmitted signal. The role of the power amplifier is to amplify the power of the signal to a sufficient level to ensure that the signal can cover the target area and penetrate the long-distance wireless communication environment.

[0076] The down-conversion module converts the RF signal output by the antenna array into a baseband signal, and the ADC module converts the fed-back baseband analog signal into a baseband digital signal, which is input into the DPD parameter estimation module to complete the extraction of the system nonlinear model factors.

[0077] The MIMO channel module is considered as a MIMO channel under microwave darkroom conditions, that is, there is only one direct wave and the transmitting and receiving antennas are in far-field conditions.

[0078] Step 2: Establish the nonlinear model of the system based on the non-ideal characteristics of each device in the millimeter wave communication transmission system:

[0079] In the millimeter-wave communication transmission system based on the subarray architecture, the nonlinear interference caused by the non-ideal characteristics of the DAC, up-conversion module, power amplifier, down-conversion module, and ADC is considered.

[0080] The resolution of the DAC determines the accuracy of the converted signal, and some distortion will inevitably occur during the signal conversion process. When the digital-to-analog converter discretizes the continuous analog signal into a digital signal, there is a quantization error. This is because the limited number of digital quantization bits cannot fully and accurately represent the continuous analog signal, resulting in a difference between the output signal and the original analog signal. The transfer function of the digital-to-analog converter is not completely linear and will introduce nonlinear distortion. The dynamic range of the input and output signals of the digital-to-analog converter is limited. When the input signal is too large or too small and exceeds the effective range of the DAC, the output signal will be distorted or unable to correctly represent the input signal. The digital-to-analog converter uses a clock signal when performing timing control. The jitter or clock noise of the clock signal will be introduced into the output of the DAC, resulting in timing inaccuracies and jitter in the output signal.

[0081] The up-conversion module is mainly composed of two modules: IQ modulator and up-converter. The incomplete balance of the transmission links of I and Q will lead to imbalance of amplitude and phase. For example, the incomplete matching of the amplifier gain and the uneven filter response on the I and Q paths will cause balance errors. The phase difference offset between I and Q paths will cause phase error. The phase error is caused by the instability of the signal source, the jitter of the clock, or the phase drift in the transmission link. Noise is a random interference introduced during the signal transmission process. Noise comes from environmental interference, the noise of the circuit components themselves, and the interference during the signal transmission process. Noise will affect the signal quality of I and Q paths. In wireless communication, the multipath effect is caused by the reflection and propagation of the signal on multiple different paths on the transmission path. This will cause the arrival time and amplitude of the signals on different paths of I and Q to be different, thus causing demodulation errors. Due to the instability of the RF signal source or the intermediate frequency clock of the receiver, the frequency offset will cause a certain frequency difference between I and Q paths, which will cause phase and time misalignment.

[0082] For broadband and high-frequency applications, power amplifiers introduce nonlinear distortion. When a power amplifier is operating, the relationship between its input and output is not linear. This causes distortion of the output signal, including harmonic distortion, intermodulation distortion, etc. Nonlinear distortion causes changes in the shape of the signal, reducing the transmission accuracy and dynamic range of the system.

[0083] The down-conversion module consists of a down-converter and an IQ demodulator, which converts the RF signal output by the antenna array into a baseband signal. The nonlinear distortion of this module is similar to that of the up-conversion module, and there is amplitude and phase imbalance caused by IQ imbalance.

[0084] There is also quantization error when ADC discretizes analog signals into digital signals. Similar to DAC, the limited number of digital quantization bits cannot fully and accurately represent continuous analog signals, resulting in differences between the output digital signal and the original analog signal. The transfer function of the analog-to-digital converter is not completely linear and will introduce nonlinear distortion. This is caused by factors such as the nonlinear characteristics of the ADC, the nonlinearity of the sample-and-hold circuit, and system deviation, which causes the output digital signal to be distorted. The sampling process of the analog-to-digital converter will introduce sampling distortion. When the input signal frequency exceeds half of the sampling frequency, an aliasing effect will occur, resulting in spectral distortion of the output signal.

[0085] In the digital beamforming architecture, the digital predistortion module is implemented after the beamforming operation, that is, each power amplifier is equipped with a dedicated predistorter to compensate for the distortion generated by itself. In this case, the signal received by the user end is in the following form:

[0086] R=H T G[F D (WD X)]

[0087] in represents the transmit signal matrix, represents the received signal matrix of different user end directions, K is the sample length, Q is the number of transmitted signal streams; W D and H are P×Q digital beamformers and the channel matrix represents the number of power amplifiers; G(·)=[g1(·),g2(·),...,g P (·)]and Represents the nonlinear transfer functions of the power amplifiers and predistorters in the array.

[0088] In the digital predistortion operation, the inverse behavioral model of the amplifier is identified and used as a predistorter, i.e. p=1,2,...,P. Therefore, the cascaded module of predistorter and power amplifier is regarded as a linear system, and the user will receive a linearized signal containing only information from the corresponding RF chain.

[0089] For a hybrid beamforming millimeter wave communication transmission system with P antennas and Q independent RF chains (Q<<P), each digital stream drives multiple antennas and power amplifiers simultaneously. After analog beamforming, the received signal is described as:

[0090] R=H T G[WF A (X)]

[0091] in represents the nonlinear function of the predistorter. In the SA (subarray) architecture, each Q RF chain is independently connected to a set of P antennas and power amplifiers through N phase shifters, resulting in an overall analog beamformer W SA becomes a block diagonal matrix:

[0092]

[0093] in q=1,...,Q. So the input signal of the power amplifier can be expressed as:

[0094]

[0095] in is the Kronecker product, Represents the input signal of each power amplifier. It can be seen from the above formula that each power amplifier in the subarray hybrid beamforming system is driven only by the phase signal of the corresponding RF chain. Therefore, the received signal of the i-th user can be expressed as:

[0096]

[0097] in i=1,2,...,Q represents a set of univariate nonlinear transfer functions. Ignoring crosstalk, the i-th beam of the SA hybrid beamforming array contains only the transmit signal from the corresponding RF chain. The digital predistortion (DPD) model configured independently for each RF chain can be equivalent to the inverse model of multiple univariate nonlinear models.

[0098] Any continuously differentiable nonlinear function can always be Taylor expanded, so the Taylor series is often used to approximate nonlinear functions. In particular, in the modeling of nonlinear behavior of power amplifiers, a model very similar to the Taylor series, the Volterra series model, is widely used. The difference between the Volterra series and the Taylor series is that the Volterra series can capture the memory effect and effectively model weak nonlinear memory systems. Its discretized bandpass expression is as follows:

[0099]

[0100] in are the bandpass and real input and output signals, M is the memory depth, P is the nonlinear order, h p (m1,…,m p ) is a p-order Volterra kernel.

[0101] In DPD processing, the signal is represented by a low-pass complex signal. Therefore, in practical applications, the band-pass Volterra series model in the above formula can be converted into the following equivalent low-pass Volterra series model representation:

[0102]

[0103] The above formulas x(n) and y(n) represent low-pass, complex input and output signals respectively.

[0104] The memory polynomial (MP) model is a simplified model derived from the classic Volterra series model. The model is an equivalent low-pass Volterra series model with m1=m2=…=m added. p The commonly used MP model expression now contains both odd-order distortion terms and even-order distortion terms, as shown in the following formula:

[0105]

[0106] Step 3: Set up a receiving antenna in the direction of the main beam, down-convert the transmitted RF signal and sample it through an analog-to-digital converter into a baseband digital signal, which is then input into the digital predistortion module:

[0107] In this step, based on the construction of the hybrid beamforming millimeter wave communication transmission system based on the subarray architecture and the nonlinear analysis of the system, a receiving antenna is set in the direction of the main beam to receive the RF signal generated by the transmission system. The signal source transmits standard test signals, such as standard dual-tone signals, QPSK signals, and 16QAM signals. The standard test signal is digitally pre-coded by the digital pre-coder, divided into several data streams and input into the RF chain for up-conversion and analog beamforming, and then the beam signal is transmitted through the power amplifier and phased array antenna array. The specific structure of the RF chain is as follows: Figure 2 As shown. In actual experiments, the position of the receiving antenna is often fixed. The main beam direction is determined by the azimuth and elevation angles between the phased array antenna array and the receiving antenna.

[0108] The structure of the digital pre-distortion module is as follows: Figure 3 As shown. The predistorter of the digital predistortion module receives the precoded digital signal as input when the system starts running. Since the DPD parameter estimation module has not received the feedback output signal at this time, the predistorter will not process the signal. The signal is input to the RF link and is affected by the nonlinearity of the DAC, up-conversion module and power amplifier. After passing through the phase shifter network, it is converted into an RF signal with nonlinear distortion. The RF signal is transmitted in the MIMO channel, and after being received by the receiving antenna, it is successively processed by down-conversion and ADC sampling, and converted into a baseband digital signal and fed back to the digital predistortion module. In this process, nonlinear interference of the down-converter and ADC is introduced.

[0109] Step 4: Perform differential processing on the signal fed back by the analog-to-digital converter and the signal input to the RF chain after digital precoding, and use the adaptive algorithm to continuously correct the collected baseband digital signal to realize the function of the digital predistortion module:

[0110] In step 3, the system's transmit signal has been down-converted and sampled into a baseband digital signal by the analog-to-digital converter. The baseband digital signal is input into the DPD parameter estimation module of the digital predistortion module to extract the nonlinear model parameter factors. After taking the inverse, the parameter matrix of the predistorter is obtained. The specific operations are as follows: first, the nonlinear model parameters are extracted by differentially processing the signal fed back by the analog-to-digital converter and the signal input to the RF chain after digital precoding. Under the action of the adaptive LMS algorithm, the parameter matrix of the predistorter is continuously learned and adjusted, and finally the difference between the standard test signal and the signal fed back by the output signal of the power amplifier approaches 0. At this time, the system response corresponding to the parameter matrix of the predistorter is exactly the inverse effect of the nonlinear response of the entire system, and a digital predistortion model suitable for non-ideal hardware conditions can be obtained.

[0111] The LMS algorithm calculates the output error at each time step and then adjusts the coefficients of the predistorter based on this error. This update process is gradual, that is, the LMS algorithm can gradually approach the ideal linearization effect by adjusting the coefficients in small steps. Such an adaptive process enables the predistorter to dynamically compensate for nonlinear distortion when the working environment changes.

[0112] The LMS algorithm simplifies the estimation of the gradient in the steepest descent algorithm and is also called the stochastic gradient algorithm. Under the condition of a limited iteration step size, the LMS algorithm converges, and the convergence speed is related to the step size factor. If the step size factor is too large, the algorithm converges faster, but it is possible that the coefficients identified by the algorithm are not the optimal solution but some suboptimal solutions around the optimal solution; conversely, it slows down the convergence speed of the algorithm. Therefore, when using the LMS algorithm to identify parameters, it is necessary to select a good step size factor.

[0113] The principle of LMS algorithm is as follows Figure 4 As shown in the figure. The LMS algorithm has a simple structure and small amount of calculation. The calculation process only contains multiplication and addition and subtraction operations, without complex division operations, and has fewer intermediate parameters. In the process of identifying the predistorter parameters, there is only the estimation error, and the error value is small. Therefore, the calculation complexity is low, and it belongs to a batch processing algorithm, and the algorithm hardware implementability is strong.

[0114] The flowchart of the LMS algorithm is as follows Figure 5 Specifically, for a given input signal x(n) and a desired output signal d(n), the LMS algorithm outputs the actual signal y(n) based on the current model parameter w(n), and then updates the next model parameter w(n+1) by comparing the difference between the actual output signal and the desired output signal. The update rules of the LMS algorithm are as follows:

[0115] w(n+1)=w(n)+μ·e(n)x(n)

[0116] e(n)=d(n)-y(n)

[0117] Among them, w(n) represents the model parameters at the nth moment, μ is the learning rate (also called step size), which determines the update amount of the model parameters in each iteration and affects the convergence speed of the algorithm, e(n) represents the degree to which the actual output signal y(n) deviates from the expected output signal d(n), and x(n) represents the current input signal. According to this update rule, the LMS algorithm will continuously adjust the model parameters so that the predicted output signal gradually approaches the expected output signal, thereby achieving the purpose of signal analysis and prediction.

[0118] When the system transmits a standard test signal under the action of beamforming, the digital predistortion model in the current main beam direction can be obtained using the LMS algorithm.

[0119] Step 5: Within the maximum scanning range of the phased array antenna array, continuously adjust the angle to achieve digital predistortion under different beam pointing directions:

[0120] Digital predistortion in all main beam directions is implemented by using the operations of step 3 and step 4. Since the phased array antenna array can change the angle of the main beam pointing by modifying the feeding phase of the phase shifter, it is also necessary to obtain digital predistortion models for all different angles by continuously adjusting the angle of the main beam pointing within the maximum scanning range of the phased array antenna array.

[0121] The beam scanning range of the phased array antenna array is determined by the azimuth angle φ and the scanning angle θ, and the scanning angle and the elevation angle are complementary. Figure 6 As shown. The range of the azimuth angle φ is 0-360°, and the range of the scanning angle θ is 0-60°. Each set of azimuth angles and scanning angles corresponds to a beam pointing, and each beam pointing corresponds to a digital pre-distortion model. While satisfying the beam pointing accuracy, all possible combinations of azimuth angles and scanning angles are traversed to obtain digital pre-distortion models for different angles.

[0122] The digital pre-distortion models trained at different angles are used to process the standard test signals at the transmitter. For each digital pre-distortion model, a set of baseband digital signals are obtained at the receiver through down-conversion and ADC sampling by continuously changing the beam pointing. The error between this set of signals and the standard test signal is calculated, and under the criterion of minimizing the error of the output signal, an optimal digital pre-distortion model is selected as the final adaptive digital pre-distortion model.

[0123] Example:

[0124] The simulation based on simulink is as follows Figure 1The hybrid beamforming millimeter wave communication transmission system based on subarray architecture is shown in the figure. In the simulation, the azimuth angle φ is set to a range of 0-360°, and the scanning angle θ is set to a range of 0-60°. During the simulation, a reasonable combination of azimuth and elevation angles is automatically generated to calculate the beamforming related parameters. The system first generates a sampling signal of a modulated baseband digital signal or a baseband analog signal from the signal source module, and then passes through the digital precoder and uses the ZF precoding method for digital precoding. The signal is then divided into four data streams and input into four such as Figure 7 The RF chain is composed of a digital-to-analog converter, an IQ modulator, an up-converter, a power divider, a phase shifter, a power amplifier, and an antenna array. Figure 8 As shown, the baseband digital signal is converted into a baseband analog signal. The upconverter is as follows Fig. 9 As shown in Figure 1, under the action of the local oscillation, the analog signal is first up-converted to the intermediate frequency by the IQ modulator, and then up-converted to the millimeter wave frequency band by the up-converter to become a radio frequency signal. Fig.10 As shown, the RF signal is divided into four paths by a power divider, analog beamforming is performed under the action of a phase shifter, and the signal is transmitted through a power amplifier and an antenna array configured as a 2×2 uniform array to enter the MIMO channel for transmission.

[0125] In the RF link, the DAC will cause certain distortion to the output signal due to the limitation of accuracy, and the ADC will also produce corresponding distortion. After the signal passes through the DAC, it is divided into I and Q paths for transmission, which will introduce IQ modulation distortion. In the up-conversion module and the down-conversion module, second-order nonlinear distortion and third-order nonlinear distortion are artificially introduced. The nonlinear characteristics of the power amplifier are simulated based on the memory polynomial model. When the analog signal passes through the power amplifier, it is affected by the memory effect of the power amplifier and the intermodulation distortion caused by the nonlinearity of the power amplifier. The system transmit signal is affected by the nonlinearity of the above non-ideal devices and is received by the receiving antenna after being transmitted through the MIMO channel. It passes through the down-conversion module and ADC in turn, and is combined into a baseband digital signal to be fed back to the digital pre-distortion module.

[0126] The digital predistortion module is located between the digital precoder and the DAC. Its structure is as follows: Fig.11 As shown in the figure, it consists of two parts: the predistorter and the DPD parameter estimation. The DPD parameter estimation part receives the feedback output signal after combination, extracts the nonlinear parameters in the system using the least mean square algorithm, and calculates the minimum mean square error e(n) between the feedback output signal and the precoded digital signal. The predistorter part performs predistortion processing on the precoded digital signal according to the nonlinear parameters and continues to input it into the RF chain. The flowchart of the LMS algorithm is shown in the figure. Figure 5By dynamically adjusting the nonlinear parameters, when e(n) approaches 0, the baseband signal output by the predistorter is linearly related to the feedback output signal after combination, completing the linearization of the system transmission signal, that is, realizing the nonlinear distortion compensation of the system components.

Claims

1. A digital predistortion method for a millimeter wave communication transmission system with a subarray architecture, characterized in that: The steps include: Step 1: Analyze the nonlinear effects of the system based on the non-ideal characteristics of each device in the millimeter wave communication transmission system; Step 2: Set up a receiving antenna in the direction of the main beam, down-convert the transmitted RF signal and sample it through an analog-to-digital converter into a baseband digital signal, which is then input into the digital predistortion module: Step 3: Perform differential processing on the signal fed back by the analog-to-digital converter and the signal input to the RF chain after digital precoding, and use an adaptive algorithm to continuously correct the collected baseband digital signal to achieve digital predistortion in the main beam direction; Step 4: Within the maximum scanning range of the phased array antenna array, continuously adjust the angle to achieve digital pre-distortion under different beam pointing directions.

2. The digital predistortion method for a millimeter wave communication transmission system with a subarray architecture according to claim 1, characterized in that: Step 1: The hybrid beamforming millimeter wave communication transmission system based on subarray architecture consists of a signal source, a digital precoder, a digital-to-analog converter DAC, an upconversion, an analog beamformer, an antenna array, a MIMO channel, a downconversion, and an analog-to-digital converter ADC. On this basis, a digital predistortion module is designed to compensate for the nonlinear distortion in the system; the DAC, the upconversion, and the analog beamformer form the RF chain; the analog beamformer consists of a phase shifter and a power amplifier; the digital predistortion module includes two parts: the predistorter DPD and the DPD parameter estimation. In the millimeter wave communication transmission system based on the subarray architecture, the nonlinear interference caused by the non-ideal characteristics of the DAC, up-conversion module, power amplifier, down-conversion module, and ADC is considered; The resolution of the DAC determines the accuracy of the converted signal, and distortion is inevitable during the signal conversion process; The up-conversion module is mainly composed of two modules: IQ modulator and up-converter. The incomplete balance of the I- and Q-path transmission links will lead to imbalance in amplitude and phase. For broadband and high-frequency applications, power amplifiers introduce nonlinear distortion; when a power amplifier is operating, the relationship between its input and output is not linear, which causes nonlinear distortion of the output signal; The down-conversion module consists of a down-converter and an IQ demodulator. The nonlinear distortion of this module is similar to that of the up-conversion module, and there is amplitude and phase imbalance caused by IQ imbalance. There is also quantization error when the ADC discretizes the analog signal into a digital signal; In the digital beamforming architecture, the digital predistortion module is implemented after the beamforming operation, that is, each power amplifier is equipped with a dedicated predistorter to compensate for the distortion generated by itself; in this case, the signal received by the user end is in the following form: R=H T G[F D (W D X)] in represents the transmit signal matrix, represents the received signal matrix of different user end directions, K is the sample length, Q is the number of transmitted signal streams; W D and H are P×Q digital beamformers and the channel matrix represents the number of power amplifiers; G(·)=[g1(·),g2(·),...,g P (·)]and represents the nonlinear transfer functions of the power amplifiers and predistorters in the array; In the digital predistortion operation, the inverse behavioral model of the amplifier is identified and used as a predistorter, i.e. Therefore, the cascaded module of the predistorter and the power amplifier is considered as a linear system, in which case the user will receive a linearized signal containing only information from the corresponding RF chain; For a hybrid beamforming millimeter wave communication transmission system with P antennas and Q independent RF chains (Q<<P), each digital stream drives multiple antennas and power amplifiers simultaneously; after analog beamforming, the received signal is described as: R=H T G[WF A (X)] in represents the nonlinear function of the predistorter; in the SA (subarray) architecture, each Q RF chain is independently connected to a set of P antennas and power amplifiers through N phase shifters, resulting in an overall analog beamformer W SA becomes a block diagonal matrix: in So the input signal of the power amplifier can be expressed as: in is the Kronecker product, represents the input signal of each power amplifier; It can be seen from the above formula that each power amplifier in the subarray hybrid beamforming system is driven only by the phase signal of the corresponding RF chain. Therefore, the received signal of the i-th user can be expressed as: in represents a set of univariate nonlinear transfer functions; ignoring crosstalk, the i-th beam of the SA hybrid beamforming array contains only the transmit signal from the corresponding RF chain; the digital predistortion (DPD) model independently configured for each RF chain can be equivalent to the inverse model of multiple univariate nonlinear models; The Volterra series model is used in the nonlinear behavior modeling of the power amplifier, and its discretized bandpass expression is as follows: in are the bandpass and real input and output signals, M is the memory depth, P is the nonlinear order, h p (m1,…,m p ) is a Volterra kernel of order p; In DPD processing, the signal is represented by a low-pass complex signal. Therefore, in practical applications, the band-pass Volterra series model in the above formula can be converted into the following equivalent low-pass Volterra series model representation: In the above formula, x(n) and y(n) represent low-pass and complex input and output signals respectively; Since the Volterra series model is only applicable to weak nonlinear system modeling and the model complexity is high, the memory polynomial model is used to replace the equivalent low-pass Volterra series model; the memory polynomial model is as follows:

3. The digital predistortion method for a millimeter wave communication transmission system with a subarray architecture according to claim 2, characterized in that: Step 2: A receiving antenna is set in the direction of the main beam to receive the RF signal generated by the transmitting system; a standard test signal is transmitted by a signal source; the standard test signal is digitally pre-coded by a digital pre-coder, divided into several data streams and input into the RFchain for up-conversion and analog beamforming, and then the beam signal is transmitted through a power amplifier and a phased array antenna array; The main beam direction is determined by the azimuth and elevation angles between the phased array antenna array and the receiving antenna; The predistorter of the digital predistortion module receives the precoded digital signal as input when the system starts running. Since the DPD parameter estimation module has not received the feedback output signal at this time, the predistorter does not process the signal. The signal is input to the RF link and is affected by the nonlinearity of the DAC, up-conversion module and power amplifier. After passing through the phase shifter network, it is converted into an RF signal with nonlinear distortion. The RF signal is transmitted in the MIMO channel, received by the receiving antenna, and then processed by down-conversion and ADC sampling, converted into a baseband digital signal and fed back to the digital pre-distortion module. In this process, nonlinear interference of the down-converter and ADC is introduced.

4. The digital predistortion method for a millimeter wave communication transmission system with a subarray architecture according to claim 3, characterized in that: The standard test signal includes but is not limited to a standard dual-tone signal, a QPSK signal or a 16QAM signal.

5. The digital predistortion method for a millimeter wave communication transmission system with a subarray architecture according to claim 3, characterized in that: Step 3: In step 2, the system's transmission signal has been down-converted and sampled into a baseband digital signal by an analog-to-digital converter. The baseband digital signal is input into the DPD parameter estimation module in the digital predistortion module to extract the nonlinear model parameters. After inversion, the parameter matrix of the predistorter is obtained. The specific operation is as follows: first, the nonlinear model parameters are extracted by performing differential processing on the signal fed back by the analog-to-digital converter and the signal input to the RF chain after digital precoding. The parameter matrix of the predistorter is continuously learned and adjusted under the action of the adaptive LMS algorithm, and finally the difference between the standard test signal and the signal fed back by the output signal of the power amplifier is close to 0. At this time, the system response corresponding to the parameter matrix of the predistorter is the inverse effect of the nonlinear response of the entire system, and a digital predistortion model suitable for non-ideal hardware conditions can be obtained. The LMS algorithm calculates the output error at each time step and then adjusts the coefficients of the predistorter based on this error. This update process is gradual, that is, the LMS algorithm can gradually approach the ideal linearization effect by adjusting the coefficients in small steps. Specifically, for a given input signal x(n) and a desired output signal d(n), the LMS algorithm outputs the actual signal y(n) based on the current model parameter w(n), and then updates the next model parameter w(n+1) by comparing the difference between the actual output signal and the desired output signal; the update rule of the LMS algorithm is as follows: w(n+1)=w(n)+μ·e(n)x(n) e(n)=d(n)-y(n) Among them, w(n) represents the model parameters at the nth moment, μ is the learning rate, e(n) represents the degree to which the actual output signal y(n) deviates from the expected output signal d(n), and x(n) represents the current input signal. According to this update rule, the LMS algorithm will continuously adjust the model parameters so that the predicted output signal gradually approaches the expected output signal, thereby achieving the purpose of signal analysis and prediction. When the system transmits a standard test signal under the action of beamforming, the digital predistortion model in the current main beam direction can be obtained using the LMS algorithm.

6. The digital predistortion method for a millimeter wave communication transmission system with a subarray architecture according to claim 5, characterized in that: Step 4: The operations of step 2 and step 3 are used to implement digital predistortion in all main beam directions. Since the phased array antenna array can change the angle of the main beam pointing by modifying the feeding phase of the phase shifter, it is also necessary to obtain digital predistortion models for all different angles by continuously adjusting the angle of the main beam pointing within the maximum scanning range of the phased array antenna array. The beam scanning range of the phased array antenna array is determined by the azimuth angle φ and the scanning angle θ, and the scanning angle is complementary to the elevation angle; the value range of the azimuth angle φ is 0-360°, and the value range of the scanning angle θ is 0-60°; each set of azimuth angles and scanning angles corresponds to a beam pointing, and each beam pointing corresponds to a digital pre-distortion model; while satisfying the beam pointing accuracy, all possible combinations of azimuth angles and scanning angles are traversed to obtain digital pre-distortion models for different angles; The digital pre-distortion models trained at different angles are used to process the standard test signals at the transmitter respectively; for each digital pre-distortion model, a group of baseband digital signals are obtained at the receiver through down-conversion and ADC sampling by continuously changing the beam pointing; the error between the group of signals and the standard test signal is calculated, and under the criterion of minimizing the output signal error, an optimal digital pre-distortion model is selected as the final adaptive digital pre-distortion model.

Images