Short wave transceiver common antenna transmit-receive converter based on PIN diode

Abstract

The invention relates to the technical field of transmit-receive converters, in particular to a short wave transceiver common antenna transmit-receive converter based on a PIN diode. The transmit-receive converter comprises a power control module used for monitoring the output power of a power amplifier in real time and dynamically adjusting the output power through a variable attenuator; the cooperative control unit is configured to adjust the attenuation amount of the output power according to the power control module; determining an adjustment strategy of a pre-distortion coefficient in a digital pre-distortion algorithm; and the digital pre-distortion module is used for determining a pre-distortion coefficient in a digital pre-distortion algorithm according to the adjusted output power and the integer strategy so as to compensate nonlinear distortion of the power amplifier through the pre-distortion coefficient.

Description

Technical Field

[0001] This application relates to the field of technology transceivers, and particularly to a shared antenna transceiver converter for short-wave transceivers based on PIN diodes. Background Art

[0002] PIN diodes are commonly used as switching elements in transceiver converters. In the receiving state, the PIN diode is in a high-impedance state, allowing the received signal to pass through; in the transmitting state, the diode switches to a low-impedance state to let the transmitted signal pass through.

[0003] Although PIN diodes exhibit linear characteristics in the ideal switching state, during the actual switching process (especially in high-power or high-frequency scenarios), the non-linearity of their junction capacitance and on-resistance can cause intermodulation distortion. Therefore, improvement is urgently needed. Summary of the Invention

[0004] Based on this, in view of the above technical problems, it is necessary to provide a shared antenna transceiver converter for short-wave transceivers based on PIN diodes that can optimize the sorting process.

[0005] This application provides a shared antenna transceiver converter for short-wave transceivers based on PIN diodes, which includes: A power control module for real-time monitoring of the output power of the power amplifier and dynamically adjusting the output power through a variable attenuator; A cooperative control unit configured to determine an adjustment strategy for the predistortion coefficient in the digital predistortion algorithm according to the attenuation amount when the power control module adjusts the output power; A digital predistortion module for determining the predistortion coefficient in the digital predistortion algorithm according to the adjusted output power and adjustment strategy, so as to compensate for the non-linear distortion of the power amplifier through the predistortion coefficient.

[0006] In one embodiment, the power control module includes: A directional coupler disposed at the end of the transmit channel to couple the output power; A logarithmic power detector connected to the directional coupler and outputting a voltage signal proportional to the output power; A microcontroller configured to compare the voltage signal with a preset threshold and determine a control signal to drive the variable attenuator.

[0007] In one embodiment, the cooperative control unit adjusts the input signal amplitude range of the digital predistortion model according to the current attenuation amount. Specifically, when the attenuation amount increases, the dynamic range of the input signal of the predistortion model is expanded to cover the high-power non-linear region; when the attenuation amount decreases, the dynamic range of the input signal is reduced to optimize the computing resources.

[0008] In one embodiment, the digital predistortion module includes: an analog-to-digital converter for digitizing the output power of the power amplifier; Field Programmable Gate Array with built-in memory polynomial model algorithm to calculate predistortion coefficients; The digital-to-analog converter inputs the pre-distorted baseband signal into the power amplifier.

[0009] In one embodiment, the digital predistortion module includes: The intelligent calculation switching unit automatically selects the most suitable compensation model according to the signal characteristics. Specifically, when the signal bandwidth is less than 5 MHz, the classic polynomial mathematical model is used for distortion compensation; when the signal bandwidth is greater than or equal to 5 MHz and the power exceeds 20 dBm, it automatically switches to the neural network model for distortion compensation.

[0010] In one embodiment, the digital predistortion module includes: The compression unit works in conjunction with the intelligent computing switching unit. Specifically, when the intelligent computing switching unit selects the neural network model, the compression unit automatically enables the structure simplification and precision compression functions; when the intelligent computing switching unit selects the polynomial model, the compression unit is in a dormant state.

[0011] In one embodiment, the transceiver converter further comprises: The other factor acquisition module is used to collect other influencing factors; so that the collaborative control unit adjusts the power threshold and determines the adjustment strategy of the pre-distortion coefficient in the digital pre-distortion algorithm according to other influencing factors and the attenuation when the power control module adjusts the output power.

[0012] In one embodiment, the collaborative control unit is configured to: adjust the power threshold and determine the adjustment strategy of the pre-distortion coefficient in the digital pre-distortion algorithm based on a deep reinforcement learning algorithm model, according to other influencing factors, and the attenuation amount when the power control module adjusts the output power.

[0013] In one embodiment, other influencing factors include temperature and humidity information, vibration information and electromagnetic spectrum.

[0014] In one embodiment, the deep reinforcement learning algorithm model includes a forward modeling network and a reverse compensation network.

[0015] The above-mentioned antenna transceiver converter for short-wave transceivers based on PIN diodes. The power control module of the present application can monitor the output power of the power amplifier in real time and dynamically adjust the output power through a variable attenuator. In the actual application of PIN diodes, high power makes the nonlinearity of the junction capacitance and on-resistance more obvious, thus causing intermodulation distortion. The power control module can stabilize the power within a suitable range, avoid exacerbating the nonlinear characteristics of the diode due to excessive power, and reduce the possibility of generating intermodulation distortion from the source. For example, when it is detected that the power is close to the threshold that may cause nonlinear distortion, the variable attenuator timely reduces the power, so that the PIN diode operates in a relatively linear region. The cooperative control unit determines the adjustment strategy of the predistortion coefficient in the digital predistortion algorithm according to the attenuation amount when the power control module adjusts the output power. The digital predistortion module determines the predistortion coefficient based on the adjusted output power and this strategy, and compensates for the nonlinear distortion of the power amplifier. Even if there is nonlinearity in the actual switching process of the PIN diode, the predistortion process can correct the signal before it enters the power amplifier, so that the signal after passing through the power amplifier and the PIN diode is as close to linear as possible, thereby effectively reducing the interference of intermodulation distortion on the received signal and improving the signal quality. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] In order to more clearly illustrate the technical solutions in the embodiments of the present application or related technologies, the following will briefly introduce the drawings required for describing the embodiments of the present application or related technologies. Obviously, the drawings in the following description are only some embodiments of the present application. For those of ordinary skill in the art, without creative efforts, other related drawings can also be obtained based on these drawings.

[0017] Figure 1 Schematic diagram of the antenna transceiver converter for short-wave transceivers based on PIN diodes in one embodiment; Figure 2 Schematic diagram of the power control module in one embodiment; Figure 3 Schematic diagram of the digital predistortion module in one embodiment. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0018] In order to make the objectives, technical solutions and advantages of the present application more clear and understandable, the following further details the present application in conjunction with the drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and are not used to limit the present application.

[0019] A short-wave transceiver generally includes the following main parts: Transmitter: Responsible for generating and amplifying radio frequency signals for transmission. Receiver: Responsible for receiving and processing radio frequency signals. Transceiver Switching Unit: Used to switch between transmission and reception modes to ensure that the antenna is connected to the transmitter during transmission and to the receiver during reception. Antenna: Used to transmit and receive radio frequency signals. In a PIN diode-based transceiver switching unit, PIN diodes are mainly used to achieve fast switching between transmission and reception modes.

[0020] Among them, the transmitter includes: Signal source, power amplifier, and output terminal: Generates low-power radio frequency signals. Power Amplifier (PA): Amplifies the low-power radio frequency signal to sufficient power for transmission through the antenna. Output terminal: The amplified signal is connected to the antenna through the transceiver switching unit.

[0021] Intermodulation Distortion (IMD) refers to the generation of new frequency components when two or more signals of different frequencies pass through a nonlinear element, and these new frequencies may interfere with the original signals. This is an important issue in radio frequency systems, especially in the design of transceiving common antennas, because transmission and reception signals may exist on the same path simultaneously.

[0022] Although PIN diodes exhibit linear characteristics in the ideal switching state, during the actual switching process (especially in high-power or high-frequency scenarios), the nonlinearity of their junction capacitance (Cj) and on-resistance (Rs) will cause intermodulation products. Transmission state: When the diode is in the low-resistance state, if the output power is too high or the drive current is insufficient, it may cause the diode not to conduct fully, leaving residual nonlinear effects. Reception state: During high-sensitivity reception, weak signals may be interfered by the nonlinear noise of the diode (such as thermal noise, harmonics).

[0023] In addition, if the isolation of the transceiver conversion module (such as <30dB) is insufficient, the transmission signal will leak into the receiving channel and mix with the receiving signal to generate intermodulation components (such as third-order intermodulation 2f1−f2). Impedance mismatch at the antenna end or in the transceiver path will cause signal reflection, forming standing waves, further exacerbating the nonlinear effect; poor filter design cannot effectively suppress out-of-band interference signals.

[0024] The specific manifestations of intermodulation distortion are as follows: Intermodulation products may fall within the receiving frequency band (such as the common 3 - 30 MHz in short - wave communication), resulting in a decrease in receiving sensitivity or an increase in the bit error rate. Example: If the transmitted signals are f1 = 10 MHz and f2 = 12 MHz, the third - order intermodulation product 2f1−f2 = 8 MHz may interfere with the communication in adjacent frequency bands. The power of intermodulation distortion is proportional to the cube of the input signal power (PIMD3∝Pin3). The higher the output power, the more significant the intermodulation problem.

[0025] In an exemplary embodiment, a transceiver antenna transceiver converter based on PIN diodes for short - wave is provided, as Figure 1 shown, including: A power control module 10, configured to monitor the output power of the power amplifier in real - time and dynamically adjust the output power through a variable attenuator; A cooperative control unit 20, configured to determine an adjustment strategy for the predistortion coefficient in the digital predistortion algorithm according to the attenuation amount when the power control module adjusts the output power; A digital predistortion module 30, configured to determine the predistortion coefficient in the digital predistortion algorithm according to the adjusted output power and the adjustment strategy, so as to compensate for the nonlinear distortion of the power amplifier through the predistortion coefficient.

[0026] It can be understood that if the output power of the power amplifier is too high, the PIN diode may not work completely linearly, thus generating intermodulation distortion. The power control module monitors the output power in real - time and dynamically adjusts the output power according to the monitoring results to keep it within a suitable range. The variable attenuator is a key component for power adjustment. By adjusting the attenuation value of the attenuator, the signal power output to the antenna can be precisely controlled. By controlling the output power, the possibility of the PIN diode working in a non - linear state is reduced, thereby reducing the risk of intermodulation distortion.

[0027] The power control module 10 monitors the output power of the power amplifier in real - time: The power control module 10 continuously monitors the output power of the power amplifier (PA) to ensure it is within a safe and optimized range. Dynamically adjust the output power: Through the variable attenuator, the module 10 can adjust the output power of the power amplifier in real - time as needed. For example, if the output power is too high, it may cause an increase in nonlinear distortion or interference to other devices, and the module 10 can reduce the power through the attenuator. The power control module 10 is the "regulator" of the entire system, ensuring that the output power of the power amplifier is always in the best state and avoiding problems caused by too high or too low power.

[0028] The collaborative control unit 20 receives the attenuation amount information from module 10 and determines the adjustment strategy of the predistortion coefficient in the digital predistortion algorithm based on this information. Its function is to ensure that the digital predistortion module (module 30) can dynamically adjust the predistortion coefficient according to the power change, thereby optimizing the compensation effect.

[0029] The digital predistortion module 30 calculates the specific predistortion coefficient according to the adjusted output power of module 10 and the adjustment strategy provided by module 20. The input signal is preprocessed using the calculated predistortion coefficient, so that the nonlinear distortion is effectively canceled after the signal passes through the power amplifier.

[0030] In an exemplary embodiment, as Figure 2 shown, the power control module 10 includes: A directional coupler 11, which is arranged at the end of the transmission channel to couple the output power; A logarithmic power detector 12, which is connected to the directional coupler and outputs a voltage signal proportional to the output power; A microcontroller 13, which is configured to compare the voltage signal with a preset threshold and determine a control signal to drive the variable attenuator.

[0031] It can be understood that in the short-wave transceiver system, the transmission channel refers to the complete path for transmitting the modulated high-frequency signal from the power amplifier to the antenna, and its core function is to ensure that the signal is transmitted with appropriate power and linearity. The power amplifier is the starting point of the transmission channel: its output signal enters the transmission channel; the transmission channel is an extension of the power amplifier: it includes all signal processing links after the PA (such as detection, attenuation, transmission).

[0032] Among them, the variable attenuator is a voltage-controlled attenuator, with an attenuation adjustment step of 0.5 dB and a dynamic range of not less than 30 dB.

[0033] The directional coupler 11 is arranged at the end of the transmission channel, and its function is to couple out part of the energy from the main signal (such as a coupling degree of 20 dB) to obtain an output signal sample for real-time monitoring. It is designed with a microstrip line structure, having an insertion loss of ≤0.5 dB and a directivity of ≥25 dB, ensuring that the main signal transmission is not interfered by the reflection of the detection circuit.

[0034] The logarithmic power detector 12 is connected to the directional coupler 11 through a coaxial cable to receive the coupled output signal sample. Based on the square-law detection characteristic of the Schottky diode, the radio frequency signal is converted into a voltage signal proportional to the output power (e.g., 0 - 5V corresponds to 10 - 100W), and the high-frequency noise is filtered by an internal RC filter (cut-off frequency 10kHz). The dynamic range is 60dB (-20dBm to +40dBm), the response time ≤1μs, and the temperature drift ≤0.1dB / ℃.

[0035] The microcontroller 13 is electrically connected to the logarithmic power detector 12 through an analog-to-digital converter (ADC) and is configured to perform the following steps: a. Read the voltage signal output by the logarithmic power detector 12 in real time; b. Compare the voltage signal with a preset threshold (such as the voltage value corresponding to the upper limit power of the linear operation of the PIN diode); c. Based on the comparison result, calculate the required attenuation ΔA through the PID algorithm and determine the PWM control signal to drive the variable attenuator.

[0036] The current attenuation ΔA is synchronized to the cooperative control unit 20 through the SPI bus for dynamically adjusting the digital predistortion coefficient. The power control module 10 realizes the dynamic adjustment of the transmit power through the following steps: The directional coupler 11 collects the transmit power signal; the logarithmic power detector 12 converts the signal into a voltage signal; the microcontroller 13 compares the voltage signal with the preset threshold to determine the control signal; the variable attenuator adjusts the attenuation according to the control signal to achieve power closed-loop control.

[0037] In this embodiment, the power control accuracy of ±0.5dB ensures that the PIN diode operates in the linear region; the response time of ≤10μs meets the fast power adjustment requirements of burst communication.

[0038] In an exemplary embodiment, the cooperative control unit 20 adjusts the input signal amplitude range of the digital predistortion model according to the current attenuation. Specifically, when the attenuation increases, the dynamic range of the input signal of the predistortion model is expanded to cover the high-power nonlinear region; when the attenuation decreases, the input signal dynamic range is reduced to optimize the computing resources.

[0039] It can be understood that when the power control module 10 detects that the output power is too high and increases the attenuation amount through the variable attenuator, it means that the power amplifier may originally operate in a high-power state. In the high-power state, the power amplifier is more likely to enter the non-linear region and generate relatively serious non-linear distortion. At this time, the cooperative control unit 20 will perform the following operations: The cooperative control unit 20 expands the dynamic range of the input signal of the digital pre-distortion model so that it can cover the non-linear region of the power amplifier in the high-power state. Specifically, it will reset the upper and lower limit values of the input signal in the digital pre-distortion model so that the model can process signals with a larger amplitude. For example, if the original amplitude range of the input signal is -A to +A, when the attenuation amount increases, the cooperative control unit 20 may expand the range to -1.5A to +1.5A.

[0040] By expanding the dynamic range of the input signal, the digital pre-distortion model can more comprehensively model and compensate for the characteristics of the power amplifier in the high-power non-linear region. In this way, even when the power amplifier operates in a high-power state, the non-linear distortion of the signal after pre-distortion processing can be effectively cancelled after passing through the power amplifier, thereby improving the linearity and signal quality of the entire system.

[0041] When the power control module 10 detects that the output power is low and reduces the attenuation amount through the variable attenuator, the power amplifier operates in a relatively low-power linear region and the non-linear distortion is relatively small. At this time, the cooperative control unit 20 will take the following measures: The cooperative control unit 20 reduces the dynamic range of the input signal of the digital pre-distortion model. For example, it reduces the original input signal amplitude range of -A to +A to -0.8A to +0.8A. Reducing the dynamic range of the input signal can reduce the amount of data that the digital pre-distortion model needs to process, thereby optimizing the utilization of computing resources. In the low-power linear region, the non-linear characteristics of the power amplifier are relatively simple and do not require too wide an input signal range for modeling and compensation. By reducing the range, the digital pre-distortion module 30 can reduce the amount of calculation when calculating the pre-distortion coefficient, improve the calculation efficiency, and at the same time reduce the power consumption and hardware cost of the system.

[0042] The cooperative control unit 20 internally stores a mapping relationship table between the attenuation amount and the dynamic range of the input signal. This table is obtained through a large number of experiments and simulations. When receiving the attenuation amount information transmitted by the power control module 10, the cooperative control unit 20 will look up the corresponding adjustment value of the dynamic range of the input signal in the mapping relationship table, and then determine the corresponding adjustment strategy and send it to the digital pre-distortion module 30. The digital pre-distortion module 30 reconfigures its internal signal processing parameters according to this adjustment strategy to achieve the adjustment of the input signal amplitude range.

[0043] In this embodiment, the adjustment strategy of the co - control unit 20 for the input signal amplitude range of the digital pre - distortion model is an adaptive optimization method based on the actual power state, which can effectively control the nonlinear distortion of the power amplifier and rationally utilize resources under different power conditions.

[0044] In an exemplary embodiment, as Figure 3 shown, the digital pre - distortion module 30 includes: An analog - to - digital converter 31, which is used to digitize the output power of the power amplifier; A field - programmable gate array 32, which has a memory polynomial model algorithm built - in to calculate the pre - distortion coefficient; A digital - to - analog converter 33, which inputs the pre - distorted baseband signal into the power amplifier.

[0045] Among them, the order of the memory polynomial model is 5, the memory depth is 3, and the coefficient update period is less than 1 ms.

[0046] It can be understood that the analog - to - digital converter 31 is set between the output end of the power amplifier and the feedback link, and is used to convert the analog output signal of the power amplifier into a digital signal (sampling rate ≥ 200 MSPS, accuracy 14 bits).

[0047] The FPGA 32 has an adaptive algorithm based on the memory polynomial model built - in. The specific parameters are: model order: 5 (x(t)+a2x²(t)+a3x³(t)+a4x 4 (t)+a5x 5 (t)); memory depth: 3 (x(t - τ), τ = 1,2,3); coefficient update period: ≤ 1 ms (iteratively updated through the LMS algorithm).

[0048] The DAC 33 is connected to the FPGA 32 through a high - speed serial interface (such as JESD204B), converts the pre - distorted digital baseband signal into an analog signal (bandwidth ≥ 20 MHz), and inputs it into the power amplifier after low - pass filtering.

[0049] Optionally, the digital pre - distortion module 30 further includes: An intelligent computing switching unit 34, which can automatically select the most suitable compensation model according to the signal characteristics. Specifically, when the signal bandwidth is less than 5 MHz, a classic polynomial mathematical model is used for distortion compensation; when the signal bandwidth is greater than or equal to 5 MHz and the power exceeds 20 dBm, it automatically switches to a neural network model for distortion compensation.

[0050] It is understandable that the intelligent computing switching unit 34 monitors the bandwidth (through FFT analysis) and power (through coupler detection) of the input signal in real time and executes the following strategies: When the signal bandwidth is less than 5MHz, the classic polynomial model is used; When the signal bandwidth is ≥5MHz and the power is >20dBm, switch to the neural network model (such as LSTM network).

[0051] Model switching delay: Based on the hardware-accelerated model loading mechanism, the switching time is ≤50μs.

[0052] The network structure of the neural network model: a 3-layer fully connected network with an input layer dimension of 1024 (IQ samples), a hidden layer dimension of 256, and an output layer dimension of 512 (pre-distortion coefficients). Training data: contains 1000 sets of nonlinear distortion samples at different powers and frequencies, covering the typical working area of ​​PIN diodes.

[0053] Optionally, the digital predistortion module 30 further includes: The compression unit 35 works in conjunction with the intelligent computing switching unit. Specifically, when the intelligent computing switching unit selects the neural network model, the compression unit automatically enables the structure simplification and precision compression functions; when the intelligent computing switching unit selects the polynomial model, the compression unit is in a dormant state.

[0054] It is understandable that when the intelligent computing switching unit 34 selects the neural network model, the compression unit 35 automatically performs: structure simplification: pruning redundant neurons (pruning rate ≥ 30%); precision compression: quantizing 32-bit floating-point coefficients to 8-bit fixed-point numbers. Resource saving effect: The model storage capacity is reduced by 70% and the computational latency is reduced by 40%, while maintaining the IMD3 compensation performance degradation ≤1dB.

[0055] In this embodiment, the digital pre-distortion module 30 realizes nonlinear distortion compensation through the following steps: the analog-to-digital converter 31 collects the output signal of the power amplifier; the FPGA 32 calculates the pre-distortion coefficient (based on the current model); the digital-to-analog converter 33 outputs the pre-distorted baseband signal; the power amplifier amplifies the signal to complete the nonlinear distortion compensation.

[0056] In an exemplary embodiment, the PIN diode-based shortwave transceiver shared antenna transceiver converter also includes other factor acquisition modules for acquiring other influencing factors so that the collaborative control unit can adjust the attenuation amount when the output power is adjusted according to other influencing factors and the power control module, adjust the power threshold and determine the adjustment strategy of the pre-distortion coefficient in the digital pre-distortion algorithm.

[0057] Among them, the other factor acquisition module integrates a temperature and humidity sensor, a vibration sensor, and an electromagnetic spectrum monitoring module to collect multi-dimensional environmental data in real time.

[0058] Among them, other influencing factors include temperature and humidity information, vibration information, and electromagnetic spectrum.

[0059] It can be understood that the temperature sensor is reasonably arranged inside the transceiver converter near key heat-generating components (such as power amplifiers, PIN diodes, etc.). A high-precision digital temperature sensor is used, such as the common DS18B20 sensor, whose measurement accuracy can reach ±0.5°C, and the measurement range is usually from -55°C to +125°C, which can meet the monitoring requirements of the short-wave transceiver under different environmental temperatures. Implementation method: The frequency monitoring module realizes the monitoring of the operating frequency by performing spectral analysis on the input and output signals. It can use a dedicated spectrum analyzer chip or a spectrum analysis algorithm based on FPGA (Field Programmable Gate Array) to complete the signal frequency detection task.

[0060] Optionally, the change in ambient temperature will significantly affect the performance of PIN diodes and power amplifiers. When the ambient temperature rises, parameters such as the on-resistance and junction capacitance of PIN diodes will change, and the gain and linearity of the power amplifier will also be affected, making the device more prone to nonlinear distortion. To avoid this situation, the cooperative control unit will dynamically reduce the power threshold according to the temperature data provided by the temperature sensor. For example, when the temperature rises by 10°C, the power threshold is reduced by 1dBm to ensure that the device can still operate stably in a high-temperature environment and reduce the generation of nonlinear distortion.

[0061] The influence of frequency on the power threshold: At different operating frequencies, the performance of PIN diodes and power amplifiers also varies. At certain specific frequencies, the device may exhibit resonance phenomena, resulting in an abnormal increase in the power amplification factor, which in turn causes nonlinear distortion. The cooperative control unit will adjust the power threshold according to the operating frequency information provided by the frequency monitoring module. For example, near the resonance frequency, the power threshold is appropriately reduced to prevent excessive power from causing device damage and exacerbating nonlinear distortion.

[0062] The influence of temperature on the predistortion coefficient: The change in temperature will change the nonlinear characteristics of power amplifiers and PIN diodes. To effectively compensate for this nonlinear distortion caused by temperature changes, the cooperative control unit will dynamically adjust the predistortion coefficient according to the temperature data. For example, through a pre-established temperature-predistortion coefficient mapping table, when the temperature rises, the predistortion coefficient is adjusted accordingly to enhance the compensation effect for nonlinear distortion.

[0063] Effect of frequency on predistortion coefficient: At different operating frequencies, the non-linear characteristics of the power amplifier and the PIN diode also vary. The cooperative control unit dynamically adjusts the predistortion coefficient according to the operating frequency information. For example, in the high-frequency band, the non-linear characteristics of the power amplifier may be more complex, and the cooperative control unit will increase the adjustment range of the predistortion coefficient to improve the compensation ability for non-linear distortion in the high-frequency band.

[0064] In this embodiment, at different ambient temperatures and operating frequencies, the power threshold and the predistortion coefficient can be adjusted in a timely manner, effectively avoiding non-linear distortion and equipment damage caused by temperature and frequency changes, and improving the stability and reliability of the transceiver converter.

[0065] In an exemplary embodiment, the deep reinforcement learning algorithm model includes a forward modeling network and a reverse compensation network.

[0066] Among them, the deep reinforcement learning algorithm model realizes non-linear compensation by the following method: constructing a dual-channel neural network model, including: a forward modeling network: inputting the baseband signal and outputting the predicted response of the power amplifier; a reverse compensation network: inputting the target linear signal and outputting the predistortion signal; through an adversarial training mechanism, the forward modeling network and the reverse compensation network are iteratively optimized with each other until the predistortion signal meets the accuracy threshold.

[0067] It can be understood that it is mainly responsible for dynamically optimizing the power control threshold and the predistortion model parameters based on the deep reinforcement learning algorithm to achieve effective compensation for the non-linear distortion of the power amplifier.

[0068] Forward modeling network: This network takes the baseband signal as input and outputs the predicted response of the power amplifier through a series of neuron layers and activation functions. Its structure can adopt a multi-layer perceptron (MLP), for example, including an input layer, multiple hidden layers, and an output layer. The input layer receives the feature vector of the baseband signal, the hidden layer performs non-linear transformation on the input information, and the output layer outputs the predicted response of the power amplifier.

[0069] Reverse compensation network: Inputting the target linear signal, after the calculation of the network, it outputs the predistortion signal. The MLP structure can also be adopted, and its purpose is to determine the appropriate predistortion signal according to the target linear signal to compensate for the non-linear characteristics of the power amplifier.

[0070] The forward modeling network and the inverse compensation network are iteratively optimized with each other through an adversarial training mechanism. During the training process, the forward modeling network attempts to accurately predict the response of the power amplifier, while the inverse compensation network tries to determine the pre-distortion signal that can make the output of the power amplifier approach the target linear signal. The two continuously confront and adjust until the pre-distortion signal meets the accuracy threshold. Specifically, by defining a loss function (such as mean square error loss), calculating the error between the prediction result of the forward modeling network and the actual response of the power amplifier, and the error between the pre-distortion signal determined by the inverse compensation network after passing through the power amplifier and the target linear signal, and then using an optimization algorithm (such as stochastic gradient descent) to update the weight parameters of the network, so that the error gradually decreases and finally meets the accuracy requirements.

[0071] Temperature and humidity sensor: Used to monitor the humidity of the environment where the transceiver converter is located in real time. A capacitive temperature and humidity sensor can be adopted, which has characteristics such as high precision and fast response, and can accurately measure the change of environmental humidity. Vibration sensor: Monitor the vibration situation of the transceiver converter. For example, a piezoelectric vibration sensor can be used, which can convert the vibration signal into an electrical signal to sense the vibration state of the device in real time. Electromagnetic spectrum monitoring module: Monitor the surrounding electromagnetic spectrum in real time and obtain relevant information about the electromagnetic environment, such as spectrum distribution, signal intensity, etc. A spectrum analyzer chip or a spectrum monitoring scheme based on software-defined radio (SDR) can be adopted. Multidimensional environmental data is collected in real time to provide rich environmental information for the cooperative control unit to more accurately adjust the working parameters of the transceiver converter.

[0072] The collected environmental data constructs a dynamic weight matrix. For example, when the humidity is high, it may affect the performance of PIN diodes and power amplifiers. At this time, the cooperative control unit will increase the weight of humidity-related factors in the weight matrix; when the vibration is large, the weight of the vibration factor will be adjusted accordingly. In this way, the influence of various environmental factors on the performance of the transceiver converter is comprehensively considered.

[0073] The cooperative control unit jointly optimizes the power threshold, attenuation step, and pre-distortion order using the dynamic weight matrix. Specifically, according to the weights of various factors in the weight matrix, the power threshold is adjusted to ensure the working stability of the power amplifier under different environmental conditions; the attenuation step is adjusted to make the power control more accurate; at the same time, the pre-distortion order is optimized to improve the compensation effect for non-linear distortion. For example, in an environment with strong electromagnetic interference, the pre-distortion order is appropriately increased to enhance the compensation ability for complex non-linear distortion.

[0074] In this embodiment, the dual-channel neural network model and the adversarial training mechanism can more accurately compensate for the non-linear distortion of the power amplifier, improving the signal quality and transmission reliability. Multidimensional environmental data is collected in real time, and the cooperative control unit performs joint optimization based on this data, enabling the transceiver converter to better adapt to different environmental conditions and reducing the impact of environmental factors on the device performance.

[0075] Those of ordinary skill in the art can understand that all or part of the processes in the methods of the above embodiments can be completed by instructing relevant hardware through a computer program. The computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above methods. Among them, any reference to a memory, database, or other medium provided in the various embodiments of the present application can include at least one of non-volatile memory and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetoresistive random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. By way of illustration and not limitation, RAM can be in various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM), etc. The databases involved in the various embodiments of the present application can include at least one of relational databases and non-relational databases. Non-relational databases can include distributed databases based on blockchain, etc., without limitation. The processors involved in the various embodiments of the present application can be general-purpose processors, central processing units, graphics processing units, digital signal processors, programmable logic devices, data processing logics based on quantum computing, artificial intelligence (AI) processors, etc., without limitation.

[0076] The technical features of the above embodiments can be combined arbitrarily. For the sake of concise description, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, it should be considered as the scope recorded in this application.

[0077] The above-described embodiments merely represent several implementation manners of this application. The description is relatively specific and detailed, but it should not be construed as a limitation on the scope of the patent of this application. It should be noted that for those of ordinary skill in the art, without departing from the concept of this application, several modifications and improvements can still be made, and these all belong to the protection scope of this application. Therefore, the protection scope of this application should be subject to the appended claims.

Claims

1. A shortwave transceiver shared antenna transceiver converter based on PIN diode, characterized in that: The transceiver converter comprises: A power control module, used to monitor the output power of the power amplifier in real time and dynamically adjust the output power through a variable attenuator; A collaborative control unit configured to adjust the attenuation of the output power according to the power control module; determine an adjustment strategy for the predistortion coefficient in the digital predistortion algorithm; The digital pre-distortion module is used to determine the pre-distortion coefficient in the digital pre-distortion algorithm according to the adjusted output power and the adjustment strategy, so as to compensate the nonlinear distortion of the power amplifier through the pre-distortion coefficient.

2. A shortwave transceiver shared antenna transceiver converter based on PIN diode according to claim 1, characterized in that: Power control module, including: A directional coupler is provided at the end of the transmission channel to couple the output power; a logarithmic power detector connected to the directional coupler and outputting a voltage signal proportional to the output power; The microcontroller is configured to compare the voltage signal with a preset threshold and determine a control signal to drive the variable attenuator.

3. The shortwave transceiver shared antenna transceiver converter based on PIN diode according to claim 1, characterized in that: The collaborative control unit adjusts the input signal amplitude range of the digital pre-distortion model according to the current attenuation. Specifically, when the attenuation increases, the input signal dynamic range of the pre-distortion model is expanded to cover the high-power nonlinear region; when the attenuation decreases, the input signal dynamic range is reduced to optimize computing resources.

4. The shortwave transceiver shared antenna transceiver converter based on PIN diode according to claim 1, characterized in that: The digital pre-distortion module includes: an analog-to-digital converter for digitizing the output power of the power amplifier; Field Programmable Gate Array with built-in memory polynomial model algorithm to calculate predistortion coefficients; The digital-to-analog converter inputs the pre-distorted baseband signal into the power amplifier.

5. A shortwave transceiver shared antenna transceiver converter based on PIN diode according to claim 4, characterized in that: The digital pre-distortion module includes: The intelligent calculation switching unit automatically selects the most suitable compensation model according to the signal characteristics. Specifically, when the signal bandwidth is less than 5 MHz, the classic polynomial mathematical model is used for distortion compensation; when the signal bandwidth is greater than or equal to 5 MHz and the power exceeds 20 dBm, it automatically switches to the neural network model for distortion compensation.

6. A shortwave transceiver shared antenna transceiver converter based on PIN diode according to claim 5, characterized in that: The digital pre-distortion module includes: The compression unit works in conjunction with the intelligent computing switching unit. Specifically, when the intelligent computing switching unit selects a neural network model, the compression unit automatically enables the structure simplification and precision compression functions; when the intelligent computing switching unit selects a polynomial model, the compression unit is in a dormant state.

7. The shortwave transceiver shared antenna transceiver converter based on PIN diode according to claim 1, characterized in that: The transceiver converter further comprises: The other factor acquisition module is used to acquire other influencing factors so that the collaborative control unit can adjust the attenuation of the output power according to the other influencing factors and the power control module, adjust the power threshold and determine the adjustment strategy of the pre-distortion coefficient in the digital pre-distortion algorithm.

8. According to the PIN diode-based shortwave transceiver shared antenna transceiver converter of claim 7, the collaborative control unit is configured to: based on the deep reinforcement learning algorithm model, adjust the attenuation amount when the power control module adjusts the output power according to the other influencing factors, adjust the power threshold and determine the adjustment strategy of the pre-distortion coefficient in the digital pre-distortion algorithm.

9. The shortwave transceiver shared antenna transceiver converter based on PIN diode according to claim 7, characterized in that: The other influencing factors include temperature and humidity information, vibration information and electromagnetic spectrum.

10. According to the PIN diode-based shortwave transceiver shared antenna transceiver converter of claim 8, the deep reinforcement learning algorithm model includes a forward modeling network and a reverse compensation network.

Images