Broadband satellite signal transmission system and low-resource digital predistortion method thereof
By combining terminal equipment and vector analyzer in the broadband satellite signal transmission system, using FPGA for complex convolution operation and coefficient processing, the nonlinear distortion problem of signal in the broadband satellite communication system is solved, and flexible multi-rate predistortion compensation is achieved. and efficient signal optimization improves communication efficiency and spectrum flatness.
Patent Information
- Application Number
- CN202411283306.2
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2024-09-13
- Publication Date
- 2024-11-26
- Estimated Expiration
- 2044-09-13
AI Technical Summary
The existing broadband satellite communication system has signal distortion due to equipment nonlinearity in the high-frequency Q/V frequency band, which affects communication efficiency. The existing compensation methods have poor flexibility or high resource consumption, making it difficult to meet the standards of vector analyzers.
Design a broadband satellite signal transmission system, combines terminal equipment and vector analyzer, and realizes data interface, coefficient processing and predistortion processing modules through FPGA, and uses the equalization filter coefficients generated by complex convolution operation and vector analysis to perform predistortion compensation, which is suitable for a variety of Rate signal.
Achieve flexible multi-rate signal predistortion processing, and the compensation effect is basically the same as the optimization effect of the equalization algorithm in the vector analyzer. It improves the spectrum flatness and EVM indicators of the signal, adapts to different nonlinear distortion scenarios without increasing the complexity of the algorithm. Spend.
Smart Images

Figure CN118826844B_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to physical layer channel transmission technology in the field of high-orbit and low-orbit broadband satellite communications, in particular to a broadband satellite signal transmission system and a low-resource digital predistortion method thereof, which can be used for broadband communications of high-orbit and low-orbit satellite gateway stations. Background Art
[0002] Broadband satellite communication systems have the characteristics of high communication rates and wide occupied signal bandwidth. The frequency band used by broadband satellite communication systems is the high-frequency Q / V band. When the baseband signal passes through devices such as frequency converters, amplifiers and filters, serious nonlinear distortion will occur due to the nonlinearity of the devices themselves, resulting in deterioration of the EVM index and the signal quality received by the receiving end, affecting communication efficiency. In order to ensure communication quality, this nonlinear distortion must be compensated.
[0003] At present, there are two main methods for compensating for nonlinear distortion of broadband signals: one is to analyze the inherent characteristics of the channel in advance by the ground transmitter and perform pre-distortion compensation. This method has poor flexibility and cannot adapt to multi-rate application requirements. The other is that the onboard receiver analyzes the received signal and uses an equalization algorithm to estimate and compensate for nonlinearity. This method will increase the resource overhead of the onboard payload and is not suitable for onboard payloads with limited resources. In addition, for the high-frequency Q / V band, it is difficult for these two methods to achieve the standards of a vector analyzer in terms of implementation results. Summary of the invention
[0004] In view of this, the present invention provides a broadband satellite signal transmission system and a low-resource digital predistortion method thereof. The present invention can perform predistortion processing on signals of various rates, has strong flexibility, and uses the coefficients generated by the vector analyzer in a closed loop to compensate the signal, and the compensation effect is basically the same as the optimization effect of the standard equalization algorithm in the vector analyzer.
[0005] The technical solution adopted by the present invention is:
[0006] A broadband satellite signal transmission system, comprising a terminal device, a back-end device and a vector analyzer, wherein the terminal device comprises a host computer, an FPGA and a digital-to-analog converter, the back-end device comprises a filter, a frequency converter and an amplifier, the FPGA is used to implement a data interface module, a coefficient processing module and a pre-distortion processing module, and the pre-distortion processing module is used to perform pre-distortion compensation on the distortion generated by the back-end device;
[0007] The data interface module receives the baseband data input from the external device according to the speed gear requirement sent by the host computer, stores the baseband data in the register, and then sends the baseband data in the register to the pre-distortion processing module;
[0008] The coefficient processing module receives the predistortion coefficient and verification information sent by the host computer, verifies and stores the predistortion coefficient, extracts the coefficient according to the speed gear requirement sent by the host computer, stores the extracted data in the register, and then sends the coefficient data in the register to the predistortion processing module;
[0009] The pre-distortion processing module performs a complex convolution operation on the baseband data sent by the data interface module and the coefficient data sent by the coefficient processing module to obtain a pre-distortion compensated digital sampling signal, and outputs it to the digital-to-analog converter;
[0010] The digital-to-analog converter performs digital-to-analog conversion on the predistortion compensated digital sampling signal output by the predistortion processing module, and outputs the converted analog signal to the back-end device;
[0011] The back-end device filters, converts and amplifies the analog signal output by the digital-to-analog converter to obtain a high-frequency satellite broadband signal with the required bandwidth, frequency and power and outputs it to the antenna end, and at the same time feeds the high-frequency satellite broadband signal back to the vector analyzer;
[0012] The vector analyzer performs vector analysis on the fed-back high-frequency satellite broadband signal, performs equalization optimization compensation on the fed-back high-frequency satellite broadband signal through the equalization filtering algorithm inside the standard instrument, generates two sets of main and auxiliary equalization filtering coefficients, and transmits the generated equalization filtering coefficients to the host computer;
[0013] The host computer generates a set of shaping filter coefficients, generates pre-distortion coefficients according to the equalization filter coefficients, then calculates the verification information of the pre-distortion coefficients, and loads the pre-distortion coefficients with the verification information into the coefficient processing module through the pre-distortion coefficient reconstruction channel; in addition, the rate gear requirement is specified by the host computer and sent to the data interface module and the coefficient processing module.
[0014] Furthermore, the specific working mode of the data interface module is:
[0015] Receive baseband data, which is divided into I-channel data and Q-channel data; for high-speed gear, I and Q are received in parallel in a 4-channel manner, for medium-speed gear, I and Q are received in parallel in a 2-channel manner, and for low-speed gear, I and Q are received in serial in a 1-channel manner;
[0016] The received I-channel data and Q-channel data are placed in the circular shift register respectively; for the high-speed gear, the data are placed in the 67-bit register and 4 data are updated in each clock cycle; for the medium-speed gear, the data are placed in the 65-bit register and 2 data are updated in each clock cycle; for the low-speed gear, the data are placed in the 64-bit register and 1 data is updated in each clock cycle;
[0017] Put the data in the circular shift register into four 64-bit registers; for the high-speed gear, put the 1st-64th bit data into the first 64-bit register, put the 2nd-65th bit data into the second 64-bit register, put the 3rd-66th bit data into the third 64-bit register, and put the 4th-67th bit data into the fourth 64-bit register; for the medium-speed gear, put the 1st-64th bit register into the first and second 64-bit registers, and put the 2nd-65th bit register into the third and fourth 64-bit registers; for the low-speed gear, put the 1st-64th bit data into four 64-bit registers;
[0018] The data in the four 64-bit registers are sent to the pre-distortion processing module.
[0019] Furthermore, the specific working mode of the coefficient processing module is:
[0020] Receive the pre-distortion coefficient and verification information sent by the host computer, store them in the register, and then verify the pre-distortion coefficient. If the verification passes, use the pre-distortion coefficient and report the successful reconstruction of the pre-distortion coefficient through the monitoring channel. If the verification fails, use the original pre-distortion coefficient and report the failure of the reconstruction of the pre-distortion coefficient through the monitoring channel, and request to resend the pre-distortion coefficient.
[0021] The successfully verified pre-distortion coefficients are divided into main coefficients and auxiliary coefficients, which are stored in two 512-bit registers respectively;
[0022] According to the speed gear requirements sent by the host computer, the main coefficients and auxiliary coefficients are extracted and processed in parallel, and then stored in 8 64-bit registers, with one main coefficient and one auxiliary coefficient stored in each register bit; among them, the high-speed gear extracts the main coefficients and auxiliary coefficients by 4 times, puts the 8i+1th main coefficient and auxiliary coefficient into the 1st, 3rd, 5th, and 7th 64-bit registers, and puts the 8i+5th main coefficient and auxiliary coefficient into the 2nd, 4th, 6th, and 8th 64-bit registers; the medium-speed gear extracts the main coefficients and auxiliary coefficients by 2 times, and puts the 8i+5th main coefficient and auxiliary coefficient into the 2nd, 4th, 6th, and 8th 64-bit registers. The i+1 main coefficient and auxiliary coefficient are placed in the 1st and 5th 64-bit registers, the 8i+3th main coefficient and auxiliary coefficient are placed in the 2nd and 6th 64-bit registers, the 8i+5th main coefficient and auxiliary coefficient are placed in the 3rd and 7th 64-bit registers, and the 8i+7th main coefficient and auxiliary coefficient are placed in the 4th and 8th 64-bit registers; the low-speed gear does not extract the main coefficient and auxiliary coefficient, and the 8i+qth main coefficient and auxiliary coefficient are placed in the qth 64-bit register, i=0,2,…,63, q=1,2,…,8;
[0023] The data in the eight 64-bit registers are sent to the pre-distortion processing module.
[0024] Further, the pre-distortion processing module performs a complex convolution operation on the baseband data of the four 64-bit registers sent by the data interface module and the coefficient data of the eight 64-bit registers sent by the coefficient processing module;
[0025] The complex convolution operation uses 512 complex multipliers and 1008 adders, and the complex multipliers are divided into 8 groups, each with 64. The specific method of the complex convolution operation is: the baseband data of the first 64-bit register is placed in the multiplier factor a of the first and second groups of multipliers, the baseband data of the second 64-bit register is placed in the multiplier factor a of the third and fourth groups of multipliers, the baseband data of the third 64-bit register is placed in the multiplier factor a of the fifth and sixth groups of multipliers, the baseband data of the fourth 64-bit register is placed in the multiplier factor a of the seventh and eighth groups of multipliers, and the baseband data of the eight 64-bit registers are placed in the multiplier factor a of the seventh and eighth groups of multipliers. The coefficient data are respectively put into the multiplication factors b of 8 groups of multipliers, and each group of multipliers performs multiplication operations in parallel. The two groups of 8×64 data of the output I and Q paths are accumulated 6 times respectively. The first accumulation obtains 8×32 data, the second accumulation obtains 8×16 data, the third accumulation obtains 8×8 data, the fourth accumulation obtains 8×4 data, the fifth accumulation obtains 8×2 data, and the sixth accumulation obtains 8×1 data. The effective number of bits of the obtained 8-path parallel data of I and Q is intercepted to obtain the pre-distortion compensated digital sampling signal, and 16 parallel paths are output to the digital-to-analog converter.
[0026] Furthermore, the host computer generates a set of shaping filter coefficients, generates pre-distortion coefficients according to the equalization filter coefficients, and then calculates the verification information of the pre-distortion coefficients, specifically in the following manner:
[0027] A set of 512-order shaping filter coefficients is generated according to the root raised cosine formula, and quantized. The two sets of equalization filter coefficients passed in are convolved with the shaping filter coefficients to generate two sets of 1024-bit convolution results, primary and secondary.
[0028] The 257-769 bits of the main and auxiliary 1024-bit convolution results are respectively intercepted to obtain the main and auxiliary coefficients, each of which contains 512 coefficients. Then, the main and auxiliary coefficients are normalized to obtain the predistortion coefficients.
[0029] The check information is calculated for the main and auxiliary coefficients respectively by accumulating and summing the byte by byte and then truncating the byte into a single byte, and the corresponding check information is added to the end of the main and auxiliary coefficients.
[0030] A low-resource digital predistortion method for a broadband satellite signal transmission system is characterized in that it is applied to the above-mentioned broadband satellite signal transmission system and comprises the following steps:
[0031] Step 1, a set of shaping filter coefficients is generated in the host computer, and a pre-distortion coefficient is generated according to the equalization filter coefficient, and then the verification information of the pre-distortion coefficient is calculated, and the pre-distortion coefficient with the verification information is loaded into the coefficient processing module through the pre-distortion coefficient reconstruction channel; in addition, the rate gear requirement is specified by the host computer and sent to the data interface module and the coefficient processing module;
[0032] Step 2: The data interface module receives the baseband data input from the external device according to the speed gear requirement sent by the host computer, stores the baseband data in the register, and then sends the baseband data in the register to the pre-distortion processing module;
[0033] Step 3, the coefficient processing module receives the pre-distortion coefficient and verification information sent by the host computer, verifies and stores the pre-distortion coefficient, and extracts the data according to the rate gear requirements sent by the host computer, stores the extracted data in the register, and then sends the coefficient data in the register to the pre-distortion processing module;
[0034] Step 4, the pre-distortion processing module performs a complex convolution operation on the baseband data sent by the data interface module and the coefficient data sent by the coefficient processing module to obtain a pre-distortion compensated digital sampling signal, and outputs it to the digital-to-analog converter;
[0035] Step 5, the digital-to-analog converter performs digital-to-analog conversion on the predistortion compensated digital sampling signal output by the predistortion processing module, and outputs the converted analog signal to the back-end device;
[0036] Step 6: The back-end device filters, converts and amplifies the analog signal output by the digital-to-analog converter to obtain a high-frequency satellite broadband signal with the required bandwidth, frequency and power, and outputs it to the antenna end, and feeds the high-frequency satellite broadband signal back to the vector analyzer.
[0037] Step 7: The vector analyzer performs vector analysis on the fed-back high-frequency satellite broadband signal, performs equalization optimization compensation on the fed-back high-frequency satellite broadband signal through the equalization filtering algorithm inside the standard instrument, generates two sets of primary and secondary equalization filtering coefficients, and transmits the generated equalization filtering coefficients to the host computer;
[0038] Repeat the above process to achieve continuous pre-distortion processing of baseband data.
[0039] Furthermore, the specific method of step 2 is:
[0040] Receive baseband data, which is divided into I-channel data and Q-channel data; for high-speed gear, I and Q are received in parallel in a 4-channel manner, for medium-speed gear, I and Q are received in parallel in a 2-channel manner, and for low-speed gear, I and Q are received in serial in a 1-channel manner;
[0041] The received I-channel data and Q-channel data are placed in the circular shift register respectively; for the high-speed gear, the data are placed in the 67-bit register and 4 data are updated in each clock cycle; for the medium-speed gear, the data are placed in the 65-bit register and 2 data are updated in each clock cycle; for the low-speed gear, the data are placed in the 64-bit register and 1 data is updated in each clock cycle;
[0042] Put the data in the circular shift register into four 64-bit registers; for the high-speed gear, put the 1st-64th bit data into the first 64-bit register, put the 2nd-65th bit data into the second 64-bit register, put the 3rd-66th bit data into the third 64-bit register, and put the 4th-67th bit data into the fourth 64-bit register; for the medium-speed gear, put the 1st-64th bit register into the first and second 64-bit registers, and put the 2nd-65th bit register into the third and fourth 64-bit registers; for the low-speed gear, put the 1st-64th bit data into four 64-bit registers;
[0043] The data in the four 64-bit registers are sent to the pre-distortion processing module.
[0044] Furthermore, the specific method of step 3 is:
[0045] Receive the pre-distortion coefficient and verification information sent by the host computer, store them in the register, and then verify the pre-distortion coefficient. If the verification passes, use the pre-distortion coefficient and report the successful reconstruction of the pre-distortion coefficient through the monitoring channel. If the verification fails, use the original pre-distortion coefficient and report the failure of the reconstruction of the pre-distortion coefficient through the monitoring channel, and request to resend the pre-distortion coefficient.
[0046] The successfully verified pre-distortion coefficients are divided into main coefficients and auxiliary coefficients, which are stored in two 512-bit registers respectively;
[0047] According to the speed gear requirements sent by the host computer, the main coefficients and auxiliary coefficients are extracted and processed in parallel, and then stored in 8 64-bit registers, with one main coefficient and one auxiliary coefficient stored in each register bit; among them, the high-speed gear extracts the main coefficients and auxiliary coefficients by 4 times, puts the 8i+1th main coefficient and auxiliary coefficient into the 1st, 3rd, 5th, and 7th 64-bit registers, and puts the 8i+5th main coefficient and auxiliary coefficient into the 2nd, 4th, 6th, and 8th 64-bit registers; the medium-speed gear extracts the main coefficients and auxiliary coefficients by 2 times, and puts the 8i+5th main coefficient and auxiliary coefficient into the 2nd, 4th, 6th, and 8th 64-bit registers. The i+1 main coefficient and auxiliary coefficient are placed in the 1st and 5th 64-bit registers, the 8i+3th main coefficient and auxiliary coefficient are placed in the 2nd and 6th 64-bit registers, the 8i+5th main coefficient and auxiliary coefficient are placed in the 3rd and 7th 64-bit registers, and the 8i+7th main coefficient and auxiliary coefficient are placed in the 4th and 8th 64-bit registers; the low-speed gear does not extract the main coefficient and auxiliary coefficient, and the 8i+qth main coefficient and auxiliary coefficient are placed in the qth 64-bit register, i=0,2,…,63, q=1,2,…,8;
[0048] The data in the eight 64-bit registers are sent to the pre-distortion processing module.
[0049] Further, in step 4, the pre-distortion processing module performs a complex convolution operation on the baseband data of the four 64-bit registers sent by the data interface module and the coefficient data of the eight 64-bit registers sent by the coefficient processing module;
[0050] The complex convolution operation uses 512 complex multipliers and 1008 adders, and the complex multipliers are divided into 8 groups, each with 64. The specific method of the complex convolution operation is: the baseband data of the first 64-bit register is placed in the multiplier factor a of the first and second groups of multipliers, the baseband data of the second 64-bit register is placed in the multiplier factor a of the third and fourth groups of multipliers, the baseband data of the third 64-bit register is placed in the multiplier factor a of the fifth and sixth groups of multipliers, the baseband data of the fourth 64-bit register is placed in the multiplier factor a of the seventh and eighth groups of multipliers, and the baseband data of the eight 64-bit registers are placed in the multiplier factor a of the seventh and eighth groups of multipliers. The coefficient data are respectively put into the multiplication factors b of 8 groups of multipliers, and each group of multipliers performs multiplication operations in parallel. The two groups of 8×64 data of the output I and Q paths are accumulated 6 times respectively. The first accumulation obtains 8×32 data, the second accumulation obtains 8×16 data, the third accumulation obtains 8×8 data, the fourth accumulation obtains 8×4 data, the fifth accumulation obtains 8×2 data, and the sixth accumulation obtains 8×1 data. The effective number of bits of the obtained 8-path parallel data of I and Q is intercepted to obtain the pre-distortion compensated digital sampling signal, and 16 parallel paths are output to the digital-to-analog converter.
[0051] Furthermore, in step 1, a set of shaping filter coefficients is generated in the host computer, and pre-distortion coefficients are generated according to the equalization filter coefficients, and then the verification information of the pre-distortion coefficients is calculated, specifically in the following manner:
[0052] A set of 512-order shaping filter coefficients is generated according to the root raised cosine formula, and quantized. The two sets of equalization filter coefficients passed in are convolved with the shaping filter coefficients to generate two sets of 1024-bit convolution results, primary and secondary.
[0053] The 257-769 bits of the main and auxiliary 1024-bit convolution results are respectively intercepted to obtain the main and auxiliary coefficients, each of which contains 512 coefficients. Then, the main and auxiliary coefficients are normalized to obtain the predistortion coefficients.
[0054] The check information is calculated for the main and auxiliary coefficients respectively by accumulating and summing the byte by byte and then truncating the byte into a single byte, and the corresponding check information is added to the end of the main and auxiliary coefficients.
[0055] The beneficial effects of the present invention are:
[0056] 1. The present invention sends the signal sent by the broadband satellite communication terminal to a vector analyzer for vector analysis to generate corresponding equalization coefficients, and sends the equalization coefficients to a host computer. The host computer performs convolution, normalization and other preprocessing on the equalization coefficients and filter coefficients to generate predistortion coefficients. The predistortion coefficients are reloaded into the predistortion algorithm through a coefficient reconstruction channel. The baseband signal and the predistortion coefficients are subjected to complex convolution operations, and the output signal is the signal after predistortion. When the signal passes through the back-end distortion device, the distortion of the signal is significantly optimized, the spectrum flatness is improved, and the EVM index performance is improved.
[0057] 2. The present invention can dynamically reconstruct coefficients according to actual application scenarios to adapt to different nonlinear distortions.
[0058] 3. The present invention can adapt to signals of various rate bandwidths without increasing the complexity of the algorithm.
[0059] 4. The present invention can be combined with a baseband shaping filter algorithm to achieve flexible switching between shaping filters and pre-distortion algorithms. BRIEF DESCRIPTION OF THE DRAWINGS
[0060] Figure 1 It is a schematic diagram of the principle of an embodiment of the present invention.
[0061] Figure 2 Schematic diagram of the principle of the pre-distortion algorithm in the embodiment of the present invention.
[0062] Figure 3 1 is a constellation diagram and an EVM without pre-distortion processing in an embodiment of the present invention.
[0063] Figure 4 1 is a constellation diagram and EVM after equalization of the vector analyzer in an embodiment of the present invention.
[0064] Figure 5 It is the constellation diagram and EVM of the existing algorithm pre-distortion compensation in the embodiment of the present invention.
[0065] Figure 6 1 is a constellation diagram and EVM of the predistortion compensation of the present invention in an embodiment of the present invention. DETAILED DESCRIPTION
[0066] The present invention is further described in detail below in conjunction with the accompanying drawings.
[0067] Broadband satellite signal transmission system, such as Figure 1 and Figure 2 As shown, it includes a terminal device, a back-end device and a vector analyzer, the terminal device includes a host computer, an FPGA and a digital-to-analog converter, the back-end device includes a filter, a frequency converter, and an amplifier, the FPGA is used to implement a data interface module, a coefficient processing module and a pre-distortion processing module, and the pre-distortion processing module is used to perform pre-distortion compensation on the distortion generated by the back-end device;
[0068] The data interface module receives the baseband data input from the external device according to the speed gear requirement sent by the host computer, stores the baseband data in the register, and then sends the baseband data in the register to the pre-distortion processing module; the specific working mode of the data interface module is as follows:
[0069] Receive baseband data, which is divided into I-channel data and Q-channel data; for high-speed gear, I and Q are received in parallel in a 4-channel manner, for medium-speed gear, I and Q are received in parallel in a 2-channel manner, and for low-speed gear, I and Q are received in serial in a 1-channel manner;
[0070] The received I-channel data and Q-channel data are placed in the circular shift register respectively; for the high-speed gear, the data are placed in the 67-bit register and 4 data are updated in each clock cycle; for the medium-speed gear, the data are placed in the 65-bit register and 2 data are updated in each clock cycle; for the low-speed gear, the data are placed in the 64-bit register and 1 data is updated in each clock cycle;
[0071] Put the data in the circular shift register into four 64-bit registers; for the high-speed gear, put the 1st-64th bit data into the first 64-bit register, put the 2nd-65th bit data into the second 64-bit register, put the 3rd-66th bit data into the third 64-bit register, and put the 4th-67th bit data into the fourth 64-bit register; for the medium-speed gear, put the 1st-64th bit register into the first and second 64-bit registers, and put the 2nd-65th bit register into the third and fourth 64-bit registers; for the low-speed gear, put the 1st-64th bit data into four 64-bit registers;
[0072] The data in the four 64-bit registers are sent to the pre-distortion processing module.
[0073] The coefficient processing module receives the predistortion coefficient and verification information sent by the host computer, verifies and stores the predistortion coefficient, extracts the coefficient according to the speed gear requirement sent by the host computer, stores the extracted data in the register, and then sends the coefficient data in the register to the predistortion processing module; the specific working mode of the coefficient processing module is as follows:
[0074] Receive the pre-distortion coefficient and verification information sent by the host computer, store them in the register, and then verify the pre-distortion coefficient. If the verification passes, use the pre-distortion coefficient and report the successful reconstruction of the pre-distortion coefficient through the monitoring channel. If the verification fails, use the original pre-distortion coefficient and report the failure of the reconstruction of the pre-distortion coefficient through the monitoring channel, and request to resend the pre-distortion coefficient.
[0075] The successfully verified pre-distortion coefficients are divided into main coefficients and auxiliary coefficients, which are stored in two 512-bit registers respectively;
[0076] According to the speed gear requirements sent by the host computer, the main coefficients and auxiliary coefficients are extracted and processed in parallel, and then stored in 8 64-bit registers, with one main coefficient and one auxiliary coefficient stored in each register bit; among them, the high-speed gear extracts the main coefficients and auxiliary coefficients by 4 times, puts the 8i+1th main coefficient and auxiliary coefficient into the 1st, 3rd, 5th, and 7th 64-bit registers, and puts the 8i+5th main coefficient and auxiliary coefficient into the 2nd, 4th, 6th, and 8th 64-bit registers; the medium-speed gear extracts the main coefficients and auxiliary coefficients by 2 times, and puts the 8i+5th main coefficient and auxiliary coefficient into the 2nd, 4th, 6th, and 8th 64-bit registers. The i+1 main coefficient and auxiliary coefficient are placed in the 1st and 5th 64-bit registers, the 8i+3th main coefficient and auxiliary coefficient are placed in the 2nd and 6th 64-bit registers, the 8i+5th main coefficient and auxiliary coefficient are placed in the 3rd and 7th 64-bit registers, and the 8i+7th main coefficient and auxiliary coefficient are placed in the 4th and 8th 64-bit registers; the low-speed gear does not extract the main coefficient and auxiliary coefficient, and the 8i+qth main coefficient and auxiliary coefficient are placed in the qth 64-bit register, i=0,2,…,63, q=1,2,…,8;
[0077] The data in the eight 64-bit registers are sent to the pre-distortion processing module.
[0078] The pre-distortion processing module performs a complex convolution operation on the baseband data of the four 64-bit registers sent by the data interface module and the coefficient data of the eight 64-bit registers sent by the coefficient processing module to obtain a pre-distortion compensated digital sampling signal, and outputs it to the digital-to-analog converter;
[0079] The complex convolution operation uses 512 complex multipliers and 1008 adders, and the complex multipliers are divided into 8 groups, each with 64. The specific method of the complex convolution operation is: the baseband data of the first 64-bit register is placed in the multiplier factor a of the first and second groups of multipliers, the baseband data of the second 64-bit register is placed in the multiplier factor a of the third and fourth groups of multipliers, the baseband data of the third 64-bit register is placed in the multiplier factor a of the fifth and sixth groups of multipliers, the baseband data of the fourth 64-bit register is placed in the multiplier factor a of the seventh and eighth groups of multipliers, and the baseband data of the eight 64-bit registers are placed in the multiplier factor a of the seventh and eighth groups of multipliers. The coefficient data are respectively put into the multiplication factors b of 8 groups of multipliers, and each group of multipliers performs multiplication operations in parallel. The two groups of 8×64 data of the output I and Q paths are accumulated 6 times respectively. The first accumulation obtains 8×32 data, the second accumulation obtains 8×16 data, the third accumulation obtains 8×8 data, the fourth accumulation obtains 8×4 data, the fifth accumulation obtains 8×2 data, and the sixth accumulation obtains 8×1 data. The effective number of bits of the obtained 8-path parallel data of I and Q is intercepted to obtain the pre-distortion compensated digital sampling signal, and 16 parallel paths are output to the digital-to-analog converter.
[0080] The digital-to-analog converter performs digital-to-analog conversion on the predistortion compensated digital sampling signal output by the predistortion processing module, and outputs the converted analog signal to the back-end device;
[0081] The back-end device filters, converts and amplifies the analog signal output by the digital-to-analog converter to obtain a high-frequency satellite broadband signal with the required bandwidth, frequency and power and outputs it to the antenna end, and at the same time feeds the high-frequency satellite broadband signal back to the vector analyzer;
[0082] The vector analyzer performs vector analysis on the fed-back high-frequency satellite broadband signal, performs equalization optimization compensation on the fed-back high-frequency satellite broadband signal through the equalization filtering algorithm inside the standard instrument, generates two sets of main and auxiliary equalization filtering coefficients, and transmits the generated equalization filtering coefficients to the host computer;
[0083] The host computer generates a set of shaping filter coefficients, generates pre-distortion coefficients according to the equalization filter coefficients, and then calculates the verification information of the pre-distortion coefficients, specifically in the following manner:
[0084] A set of 512-order shaping filter coefficients is generated according to the root raised cosine formula, and quantized. The two sets of equalization filter coefficients passed in are convolved with the shaping filter coefficients to generate two sets of 1024-bit convolution results, primary and secondary.
[0085] The 257-769 bits of the main and auxiliary 1024-bit convolution results are respectively intercepted to obtain the main and auxiliary coefficients, each of which contains 512 coefficients. Then, the main and auxiliary coefficients are normalized to obtain the predistortion coefficients.
[0086] The check information is calculated for the main and auxiliary coefficients respectively by accumulating and summing the byte by byte and then truncating the byte into a single byte, and the corresponding check information is added to the end of the main and auxiliary coefficients.
[0087] Then, the host computer loads the predistortion coefficients with verification information into the coefficient processing module through the predistortion coefficient reconstruction channel; in addition, the rate gear requirement can be specified by the host computer and sent to the data interface module and the coefficient processing module.
[0088] A low-resource digital predistortion method for a broadband satellite signal transmission system comprises the following steps:
[0089] Step 1, a set of shaping filter coefficients is generated in the host computer, and a pre-distortion coefficient is generated according to the equalization filter coefficient, and then the verification information of the pre-distortion coefficient is calculated, and the pre-distortion coefficient with the verification information is loaded into the coefficient processing module through the pre-distortion coefficient reconstruction channel; in addition, the rate gear requirement is specified by the host computer and sent to the data interface module and the coefficient processing module;
[0090] Step 2: The data interface module receives the baseband data input from the external device according to the speed gear requirement sent by the host computer, stores the baseband data in the register, and then sends the baseband data in the register to the pre-distortion processing module;
[0091] Step 3, the coefficient processing module receives the pre-distortion coefficient and verification information sent by the host computer, verifies and stores the pre-distortion coefficient, and extracts the data according to the rate gear requirements sent by the host computer, stores the extracted data in the register, and then sends the coefficient data in the register to the pre-distortion processing module;
[0092] Step 4, the pre-distortion processing module performs a complex convolution operation on the baseband data sent by the data interface module and the coefficient data sent by the coefficient processing module to obtain a pre-distortion compensated digital sampling signal, and outputs it to the digital-to-analog converter;
[0093] Step 5, the digital-to-analog converter performs digital-to-analog conversion on the predistortion compensated digital sampling signal output by the predistortion processing module, and outputs the converted analog signal to the back-end device;
[0094] Step 6: The back-end device filters, converts and amplifies the analog signal output by the digital-to-analog converter to obtain a high-frequency satellite broadband signal with the required bandwidth, frequency and power, and outputs it to the antenna end, and feeds the high-frequency satellite broadband signal back to the vector analyzer.
[0095] Step 7: The vector analyzer performs vector analysis on the fed-back high-frequency satellite broadband signal, performs equalization optimization compensation on the fed-back high-frequency satellite broadband signal through the equalization filtering algorithm inside the standard instrument, generates two sets of primary and secondary equalization filtering coefficients, and transmits the generated equalization filtering coefficients to the host computer;
[0096] Repeat the above process to achieve continuous pre-distortion processing of baseband data.
[0097] The brief working principle of the present invention is as follows:
[0098] Reference Figure 1 to Figure 2 The broadband satellite signal transmission system generates a baseband shaped and unpredistorted baseband sampling signal and generates a nonlinear distorted signal through the back-end frequency converter filter amplifier and other equipment; the signal is connected to the vector analyzer for equalization optimization processing, and the equalization coefficient generated by the equalization is transmitted to the host computer; the host computer generates a shaping filter coefficient and performs complex convolution and normalization operations on the shaping filter coefficient and the equalization coefficient to generate a predistortion coefficient and verification information; the host computer reloads the predistortion coefficient and verification information into the FPGA through the reconstruction channel; the FPGA predistortion processing module verifies the predistortion coefficient and feeds back the verification result; the FPGA data interface module and the coefficient processing module respectively process the baseband data and the predistortion coefficient in parallel according to different rate requirements and store them in the register; the predistortion processing module in the FPGA performs complex convolution operations on the baseband data in the register with the predistortion coefficient to output the baseband sampling signal, and the baseband sampling signal is then passed through the back-end frequency converter filter amplifier and other equipment to compensate for the nonlinear distortion of the back-end equipment, thereby outputting a standard high-frequency signal.
[0099] The effects of the present invention are as follows:
[0100] Broadband satellite signal transmission system and low-resource digital predistortion method thereof, referring to Figures 3 to 6 ,like Figure 3 As shown in the figure, the constellation diagram and EVM without pre-distortion processing are shown. The constellation points in the constellation diagram are relatively scattered, and the EVM value is 13.35, indicating that the high-frequency signal without pre-distortion processing has serious nonlinear distortion when passing through the back-end equipment; Figure 4 As shown in the figure, the constellation diagram and EVM after vector analyzer equalization, the constellation points in the constellation diagram are relatively concentrated, and the EVM value is 3.62, indicating that the high-frequency signal can be processed by standard instruments to achieve a relatively ideal constellation diagram and EVM; Figure 5As shown in the figure, the constellation diagram and EVM of the existing algorithm for pre-distortion compensation are shown. The degree of aggregation of the constellation points in the constellation diagram has been improved to a certain extent, and the EVM value is 7.62, indicating that the high-frequency signal has been compensated for the nonlinear distortion to a certain extent through the existing algorithm; Figure 6 As shown in the figure, the constellation diagram and EVM of the pre-distortion compensation of the present invention, the constellation points in the constellation diagram are also relatively concentrated, and the EVM value is 3.95, which is basically close to the compensation result of the vector analyzer; in summary, the pre-distortion compensation effect of the present invention is basically the same as the optimization effect of the standard equalization algorithm in the vector analyzer, and the compensation capability is significantly improved compared with the existing pre-distortion algorithm.
[0101] The present invention can compensate for the nonlinear distortion of broadband satellite signals generated by modulation of high-orbit broadband satellite communication terminals when passing through back-end devices such as frequency converters, filters, and amplifiers, improve the flatness of the signal spectrum of large bandwidth and high local oscillator output by the ground system, and improve the error vector magnitude (EVM) index of the system. Without increasing the complexity of the algorithm, the present invention uses parallel processing to achieve dynamic adjustment of 3-speed symbol rates; realizes the function of dynamically reconstructing the pre-distortion coefficient according to different back-end distortion devices to cope with different application scenarios; combines the baseband shaping algorithm with the pre-distortion algorithm, and realizes the flexible switching of baseband shaping and pre-distortion in combination with the dynamic reconstruction function, optimizes the resource utilization of the algorithm, and has engineering significance.
Claims
1. A broadband satellite signal transmission system, characterized in that: It includes a terminal device, a back-end device and a vector analyzer. The terminal device includes a host computer, an FPGA and a digital-to-analog converter. The back-end device includes a filter, a frequency converter and an amplifier. The FPGA is used to implement a data interface module, a coefficient processing module and a pre-distortion processing module. The pre-distortion processing module is used to perform pre-distortion compensation on the distortion generated by the back-end device. The data interface module receives the baseband data input from the external device according to the speed gear requirement sent by the host computer, stores the baseband data in the register, and then sends the baseband data in the register to the pre-distortion processing module; The coefficient processing module receives the predistortion coefficient and verification information sent by the host computer, verifies and stores the predistortion coefficient, extracts the coefficient according to the speed gear requirement sent by the host computer, stores the extracted data in the register, and then sends the coefficient data in the register to the predistortion processing module; The pre-distortion processing module performs a complex convolution operation on the baseband data sent by the data interface module and the coefficient data sent by the coefficient processing module to obtain a pre-distortion compensated digital sampling signal, and outputs it to the digital-to-analog converter; The digital-to-analog converter performs digital-to-analog conversion on the predistortion compensated digital sampling signal output by the predistortion processing module, and outputs the converted analog signal to the back-end device; The back-end device filters, converts and amplifies the analog signal output by the digital-to-analog converter to obtain a high-frequency satellite broadband signal with the required bandwidth, frequency and power and outputs it to the antenna end, and at the same time feeds the high-frequency satellite broadband signal back to the vector analyzer; The vector analyzer performs vector analysis on the fed-back high-frequency satellite broadband signal, performs equalization optimization compensation on the fed-back high-frequency satellite broadband signal through the equalization filtering algorithm inside the standard instrument, generates two sets of main and auxiliary equalization filtering coefficients, and transmits the generated equalization filtering coefficients to the host computer; The host computer generates a set of shaping filter coefficients, generates pre-distortion coefficients according to the equalization filter coefficients, then calculates the verification information of the pre-distortion coefficients, and loads the pre-distortion coefficients with the verification information into the coefficient processing module through the pre-distortion coefficient reconstruction channel; in addition, the rate gear requirement is specified by the host computer and sent to the data interface module and the coefficient processing module.
2. The broadband satellite signal transmission system according to claim 1, characterized in that: The specific working mode of the data interface module is: Receive baseband data, which is divided into I-channel data and Q-channel data; for high-speed gear, I and Q are received in parallel in a 4-channel manner, for medium-speed gear, I and Q are received in parallel in a 2-channel manner, and for low-speed gear, I and Q are received in serial in a 1-channel manner; The received I-channel data and Q-channel data are placed in the circular shift register respectively; for the high-speed gear, the data are placed in the 67-bit register and 4 data are updated in each clock cycle; for the medium-speed gear, the data are placed in the 65-bit register and 2 data are updated in each clock cycle; for the low-speed gear, the data are placed in the 64-bit register and 1 data is updated in each clock cycle; Put the data in the circular shift register into four 64-bit registers; for the high-speed gear, put the 1st-64th bit data into the first 64-bit register, put the 2nd-65th bit data into the second 64-bit register, put the 3rd-66th bit data into the third 64-bit register, and put the 4th-67th bit data into the fourth 64-bit register; for the medium-speed gear, put the 1st-64th bit register into the first and second 64-bit registers, and put the 2nd-65th bit register into the third and fourth 64-bit registers; for the low-speed gear, put the 1st-64th bit data into four 64-bit registers; The data in the four 64-bit registers are sent to the pre-distortion processing module.
3. The broadband satellite signal transmission system according to claim 2, characterized in that: The specific working mode of the coefficient processing module is: Receive the pre-distortion coefficient and verification information sent by the host computer, store them in the register, and then verify the pre-distortion coefficient. If the verification passes, use the pre-distortion coefficient and report the successful reconstruction of the pre-distortion coefficient through the monitoring channel. If the verification fails, use the original pre-distortion coefficient and report the failure of the reconstruction of the pre-distortion coefficient through the monitoring channel, and request to resend the pre-distortion coefficient. The successfully verified pre-distortion coefficients are divided into main coefficients and auxiliary coefficients, which are stored in two 512-bit registers respectively; According to the speed gear requirements sent by the host computer, the main coefficient and the auxiliary coefficient are extracted and processed in parallel, and then stored in 8 64-bit registers, with one main coefficient and one auxiliary coefficient stored in each register bit; among them, the high-speed gear extracts the main coefficient and the auxiliary coefficient by 4 times, puts the 8i+1th main coefficient and the auxiliary coefficient into the 1st, 3rd, 5th, and 7th 64-bit registers, and puts the 8i+5th main coefficient and the auxiliary coefficient into the 2nd, 4th, 6th, and 8th 64-bit registers; the medium-speed gear extracts the main coefficient and the auxiliary coefficient by 2 times, puts the 8i+1th main coefficient and the auxiliary coefficient into the 1st and 5th 64-bit registers, puts the 8i+3th main coefficient and the auxiliary coefficient into the 2nd and 6th 64-bit registers, puts the 8i+5th main coefficient and the auxiliary coefficient into the 3rd and 7th 64-bit registers, and puts the 8i+7th main coefficient and the auxiliary coefficient into the 4th and 8th 64-bit registers; The low-speed gear does not extract the main coefficient and auxiliary coefficient, and puts the 8i+qth main coefficient and auxiliary coefficient into the qth 64-bit register, i=0,2,…,63, q=1,2,…,8; The data in the eight 64-bit registers are sent to the pre-distortion processing module.
4. The broadband satellite signal transmission system according to claim 3, characterized in that: The pre-distortion processing module performs a complex convolution operation on the baseband data of the four 64-bit registers sent by the data interface module and the coefficient data of the eight 64-bit registers sent by the coefficient processing module; The complex convolution operation uses 512 complex multipliers and 1008 adders, and the complex multipliers are divided into 8 groups, each with 64. The specific method of the complex convolution operation is: the baseband data of the first 64-bit register is placed in the multiplier factor a of the first and second groups of multipliers, the baseband data of the second 64-bit register is placed in the multiplier factor a of the third and fourth groups of multipliers, the baseband data of the third 64-bit register is placed in the multiplier factor a of the fifth and sixth groups of multipliers, the baseband data of the fourth 64-bit register is placed in the multiplier factor a of the seventh and eighth groups of multipliers, and the baseband data of the eight 64-bit registers are placed in the multiplier factor a of the seventh and eighth groups of multipliers. The coefficient data are respectively put into the multiplication factors b of 8 groups of multipliers, and each group of multipliers performs multiplication operations in parallel. The two groups of 8×64 data of the output I and Q paths are accumulated 6 times respectively. The first accumulation obtains 8×32 data, the second accumulation obtains 8×16 data, the third accumulation obtains 8×8 data, the fourth accumulation obtains 8×4 data, the fifth accumulation obtains 8×2 data, and the sixth accumulation obtains 8×1 data. The effective number of bits of the obtained 8-path parallel data of I and Q is intercepted to obtain the pre-distortion compensated digital sampling signal, and 16 parallel paths are output to the digital-to-analog converter.
5. The broadband satellite signal transmission system according to claim 4, characterized in that: The host computer generates a set of shaping filter coefficients, generates pre-distortion coefficients according to the equalization filter coefficients, and then calculates the verification information of the pre-distortion coefficients, specifically in the following manner: A set of 512-order shaping filter coefficients is generated according to the root raised cosine formula, and quantized. The two sets of equalization filter coefficients passed in are convolved with the shaping filter coefficients to generate two sets of 1024-bit convolution results, primary and secondary. The 257-769 bits of the main and auxiliary 1024-bit convolution results are respectively intercepted to obtain the main and auxiliary coefficients, each of which contains 512 coefficients. Then, the main and auxiliary coefficients are normalized to obtain the predistortion coefficients. The check information is calculated for the main and auxiliary coefficients respectively by accumulating and summing the byte by byte and then truncating the byte into a single byte, and the corresponding check information is added to the end of the main and auxiliary coefficients.
6. A low-resource digital predistortion method for a broadband satellite signal transmission system, characterized in that: The broadband satellite signal transmission system as claimed in claim 1 comprises the following steps: Step 1, a set of shaping filter coefficients is generated in the host computer, and a pre-distortion coefficient is generated according to the equalization filter coefficient, and then the verification information of the pre-distortion coefficient is calculated, and the pre-distortion coefficient with the verification information is loaded into the coefficient processing module through the pre-distortion coefficient reconstruction channel; in addition, the rate gear requirement is specified by the host computer and sent to the data interface module and the coefficient processing module; Step 2: The data interface module receives the baseband data input from the external device according to the speed gear requirement sent by the host computer, stores the baseband data in the register, and then sends the baseband data in the register to the pre-distortion processing module; Step 3, the coefficient processing module receives the pre-distortion coefficient and verification information sent by the host computer, verifies and stores the pre-distortion coefficient, and extracts the data according to the rate gear requirements sent by the host computer, stores the extracted data in the register, and then sends the coefficient data in the register to the pre-distortion processing module; Step 4, the pre-distortion processing module performs a complex convolution operation on the baseband data sent by the data interface module and the coefficient data sent by the coefficient processing module to obtain a pre-distortion compensated digital sampling signal, and outputs it to the digital-to-analog converter; Step 5, the digital-to-analog converter performs digital-to-analog conversion on the predistortion compensated digital sampling signal output by the predistortion processing module, and outputs the converted analog signal to the back-end device; Step 6: The back-end device filters, converts and amplifies the analog signal output by the digital-to-analog converter to obtain a high-frequency satellite broadband signal with the required bandwidth, frequency and power, and outputs it to the antenna end, and feeds the high-frequency satellite broadband signal back to the vector analyzer. Step 7: The vector analyzer performs vector analysis on the fed-back high-frequency satellite broadband signal, performs equalization optimization compensation on the fed-back high-frequency satellite broadband signal through the equalization filtering algorithm inside the standard instrument, generates two sets of primary and secondary equalization filtering coefficients, and transmits the generated equalization filtering coefficients to the host computer; Repeat the above process to achieve continuous pre-distortion processing of baseband data.
7. The low-resource digital predistortion method according to claim 6, characterized in that: The specific method of step 2 is: Receive baseband data, which is divided into I-channel data and Q-channel data; for high-speed gear, I and Q are received in parallel in a 4-channel manner, for medium-speed gear, I and Q are received in parallel in a 2-channel manner, and for low-speed gear, I and Q are received in serial in a 1-channel manner; The received I-channel data and Q-channel data are placed in the circular shift register respectively; for the high-speed gear, the data are placed in the 67-bit register and 4 data are updated in each clock cycle; for the medium-speed gear, the data are placed in the 65-bit register and 2 data are updated in each clock cycle; for the low-speed gear, the data are placed in the 64-bit register and 1 data is updated in each clock cycle; Put the data in the circular shift register into four 64-bit registers; for the high-speed gear, put the 1st-64th bit data into the first 64-bit register, put the 2nd-65th bit data into the second 64-bit register, put the 3rd-66th bit data into the third 64-bit register, and put the 4th-67th bit data into the fourth 64-bit register; for the medium-speed gear, put the 1st-64th bit register into the first and second 64-bit registers, and put the 2nd-65th bit register into the third and fourth 64-bit registers; for the low-speed gear, put the 1st-64th bit data into four 64-bit registers; The data in the four 64-bit registers are sent to the pre-distortion processing module.
8. The low-resource digital predistortion method according to claim 7, characterized in that: The specific method of step 3 is: Receive the pre-distortion coefficient and verification information sent by the host computer, store them in the register, and then verify the pre-distortion coefficient. If the verification passes, use the pre-distortion coefficient and report the successful reconstruction of the pre-distortion coefficient through the monitoring channel. If the verification fails, use the original pre-distortion coefficient and report the failure of the reconstruction of the pre-distortion coefficient through the monitoring channel, and request to resend the pre-distortion coefficient. The successfully verified pre-distortion coefficients are divided into main coefficients and auxiliary coefficients, which are stored in two 512-bit registers respectively; According to the speed gear requirements sent by the host computer, the main coefficient and the auxiliary coefficient are extracted and processed in parallel, and then stored in 8 64-bit registers, with one main coefficient and one auxiliary coefficient stored in each register bit; among them, the high-speed gear extracts the main coefficient and the auxiliary coefficient by 4 times, puts the 8i+1th main coefficient and the auxiliary coefficient into the 1st, 3rd, 5th, and 7th 64-bit registers, and puts the 8i+5th main coefficient and the auxiliary coefficient into the 2nd, 4th, 6th, and 8th 64-bit registers; the medium-speed gear extracts the main coefficient and the auxiliary coefficient by 2 times, puts the 8i+1th main coefficient and the auxiliary coefficient into the 1st and 5th 64-bit registers, puts the 8i+3th main coefficient and the auxiliary coefficient into the 2nd and 6th 64-bit registers, puts the 8i+5th main coefficient and the auxiliary coefficient into the 3rd and 7th 64-bit registers, and puts the 8i+7th main coefficient and the auxiliary coefficient into the 4th and 8th 64-bit registers; The low-speed gear does not extract the main coefficient and auxiliary coefficient, and puts the 8i+qth main coefficient and auxiliary coefficient into the qth 64-bit register, i=0,2,…,63, q=1,2,…,8; The data in the eight 64-bit registers are sent to the pre-distortion processing module.
9. The low-resource digital predistortion method according to claim 8, characterized in that: In step 4, the pre-distortion processing module performs a complex convolution operation on the baseband data of the four 64-bit registers sent by the data interface module and the coefficient data of the eight 64-bit registers sent by the coefficient processing module; The complex convolution operation uses 512 complex multipliers and 1008 adders, and the complex multipliers are divided into 8 groups, each with 64. The specific method of the complex convolution operation is: the baseband data of the first 64-bit register is placed in the multiplier factor a of the first and second groups of multipliers, the baseband data of the second 64-bit register is placed in the multiplier factor a of the third and fourth groups of multipliers, the baseband data of the third 64-bit register is placed in the multiplier factor a of the fifth and sixth groups of multipliers, the baseband data of the fourth 64-bit register is placed in the multiplier factor a of the seventh and eighth groups of multipliers, and the baseband data of the eight 64-bit registers are placed in the multiplier factor a of the seventh and eighth groups of multipliers. The coefficient data are respectively put into the multiplication factors b of 8 groups of multipliers, and each group of multipliers performs multiplication operations in parallel. The two groups of 8×64 data of the output I and Q paths are accumulated 6 times respectively. The first accumulation obtains 8×32 data, the second accumulation obtains 8×16 data, the third accumulation obtains 8×8 data, the fourth accumulation obtains 8×4 data, the fifth accumulation obtains 8×2 data, and the sixth accumulation obtains 8×1 data. The effective number of bits of the obtained 8-path parallel data of I and Q is intercepted to obtain the pre-distortion compensated digital sampling signal, and 16 parallel paths are output to the digital-to-analog converter.
10. The low-resource digital predistortion method according to claim 9, characterized in that: In step 1, a set of shaping filter coefficients is generated in the host computer, and pre-distortion coefficients are generated according to the equalization filter coefficients, and then the verification information of the pre-distortion coefficients is calculated, specifically in the following manner: A set of 512-order shaping filter coefficients is generated according to the root raised cosine formula, and quantized. The two sets of equalization filter coefficients passed in are convolved with the shaping filter coefficients to generate two sets of 1024-bit convolution results, primary and secondary. The 257-769 bits of the main and auxiliary 1024-bit convolution results are respectively intercepted to obtain the main and auxiliary coefficients, each of which contains 512 coefficients. Then, the main and auxiliary coefficients are normalized to obtain the predistortion coefficients. The check information is calculated for the main and auxiliary coefficients respectively by accumulating and summing the byte by byte and then truncating the byte into a single byte, and the corresponding check information is added to the end of the main and auxiliary coefficients.
Citation Information
Patent Citations
Wideband adaptive digital predistortion engine apparatus based on programmable device
CN101286963A
Digital predistortion device, predistortion control method and emitter
CN104079514A
Method of processing in-band unevenness caused by DAC preprocessing
CN109474285A
Vector signal generation method and device with adjustable error vector amplitude
CN112217579A
