FPGA-based integrated broadband Fourier transform microwave spectrometer storage and processing method and system, and storage medium

By integrating data storage and processing methods with FPGA hardware, and combining DDR3 memory with asynchronous FIFO, radix-4 FFT and CORDIC algorithms, the problems of long data transmission time and high packet loss risk in traditional spectrometers are solved, achieving high-speed storage and efficient processing, and improving the integration level and data processing performance of the spectrometer.

CN121858504APending Publication Date: 2026-04-14NANJING UNIV OF SCI & TECH
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202511933449.8
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2025-12-19
Publication Date
2026-04-14

AI Technical Summary

Technical Problem

Traditional broadband Fourier transform microwave spectrometers rely on a host computer for data processing, resulting in long transmission times, high risk of packet loss, and impact on real-time performance and convenience.

Method used

Data storage and processing are achieved through FPGA hardware integration. By combining DDR3 memory with asynchronous FIFO, and using radix-4 FFT and CORDIC algorithms, high-speed signal caching, stable transmission across clock domains, and efficient signal processing are accomplished.

Benefits of technology

It achieves high-speed data storage and efficient processing, reduces hardware resource consumption, improves signal processing efficiency and spectral data quality, and solves the problems of long transmission time and high risk of packet loss.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN121858504A_ABST
    Figure CN121858504A_ABST
Patent Text Reader

Abstract

The invention discloses a storage and processing method and system of an integrated broadband Fourier transform microwave spectrometer based on an FPGA (Field Programmable Gate Array), and a storage medium. According to the method, time domain signals collected by an ADC are combined and written into a DDR3 through an asynchronous FIFO, and are split and output during reading; a radix-4 FFT algorithm is adopted, a ping-pong storage structure and an assembly line butterfly operation unit are combined, and conflict-free parallel FFT calculation is completed through six-stage extraction through cooperation of an RAM module, a twiddle factor module and a storage management unit; solving a frequency domain complex signal module value through pure shift addition operation by using a CORDIC algorithm, and correcting amplitude attenuation by using a compensation factor; data are stored through a dual-port RAM, four states are controlled by a finite-state machine to be switched, and an average result is output after corresponding points are accumulated. The system comprises a data storage module, a Fourier transform processing module, an amplitude calculation module and an accumulation averaging module. According to the invention, the data throughput efficiency and the real-time processing capability of the broadband microwave spectrometer are obviously improved.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] This invention belongs to the field of Fourier transform microwave spectrometer technology, and in particular relates to a storage and processing method, system, and storage medium for an integrated broadband Fourier transform microwave spectrometer based on FPGA. Background Technology

[0002] Fourier transform microwave (FTMW) spectrometers, as core tools for detecting molecular rotational spectra, possess ultra-high resolution and sensitivity. They can precisely resolve key parameters such as the three-dimensional structure, bond lengths, bond angles, and dipole moments of molecules through the characteristics of molecular rotational energy level transitions, making them akin to a molecular "fingerprint" identification technology. They play an irreplaceable role in fields such as physical chemistry, molecular physics, astrochemistry, and environmental monitoring. With technological advancements, broadband Fourier transform microwave spectrometers are evolving towards integration and miniaturization, while data processing in traditional spectrometers primarily relies on host computers.

[0003] However, traditional host computer processing methods have significant limitations in data transmission for integrated spectrometers: when large amounts of raw spectral data are uploaded to the host computer via the transmission link, problems such as excessively long transmission times and data packet loss easily occur, affecting real-time processing. Furthermore, host computer processing requires manual preprocessing of the raw data, hindering the ease of use of integrated spectrometers. Therefore, a solution capable of efficient storage and processing at the hardware level is needed to replace the traditional host computer processing mode and meet the high-performance requirements of integrated spectrometers. Summary of the Invention

[0004] The purpose of this invention is to provide a storage and processing method, system, and storage medium for an integrated broadband Fourier transform microwave spectrometer based on FPGA. By integrating FPGA hardware, high-speed storage and efficient processing of data are achieved, replacing the traditional host computer processing flow and improving the integration level and data processing performance of the spectrometer.

[0005] To achieve the objectives of this invention, on one hand, this invention provides a storage and processing method for an integrated broadband Fourier transform microwave spectrometer based on FPGA, comprising the following steps:

[0006] S1. The ADC acquisition data and DDR3 memory bit width adaptation and cross-clock domain synchronization are realized through asynchronous FIFO. The 16-bit time domain signal acquired by the ADC is written into the DDR3 memory in sequence, and then read out from the DDR3 memory in preset batches and sent to the subsequent signal processing link.

[0007] S2. For the time-domain signal sent into the signal processing link, the radix-4 FFT algorithm is used, combined with a ping-pong storage structure and a pipelined butterfly operation unit to complete the conversion of the signal from the time domain to the frequency domain, generating a frequency domain complex signal.

[0008] S3. The frequency domain complex signal is calculated using the CORDIC algorithm. The magnitude calculation is specifically performed through pure shift and addition operations to obtain the frequency domain amplitude data.

[0009] S4. Perform at least 4096 corresponding point accumulation operations on the frequency domain amplitude data through a dual-port RAM and a finite state machine, and output the average result after the accumulation is completed.

[0010] On the other hand, the present invention also provides a system for storage and processing of an integrated broadband Fourier transform microwave spectrometer based on FPGA, comprising the following modules:

[0011] The data storage module is used to achieve bit width adaptation and cross-clock domain synchronization between ADC acquisition data and DDR3 memory through asynchronous FIFO. The 16-bit time domain signal acquired by ADC is written into DDR3 memory in sequence, and then read out from DDR3 memory in preset batches and sent to the subsequent signal processing link.

[0012] The Fourier transform processing module is used to convert the time-domain signal of the input signal processing link into the frequency domain using the radix-4 FFT algorithm, combined with a ping-pong storage structure and a pipelined butterfly operation unit, to generate a frequency-domain complex signal.

[0013] The amplitude calculation module is used to calculate the magnitude of the frequency domain complex signal using the CORDIC algorithm. The magnitude calculation is specifically completed through pure shift and addition operations to obtain frequency domain amplitude data.

[0014] The cumulative averaging module is used to perform at least 4096 cumulative operations on the frequency domain amplitude data through a dual-port RAM and a finite state machine, and output the average result after the accumulation is completed.

[0015] An electronic device includes a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor, when executing the program, implements the storage and processing methods described above.

[0016] A non-transitory computer-readable storage medium stores computer instructions for causing the computer to perform the above-described storage and processing methods.

[0017] A computer program product includes computer program instructions that, when executed on a computer, cause the computer to perform the aforementioned storage and processing methods.

[0018] The significant advancement of this invention compared to existing technologies lies in:

[0019] (1) This invention realizes the entire process of data storage and processing through FPGA hardware integration, replacing the traditional host computer processing mode, and effectively solving the problems of long data transmission time, high risk of packet loss, and cumbersome operation in traditional solutions. Through the storage architecture combining DDR3 memory and asynchronous FIFO, high-speed caching of a large amount of acquired data and stable transmission across clock domains are realized; the hardware implementation of the radix-4 FFT algorithm and CORDIC algorithm greatly improves signal processing efficiency and reduces hardware resource consumption; the cumulative averaging module achieves noise suppression through automatic control, improving the quality of spectral data.

[0020] To more clearly illustrate the functional characteristics and structural parameters of the present invention, further explanation is provided below in conjunction with the accompanying drawings and specific embodiments. Attached Figure Description

[0021] The accompanying drawings, which are included to provide a further understanding of the invention and form part of this application, illustrate exemplary embodiments of the invention and, together with their description, serve to explain the invention and do not constitute an undue limitation thereof. In the drawings:

[0022] Figure 1 This is a flowchart of the steps of the present invention:

[0023] Figure 2 This is a timing diagram of the microwave spectrometer of the present invention;

[0024] Figure 3 This is a structural diagram of the DDR3 read / write control module of the present invention;

[0025] Figure 4 This is the overall design structure diagram of the FFT of this invention;

[0026] Figure 5 This is the cumulative average design structure diagram of the present invention. Detailed Implementation

[0027] The technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0028] The core hardware selection for the test platform built in this embodiment is as follows: The FPGA chip is the Xilinx Spartan-7 series XC7A35TFGG484, which has sufficient logic units and DSP resources to meet the requirements of high-speed data processing; the ADC acquisition module uses the LS08D2000 high-speed board, which supports a sampling rate of 600MSPS and 8-bit sampling accuracy to realize the synchronous acquisition of dual-channel intermediate frequency signals; the DDR3 memory is the Micron MT41J128M16HA-125, with a capacity of 128MB and a maximum operating frequency of 400MHz, to ensure high-bandwidth storage; the development environment uses Xilinx Vivado 2018.3 to complete the configuration and logic design of each module.

[0029] Based on the above environment, the present invention provides a storage and processing method for an integrated broadband Fourier transform microwave spectrometer based on FPGA, combined with... Figure 1 Specific processing timing diagrams combined Figure 2 This includes the following steps:

[0030] Step 1: Acquire the dual-channel intermediate frequency signals via ADC, adapt the asynchronous FIFO bit width and synchronize across clock domains, and store them in DDR3 memory. Figure 3 , to obtain a 16-bit time-domain signal that is cached in batches;

[0031] Step 1-1: Acquire the dual intermediate frequency signals using the LS08D2000 high-speed sensor, converting each channel's 8-bit acquisition data into 16-bit acquisition data, and setting the sampling clock to 600MHz;

[0032] Steps 1-2: The 16-bit acquired data is converted into a bit width and synchronized across clock domains through an asynchronous FIFO to obtain a 16-bit time domain signal adapted to DDR3 storage and back-end processing, with a read clock of 200MHz;

[0033] Steps 1-3: Store the adapted 16-bit time domain signal through the MIG IP core and DDR3 incremental address to obtain a 16-bit time domain signal that is cached in batches of 4096 points.

[0034] Step 2: After data preprocessing, the 16-bit time-domain signal is subjected to radix-4 FFT operation to obtain a 4096-point frequency-domain complex signal;

[0035] Step 2-1: Extend the 16-bit time-domain signal using 32-bit signed data to obtain a 32-bit signed time-domain signal that avoids computational overflow. Figure 4 ;

[0036] Step 2-2: The 32-bit signed time-domain signal is processed through a ping-pong storage structure, parallel reading of the rotation factor, and a six-stage pipelined butterfly operation to obtain a 4096-point frequency-domain complex signal.

[0037] Step 3: Obtain 16-bit frequency domain amplitude data from the frequency domain complex signal using the CORDIC algorithm;

[0038] Step 3-1: The frequency domain complex signal is subjected to 16 levels of iterative micro-rotation to obtain a vector signal rotated to the real axis direction;

[0039] Step 3-2: Correct the rotated vector signal using an amplitude compensation factor to obtain high-precision frequency domain amplitude data.

[0040] Step 4: Store the frequency domain amplitude data in a dual-port RAM and control it with a finite state machine to obtain the noise-reduced average spectrum data, and combine it with... Figure 5 ;

[0041] Step 4-1: Initialize the frequency domain amplitude data with the state machine through a 4096×28bit dual-port RAM to obtain a ready accumulation storage environment;

[0042] Step 4-2: The frequency domain amplitude data is accumulated at corresponding points 4096 times by controlling the finite state machine. After being right-shifted by 12 bits and averaged, the denoised 16-bit average frequency domain amplitude data is obtained and output to the host computer through the USB 2.0 interface.

[0043] The system based on the above-mentioned storage and processing method of an integrated broadband Fourier transform microwave spectrometer based on FPGA includes the following modules:

[0044] The data storage module is used to achieve bit width adaptation and cross-clock domain synchronization between ADC acquisition data and DDR3 memory through asynchronous FIFO. The 16-bit time domain signal acquired by ADC is written into DDR3 memory in sequence, and then read out from DDR3 memory in preset batches and sent to the subsequent signal processing link.

[0045] The Fourier transform processing module is used to convert the time-domain signal of the input signal processing link into the frequency domain using the radix-4 FFT algorithm, combined with a ping-pong storage structure and a pipelined butterfly operation unit, to generate a frequency-domain complex signal.

[0046] The amplitude calculation module is used to calculate the magnitude of the frequency domain complex signal using the CORDIC algorithm. The magnitude calculation is specifically completed through pure shift and addition operations to obtain frequency domain amplitude data.

[0047] The cumulative averaging module is used to perform at least 4096 cumulative operations on the frequency domain amplitude data through a dual-port RAM and a finite state machine, and output the average result after the accumulation is completed.

[0048] The Fourier transform processing module specifically includes the following modules:

[0049] The RAM storage module is used to store time-domain signals, FFT intermediate results, and frequency-domain complex signals. Two sets of dual-port RAMs are accessed in parallel, and ping-pong operation is used to prevent collisions.

[0050] The rotation factor storage module is used to pre-store the rotation factor required for the radix-4 FFT, and three ROMs are output in parallel to adapt to the operation timing.

[0051] The radix-4 butterfly operation module is used to receive four data inputs from the RAM storage module and three rotation factors input from the rotation factor storage module, perform butterfly operations, and output intermediate results.

[0052] The storage management module is used to generate conflict-free addresses, control RAM switching, schedule data transmission, and control the six-level extraction operation process.

[0053] The specific implementation of the system is the same as that of the above method, and will not be repeated here.

[0054] It should be noted that, in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or apparatus.

[0055] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A storage and processing method for an integrated broadband Fourier transform microwave spectrometer based on FPGA, characterized in that, Includes the following steps: S1. The ADC acquisition data and DDR3 memory bit width adaptation and cross-clock domain synchronization are realized through asynchronous FIFO. The 16-bit time domain signal acquired by the ADC is written into the DDR3 memory in sequence, and then read out from the DDR3 memory in preset batches and sent to the subsequent signal processing link. S2. For the time-domain signal sent into the signal processing link, the radix-4 FFT algorithm is used, combined with a ping-pong storage structure and a pipelined butterfly operation unit to complete the conversion of the signal from the time domain to the frequency domain, generating a frequency domain complex signal. S3. The frequency domain complex signal is calculated using the CORDIC algorithm. The magnitude calculation is specifically performed through pure shift and addition operations to obtain the frequency domain amplitude data. S4. Perform at least 4096 corresponding point accumulation operations on the frequency domain amplitude data through a dual-port RAM and a finite state machine, and output the average result after the accumulation is completed.

2. The storage and processing method for an FPGA-based integrated broadband Fourier transform microwave spectrometer according to claim 1, characterized in that, The asynchronous FIFO of S1 combines four 16-bit parallel time-domain signals into 64-bit data and writes it into DDR3. When reading out, it splits the 64-bit data into 16-bit parallel outputs, realizing the cross-clock domain transition between the ADC sampling clock and the back-end processing clock.

3. The storage and processing method for an FPGA-based integrated broadband Fourier transform microwave spectrometer according to claim 1, characterized in that, The S2 radix-4 FFT algorithm, through the collaboration of the RAM storage module, twitch factor storage module, radix-4 butterfly operation module, and storage management unit, completes a 4096-point FFT calculation through six levels of extraction, specifically including the following steps: S21. The storage management unit is initialized and configured to store the time-domain signal into two groups of modules of the RAM storage unit in a certain sequence. Each group contains 4 dual-port RAMs to achieve 4-way parallel access. S22. The storage management unit generates a conflict-free address, synchronously reads four data streams from the RAM storage unit, and reads three preset rotation factors in parallel from the rotation factor storage module, and inputs them into the radix-4 butterfly operation module. S23. The radix-4 butterfly operation module performs butterfly operation and outputs 4 intermediate results. The storage management unit writes the results into another set of RAM storage modules according to the ping-pong operation rules to avoid read and write conflicts. S24. Repeat S22-S23 to complete the six-level decimation operation. The two sets of RAM modules alternately serve as input or output carriers. Finally, the storage management unit controls the output of the frequency domain complex signal to complete the FFT calculation.

4. The storage and processing method for an FPGA-based integrated broadband Fourier transform microwave spectrometer according to claim 1, characterized in that, The CORDIC algorithm of S3 adopts a vector mode in circular coordinates. It rotates the complex plane vector to the real axis direction through iterative micro-rotation and combines the compensation factor to correct the amplitude attenuation, thus completing the high-precision solution of frequency domain complex signal to frequency domain amplitude data.

5. The storage and processing method for an FPGA-based integrated broadband Fourier transform microwave spectrometer according to claim 1, characterized in that, The dual-port RAM of the S4 is set to a storage depth of 4096 points to adapt to the output scale of frequency domain amplitude data. The finite state machine realizes automatic data accumulation and output by switching between four states: idle initialization, waiting judgment, accumulation operation, and result output.

6. A system for storing and processing data in an integrated broadband Fourier transform microwave spectrometer based on FPGA, as described in any one of claims 1-5, characterized in that... Includes the following modules: The data storage module is used to achieve bit width adaptation and cross-clock domain synchronization between ADC acquisition data and DDR3 memory through asynchronous FIFO. The 16-bit time domain signal acquired by ADC is written into DDR3 memory in sequence, and then read out from DDR3 memory in preset batches and sent to the subsequent signal processing link. The Fourier transform processing module is used to convert the time-domain signal of the input signal processing link into the frequency domain using the radix-4 FFT algorithm, combined with a ping-pong storage structure and a pipelined butterfly operation unit, to generate a frequency-domain complex signal. The amplitude calculation module is used to calculate the magnitude of the frequency domain complex signal using the CORDIC algorithm. The magnitude calculation is specifically completed through pure shift and addition operations to obtain frequency domain amplitude data. The cumulative averaging module is used to perform at least 4096 cumulative operations on the frequency domain amplitude data through a dual-port RAM and a finite state machine, and output the average result after the accumulation is completed.

7. The system for storage and processing of an integrated broadband Fourier transform microwave spectrometer based on FPGA according to claim 6, characterized in that, The Fourier transform processing module includes the following modules: The RAM storage module is used to store time-domain signals, FFT intermediate results, and frequency-domain complex signals. Two sets of dual-port RAMs are accessed in parallel, and ping-pong operation is used to prevent collisions. The rotation factor storage module is used to pre-store the rotation factor required for the radix-4 FFT, and three ROMs are output in parallel to adapt to the operation timing. The radix-4 butterfly operation module is used to receive four data inputs from the RAM storage module and three rotation factors input from the rotation factor storage module, perform butterfly operations, and output intermediate results. The storage management module is used to generate conflict-free addresses, control RAM switching, schedule data transmission, and control the six-level extraction operation process.

8. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the program, it implements the method as described in any one of claims 1 to 7.

9. A non-transitory computer-readable storage medium, characterized in that, The non-transitory computer-readable storage medium stores computer instructions for causing the computer to perform the method according to any one of claims 1 to 7.

10. A computer program product comprising computer program instructions, characterized in that, When the computer program instructions are executed on a computer, the computer causes the computer to perform the method as described in any one of claims 1 to 7.