APSK soft demodulation method based on decision feedback mechanism

Abstract

This invention relates to the field of communication technology, specifically to an APSK soft demodulation method based on a decision feedback mechanism. The method includes the following steps: S1, for each constellation point of the APSK modulated signal, a unique decision region is defined according to the distribution characteristics of amplitude and phase; S2, the received APSK modulated signal is acquired, the amplitude and phase values ​​of the receiving point are calculated, all constellation reference points within the decision region are determined, and a reference point set is constructed; S3, the average phase of all reference points in the reference point set of a set length is calculated to obtain the corrected phase value of the receiving point; S4, the original amplitude of the receiving point is retained, and the Euclidean distance between the receiving point and each reference point in the reference point set is calculated; S5, the minimum Euclidean distance between the receiving point and each reference point in the reference point set is determined, completing the soft demodulation mapping of the APSK modulated signal. This invention solves the technical problems of high computational complexity and low demodulation accuracy in existing soft demodulation methods.

Description

Technical Field

[0001] This invention relates to the field of communication technology, and specifically to an APSK soft demodulation method based on a decision feedback mechanism. Background Technology

[0002] Amplitude Phase Shift Keying (APSK), as a high-order digital modulation method, has become the mainstream modulation technology in satellite communication, digital broadcasting, and other fields due to its high spectral efficiency. The high reliability requirements of communication systems make soft demodulation a core component of APSK receivers. Soft demodulation outputs the log-likelihood ratio, reflecting the confidence level of bit demodulation, providing soft information for subsequent channel decoding such as LDPC and Turbo. Compared to hard demodulation, which only outputs bit values, it significantly improves the system's anti-interference capability and reduces the overall bit error rate, making it a key technological support for achieving stable transmission of high-order APSK.

[0003] Existing APSK soft demodulation methods are mainly divided into two categories. One category is a demodulation scheme based on carrier recovery loop, which completes carrier synchronization and phase compensation through Costas loop, phase-locked loop, etc., and then calculates the Euclidean distance between the receiver point and the constellation reference point, and then solves the log-likelihood ratio to achieve soft demodulation. The other category is a simplified hard-decision feedback demodulation scheme, which makes simple corrections to the phase of the receiver point through the hard decision result of a single reference point, reducing complex calculations. Some schemes also simplify the calculation process of log-likelihood ratio by optimizing the constellation point traversal method.

[0004] While carrier recovery loop-based schemes offer high demodulation accuracy, the loop involves complex operations such as filtering and integration, resulting in high computational complexity. This hinders engineering implementation on platforms like FPGAs and DSPs, and also limits data throughput. Simple hard-decision feedback schemes reduce complexity, but single-reference-point correction is susceptible to channel noise and phase shift, leading to low phase compensation accuracy and a relatively high demodulation bit error rate. Existing methods fail to simultaneously meet the requirements of low computational complexity, high demodulation accuracy, and flexible engineering adaptability, becoming a bottleneck for the practical application of high-order APSK modulation technology. Therefore, there is an urgent need for an APSK soft demodulation method that balances demodulation accuracy and computational complexity while being adaptable to multi-platform engineering implementation. Summary of the Invention

[0005] This invention provides an APSK soft demodulation method based on a decision feedback mechanism to solve the technical problems of high computational complexity and low demodulation accuracy in existing soft demodulation methods.

[0006] To address the above problems, the APSK soft demodulation method based on a decision feedback mechanism provided by this invention adopts the following technical solution: An APSK soft demodulation method based on a decision feedback mechanism includes the following steps: S1. For the constellation points of the APSK modulated signal, a unique decision region is divided according to the distribution characteristics of amplitude and phase. Each decision region corresponds to a preset constellation reference point range. S2. Acquire the received APSK modulated signal, calculate the amplitude and phase values ​​of the receiving point, match the decision region to which it belongs based on the amplitude and phase values, determine all constellation reference points within the decision region, and construct a reference point set. S3. Select the set of reference points of a set length, calculate the average phase of all reference points in the set of reference points of the set length, and use the average phase to compensate and correct the original phase value of the receiving point to obtain the corrected phase value of the receiving point. S4. Retain the original amplitude of the receiving point, and calculate the Euclidean distance between the receiving point and each reference point in the reference point set by combining the corrected phase value of the receiving point; S5: Determine the minimum Euclidean distance between the receiving point and each reference point in the reference point set, and calculate the log-likelihood ratio of each bit based on the minimum Euclidean distance to complete the soft demodulation mapping of the APSK modulated signal.

[0007] The beneficial effects of the above technical solution are: accurate phase compensation of the receiving point is achieved by averaging the phase of the reference point set, effectively reducing the demodulation error rate; the calculation range is narrowed by defining the decision region, simplifying the core calculation logic and significantly reducing the overall computational complexity; the log-likelihood ratio is accurately calculated based on the minimum Euclidean distance, and reliable soft information is output to support subsequent decoding. Overall, it balances demodulation accuracy and computational efficiency, improves the data throughput and application flexibility of APSK soft demodulation, and is adaptable to multi-platform engineering implementation.

[0008] Further, in step S1, the amplitude is divided into continuous intervals corresponding to the amplitude rings according to the number of amplitude rings distributed at the APSK adjustment signal constellation points. Then, the phase domain corresponding to each amplitude ring is evenly divided into several phase sectors. The two-dimensional intersection of the amplitude intervals and the phase sectors constitutes the decision region.

[0009] The beneficial effects of the above technical solution are: the division method conforms to the distribution characteristics of the amplitude loop and phase sector of the APSK constellation, making the decision area boundary clear and unique, and realizing fast and accurate matching of the constellation range to which the receiving point belongs; it can be flexibly adjusted according to the number of APSK amplitude loops to adapt to different orders of APSK modulation; at the same time, it greatly reduces the subsequent reference point calculation range, reduces invalid calculations, and improves the calculation efficiency of reference point matching and subsequent demodulation steps.

[0010] Further, in step S2, the received APSK modulated baseband complex signal is converted into polar coordinates, the amplitude of the receiving point is obtained by modulus calculation, and the phase value of the receiving point is obtained by argument calculation; the reference point set is all constellation reference points within the decision region to which the receiving point belongs, and each reference point in the reference point set carries corresponding bit mapping information.

[0011] The beneficial effects of the above technical solution are: converting the baseband complex signal to polar coordinates to calculate the amplitude and phase is simple and accurately reflects the signal characteristics, reducing parameter calculation errors; the reference point set consists of all constellation points in the decision region, providing effective statistical samples for subsequent average phase calculation and improving phase correction accuracy; the reference points carry bit mapping information, reducing subsequent information completion calculations for soft demodulation and improving the continuity and computational efficiency of the demodulation process.

[0012] Furthermore, in step S3, a sliding window is used to select a subset of reference points of length N from the continuously acquired set of reference points, where N is a positive integer and can be flexibly adjusted according to the channel environment; when calculating the average phase, the phase values ​​of each reference point in the subset of reference points are first normalized to [0, 2π], and then the normalized phase values ​​are arithmetically averaged to obtain the average phase.

[0013] The beneficial effects of the above technical solution are: the sliding window selects a subset of reference points of a set length, and the length N can be flexibly adjusted to match different channel environments, achieving a balance between phase correction effect and demodulation real-time performance; the phase value is first normalized and then arithmetically averaged to avoid cross-interval calculation deviation, improve the accuracy of average phase, provide a reliable benchmark for phase compensation at the receiving point, and ensure the phase correction effect.

[0014] Furthermore, in step S3, when performing compensation correction, the original phase value of the receiving point is compensated for by the difference between the average phase. If the compensated phase value exceeds the range of [0, 2π], phase folding is performed on it to obtain the corrected phase value of the receiving point, and the original amplitude of the receiving point remains unchanged during the correction process.

[0015] The beneficial effects of the above technical solution are: it accurately cancels the phase shift of the receiving point through differential compensation, thereby improving phase accuracy; it folds out-of-range phase values ​​to avoid subsequent calculation errors and ensures a smooth demodulation process; it retains the original amplitude during phase correction, fully preserves the signal amplitude characteristics, avoids the introduction of additional errors, and improves the accuracy of subsequent Euclidean distance calculation.

[0016] Furthermore, in step S4, when calculating the Euclidean distance, the receiver point with the original amplitude and the phase value combined with the corrected phase value, as well as each reference point in the reference point set, are all converted into rectangular coordinate form. Then, according to the rectangular coordinate Euclidean distance calculation formula, the Euclidean distance between the receiver point and each reference point is calculated respectively.

[0017] The beneficial effects of the above technical solution are: it uniformly converts to rectangular coordinates to calculate Euclidean distance, which conforms to the classic calculation logic, simplifies the operation and reduces the computational complexity, and adapts to engineering calculation needs; it eliminates the deviation caused by inconsistent coordinate forms, makes the distance calculation consistent, improves the accuracy of Euclidean distance calculation, and provides a reliable data foundation for subsequent minimum distance selection.

[0018] Furthermore, in step S5, when determining the minimum Euclidean distance between the receiving point and each reference point in the set of reference points, all Euclidean distances obtained in step S4 are sorted and filtered to obtain the minimum Euclidean distance value. At the same time, the target reference point corresponding to the minimum Euclidean distance value is recorded, and the bit mapping relationship of the target reference point is extracted.

[0019] The beneficial effects of the above technical solution are: it can quickly locate the minimum Euclidean distance by sorting and filtering, which is efficient and adaptable to the demodulation requirements of high data throughput; it can simultaneously record the target reference point and extract the bit mapping relationship, which directly provides core information for subsequent log-likelihood ratio calculation, avoids secondary retrieval and extraction, reduces redundant calculations, and improves the efficiency of the software demodulation process.

[0020] Further, in step S5, when calculating the log-likelihood ratio of each bit, the difference between the minimum Euclidean distance value and the corresponding Euclidean distance values ​​of other reference points in the reference point set is calculated. Combined with the preset channel noise variance, the log-likelihood ratio of each bit when it is 0 and 1 is solved according to the soft demodulation log-likelihood ratio calculation model. The absolute value of the log-likelihood ratio reflects the confidence level of the demodulation result of the corresponding bit.

[0021] The beneficial effects of the above technical solution are: combining the Euclidean distance difference with the channel noise variance to calculate the log-likelihood ratio, making the soft information more consistent with the actual demodulation situation and improving the accuracy of the log-likelihood ratio; and correlating the absolute value with the demodulation confidence to provide quantitative soft information support for subsequent channel decoding, helping to correct errors in the decoding process and improving the anti-interference capability of the entire communication system.

[0022] Furthermore, the method is implemented in engineering applications based on a hardware or software platform, which includes one or more of FPGA, DSP, and CPU. If it is implemented based on an FPGA platform, a parallel computing architecture is used to implement the hardware logic of the operation process in steps S2 to S5. If it is implemented based on a DSP platform, a fixed-point operation optimization strategy is used to reduce the computational complexity.

[0023] The beneficial effects of the above technical solutions are: they support implementation on multiple platforms such as FPGA, DSP, and CPU, breaking the limitations of a single platform and improving the flexibility and adaptability of engineering applications; for FPGA, a parallel computing architecture is adopted to give full play to its hardware advantages, improve demodulation speed and increase data throughput; for DSP, fixed-point computing optimization is adopted to reduce hardware resource consumption, improve computing efficiency, and adapt to engineering scenarios with limited resources.

[0024] Furthermore, the APSK modulation order adapted to this method includes one or more higher-order APSKs such as 16APSK, 32APSK, and 64APSK.

[0025] The beneficial effects of the above technical solution are: by simply adjusting the decision region division rules to adapt to different high-order APSKs, without modifying the core demodulation steps, it improves the compatibility and adaptability to different order modulations and reduces the system deployment cost and difficulty.

[0026] The beneficial effects of the APSK soft demodulation method based on a decision feedback mechanism provided by this invention are as follows: Accurate phase compensation at the receiving point is achieved by averaging the phase of the reference point set, effectively offsetting errors caused by channel noise and phase offset, and significantly reducing the demodulation bit error rate. Simultaneously, relying on the amplitude-phase decision region division of constellation points, the computational range is significantly reduced, the core computational logic is simplified, and the overall computational complexity is significantly reduced. The output log-likelihood ratio has higher accuracy than the soft information, providing reliable support for subsequent channel decoding. The solution is adaptable to multiple platforms such as FPGA, DSP, and CPU for engineering implementation, supports multiple high-order APSK modulations such as 16APSK and 32APSK, improves the data throughput and application flexibility of demodulation, and can meet the actual transmission needs of scenarios such as satellite communication and digital broadcasting.

[0027] In summary, through the above-mentioned design, this invention solves the technical problems of high computational complexity and low demodulation accuracy in existing soft demodulation methods. Attached Figure Description

[0028] Figure 1 This is a flowchart of the APSK soft demodulation method based on a decision feedback mechanism provided by the present invention. Detailed Implementation

[0029] The principles and spirit of the present invention will be explained in detail below with reference to several representative embodiments.

[0030] An embodiment of the APSK soft demodulation method based on a decision feedback mechanism provided by this invention: like Figure 1 As shown, an APSK soft demodulation method based on a decision feedback mechanism includes the following steps: S1. For the constellation points of the APSK modulated signal, a unique decision region is divided according to the distribution characteristics of amplitude and phase. Each decision region corresponds to a preset constellation reference point range. S2. Acquire the received APSK modulated signal, calculate the amplitude and phase values ​​of the receiving point, match the decision region to which it belongs based on the amplitude and phase values, determine all constellation reference points within the decision region, and construct a reference point set. S3. Select the set of reference points of a set length, calculate the average phase of all reference points in the set of reference points of the set length, and use the average phase to compensate and correct the original phase value of the receiving point to obtain the corrected phase value of the receiving point. S4. Retain the original amplitude of the receiving point, and calculate the Euclidean distance between the receiving point and each reference point in the reference point set by combining the corrected phase value of the receiving point; S5: Determine the minimum Euclidean distance between the receiving point and each reference point in the reference point set, and calculate the log-likelihood ratio of each bit based on the minimum Euclidean distance to complete the soft demodulation mapping of the APSK modulated signal.

[0031] Specifically, in step S1, based on the number of amplitude rings distributed among the APSK adjustment signal constellation points, the amplitude is divided into continuous intervals corresponding one-to-one with the amplitude rings. Then, the phase domain corresponding to each amplitude ring is evenly divided into several phase sectors. The two-dimensional intersection of the amplitude intervals and the phase sectors constitutes the decision region. This division method closely matches the distribution characteristics of the APSK constellation point amplitude rings and phase sectors, ensuring clear boundaries and unique attribution of the decision region, achieving rapid and accurate matching of the constellation range to which the receiving point belongs. It can be flexibly adjusted according to the number of APSK amplitude rings to adapt to different orders of APSK modulation. Simultaneously, it significantly reduces the subsequent reference point calculation range, decreases invalid calculations, and improves the computational efficiency of reference point matching and subsequent demodulation steps.

[0032] In step S2, the received APSK modulated baseband complex signal is converted into polar coordinates. The amplitude of the receiving point is obtained through modulus calculation, and the phase value of the receiving point is obtained through argument calculation. The reference point set consists of all constellation reference points within the decision region to which the receiving point belongs, and each reference point in the reference point set carries corresponding bit mapping information. Converting the baseband complex signal to polar coordinates for amplitude and phase calculation is simple and accurately reflects the signal characteristics, reducing parameter calculation errors. The reference point set consists of all constellation points within the decision region, providing effective statistical samples for subsequent average phase calculation and improving phase correction accuracy. The reference points carrying bit mapping information reduce subsequent information completion calculations for soft demodulation, improving the coherence and computational efficiency of the demodulation process.

[0033] In step S3, a sliding window is used to select a subset of reference points of length N from the continuously acquired set of reference points. N is a positive integer and can be flexibly adjusted according to the channel environment. When calculating the average phase, the phase values ​​of each reference point in the subset are first normalized to [0, 2π], and then the normalized phase values ​​are arithmetically averaged to obtain the average phase. The sliding window selects a subset of reference points of a set length, and the length N can be flexibly adjusted to match different channel environments, achieving a balance between phase correction effect and demodulation real-time performance. The phase values ​​are normalized first and then arithmetically averaged to avoid cross-interval calculation errors, improve the accuracy of the average phase, provide a reliable benchmark for phase compensation at the receiving point, and ensure the phase correction effect.

[0034] Furthermore, in step S3, during compensation correction, the original phase value of the receiving point is compensated for by the difference between the original phase value and the average phase. If the compensated phase value exceeds the range of [0, 2π], phase folding is performed to obtain the corrected phase value of the receiving point, while maintaining the original amplitude of the receiving point unchanged during the correction process. This differential compensation accurately offsets the phase shift of the receiving point, improving phase accuracy; folding is performed on out-of-range phase values ​​to avoid subsequent calculation errors and ensure a smooth demodulation process; and the original amplitude is preserved during phase correction, fully retaining the signal amplitude characteristics, avoiding the introduction of additional errors, and improving the accuracy of subsequent Euclidean distance calculations.

[0035] In step S4, when calculating the Euclidean distance, the receiver points retaining the original amplitude and combined with the corrected phase values, as well as all reference points in the reference point set, are all converted to rectangular coordinates. Then, according to the rectangular coordinate Euclidean distance calculation formula, the Euclidean distance between the receiver point and each reference point is calculated separately. This unified conversion to rectangular coordinates for Euclidean distance calculation aligns with classic calculation logic, simplifies the operation, reduces computational complexity, and adapts to engineering calculation needs. It eliminates deviations caused by inconsistent coordinate forms, ensuring consistency in distance calculation, improving the accuracy of Euclidean distance calculation, and providing a reliable data foundation for subsequent minimum distance selection.

[0036] In step S5, when determining the minimum Euclidean distance between the receiving point and each reference point in the reference point set, all Euclidean distances obtained in step S4 are sorted and filtered to obtain the minimum Euclidean distance value. Simultaneously, the target reference point corresponding to the minimum Euclidean distance value is recorded, and the bit mapping relationship of the target reference point is extracted. This sorting and filtering method quickly locates the minimum Euclidean distance, resulting in high computational efficiency and adapting to high data throughput demodulation requirements. Simultaneously recording the target reference point and extracting the bit mapping relationship directly provides core information for subsequent log-likelihood ratio calculations, avoiding secondary retrieval and extraction, reducing redundant calculations, and improving the overall efficiency of the software demodulation process.

[0037] Simultaneously, in step S5, when calculating the log-likelihood ratio for each bit, the difference between the minimum Euclidean distance value and the corresponding Euclidean distance values ​​of other reference points in the reference point set is calculated. Combined with the preset channel noise variance, and based on the soft demodulation log-likelihood ratio calculation model, the log-likelihood ratio for each bit being 0 and 1 is solved respectively. The absolute value of the log-likelihood ratio reflects the confidence level of the demodulation result for the corresponding bit. Calculating the log-likelihood ratio by combining the Euclidean distance difference with the channel noise variance makes the soft information more closely reflect the actual demodulation situation, improving the accuracy of the log-likelihood ratio. Correlating the absolute value with the demodulation confidence level provides quantitative soft information support for subsequent channel decoding, assisting in error correction in the decoding process and improving the anti-interference capability of the entire communication system.

[0038] Finally, the method of this invention is implemented in engineering applications based on a hardware or software platform, which includes one or more of FPGA, DSP, and CPU. If implemented on an FPGA platform, a parallel computing architecture is used to implement the hardware logic of the computation process in steps S2 to S5. If implemented on a DSP platform, a fixed-point computation optimization strategy is used to reduce computational complexity. It supports implementation on multiple platforms such as FPGA, DSP, and CPU, breaking the limitations of a single platform and improving the flexibility and adaptability of engineering applications. For FPGA, a parallel computing architecture is used to fully leverage its hardware advantages, improving demodulation speed and increasing data throughput. For DSP, fixed-point computation optimization is used to reduce hardware resource consumption, improve computational efficiency, and adapt to resource-constrained engineering scenarios.

[0039] The method of this invention is compatible with one or more high-order APSK modulation orders, including 16APSK, 32APSK, and 64APSK. By simply adjusting the decision region division rules, it adapts to different high-order APSKs without modifying the core demodulation steps, improving compatibility with different modulation orders and reducing system deployment costs and complexity.

[0040] The working principle of the APSK soft demodulation method based on a decision feedback mechanism provided by this invention is as follows: First, for the constellation points of the APSK modulated signal, a unique decision region is divided according to the amplitude and phase distribution characteristics. Each region corresponds to a preset constellation reference point range. After acquiring the received APSK modulated signal, it is converted into polar coordinates to calculate the amplitude and phase values, matched to the decision region, and a set of reference points within that region is constructed. A sliding window is used to select a subset of reference points of a set length. After phase normalization, the average phase is calculated. This average phase is used to compensate for the difference in the original phase value of the received point. Phase values ​​outside the range are folded while retaining the original amplitude. Subsequently, both the received point and the reference points are converted into rectangular coordinates, and the Euclidean distance between them is calculated. The minimum Euclidean distance is selected and the corresponding target reference point is recorded. Finally, combined with the channel noise variance, the log-likelihood ratio of each bit is calculated with the minimum Euclidean distance as the core, completing the soft demodulation mapping of the APSK modulated signal.

[0041] Based on the above description in this specification, those skilled in the art will also understand that the following terms, such as "upper," "lower," "front," "rear," "left," "right," "width," "horizontal," "top," "bottom," "inner," and "outer," which indicate orientation or positional relationships, are based on the orientation or positional relationships shown in the accompanying drawings of this specification. They are only for the purpose of facilitating the explanation of the present invention and simplifying the description, and do not explicitly or implicitly suggest that the device or element involved must have the specific orientation, or be constructed and operated in a specific orientation. Therefore, the above-mentioned orientation or positional relationship terms should not be understood or interpreted as limitations on the present invention.

[0042] In addition, in the description of this specification, "multiple" means at least two, such as two, three or more, etc., unless otherwise expressly and specifically defined.

Claims

1. An APSK soft demodulation method based on a decision feedback mechanism, characterized in that, Includes the following steps: S1. For the constellation points of the APSK modulated signal, a unique decision region is divided according to the distribution characteristics of amplitude and phase. Each decision region corresponds to a preset constellation reference point range. S2. Acquire the received APSK modulated signal, calculate the amplitude and phase values ​​of the receiving point, match the decision region to which it belongs based on the amplitude and phase values, determine all constellation reference points within the decision region, and construct a reference point set. S3. Select the set of reference points of a set length, calculate the average phase of all reference points in the set of reference points of the set length, and use the average phase to compensate and correct the original phase value of the receiving point to obtain the corrected phase value of the receiving point. S4. Retain the original amplitude of the receiving point, and calculate the Euclidean distance between the receiving point and each reference point in the reference point set by combining the corrected phase value of the receiving point; S5: Determine the minimum Euclidean distance between the receiving point and each reference point in the reference point set, and calculate the log-likelihood ratio of each bit based on the minimum Euclidean distance to complete the soft demodulation mapping of the APSK modulated signal.

2. The APSK soft demodulation method based on a decision feedback mechanism according to claim 1, characterized in that: In step S1, the amplitude is divided into continuous intervals corresponding to the amplitude rings according to the number of amplitude rings distributed in the APSK adjustment signal constellation points. Then, the phase domain corresponding to each amplitude ring is evenly divided into several phase sectors. The two-dimensional intersection of the amplitude intervals and the phase sectors constitutes the decision region.

3. The APSK soft demodulation method based on a decision feedback mechanism according to claim 2, characterized in that: In step S2, the received APSK modulated baseband complex signal is converted into polar coordinates, the amplitude of the receiving point is obtained by modulus calculation, and the phase value of the receiving point is obtained by argument calculation; the reference point set is all constellation reference points in the decision region to which the receiving point belongs, and each reference point in the reference point set carries corresponding bit mapping information.

4. The APSK soft demodulation method based on a decision feedback mechanism according to any one of claims 1 to 3, characterized in that: In step S3, a sliding window is used to select a subset of reference points of length N from the continuously acquired set of reference points. N is a positive integer and can be flexibly adjusted according to the channel environment. When calculating the average phase, the phase values ​​of each reference point in the subset of reference points are first normalized to [0, 2π], and then the normalized phase values ​​are arithmetically averaged to obtain the average phase.

5. The APSK soft demodulation method based on a decision feedback mechanism according to claim 4, characterized in that: In step S3, when performing compensation correction, the original phase value of the receiving point is compensated for by the difference between the average phase. If the compensated phase value exceeds the range of [0, 2π], phase folding is performed on it to obtain the corrected phase value of the receiving point, and the original amplitude of the receiving point remains unchanged during the correction process.

6. The APSK soft demodulation method based on a decision feedback mechanism according to claim 5, characterized in that: In step S4, when calculating the Euclidean distance, the receiver point with the original amplitude and the phase value combined with the corrected phase value, as well as each reference point in the reference point set, are all converted into rectangular coordinates. Then, according to the rectangular coordinate Euclidean distance calculation formula, the Euclidean distance between the receiver point and each reference point is calculated respectively.

7. The APSK soft demodulation method based on a decision feedback mechanism according to claim 6, characterized in that: In step S5, when determining the minimum Euclidean distance between the receiving point and each reference point in the reference point set, all Euclidean distances obtained in step S4 are sorted and filtered to obtain the minimum Euclidean distance value. At the same time, the target reference point corresponding to the minimum Euclidean distance value is recorded, and the bit mapping relationship of the target reference point is extracted.

8. The APSK soft demodulation method based on a decision feedback mechanism according to claim 7, characterized in that: In step S5, when calculating the log-likelihood ratio of each bit, the difference between the minimum Euclidean distance value and the corresponding Euclidean distance values ​​of other reference points in the reference point set is calculated. Combined with the preset channel noise variance, the log-likelihood ratio of each bit when it is 0 and 1 is solved according to the soft demodulation log-likelihood ratio calculation model. The absolute value of the log-likelihood ratio reflects the confidence level of the demodulation result of the corresponding bit.

9. The APSK soft demodulation method based on a decision feedback mechanism according to claim 8, characterized in that: This method is implemented in engineering applications based on a hardware or software platform, which includes one or more of FPGA, DSP, and CPU. If it is implemented based on an FPGA platform, a parallel computing architecture is used to implement the hardware logic of the operation process from step S2 to S5. If it is implemented based on a DSP platform, a fixed-point operation optimization strategy is used to reduce the computational complexity.

10. The APSK soft demodulation method based on a decision feedback mechanism according to claim 9, characterized in that: The APSK modulation order that this method is compatible with includes one or more higher-order APSKs such as 16APSK, 32APSK, and 64APSK.

Images