一种基于三角模糊数层次分析理论的短波多通道分集融合方法

Abstract

本发明涉及短波通信领域，特别涉及一种基于三角模糊数层次分析理论的短波多通道分集融合方法，包括获取接收端解调后、译码前的符号软信息，并估计各个信道的多径时延、多普勒扩展以及信噪比参数；利用三角模糊层次分析法，确认各个信道的多径时延、多普勒扩展以及信噪比参数的影响程度值；根据各个信道的多径时延、多普勒扩展以及信噪比参数及其对应的影响程度值计算折扣因子；利用折扣因子对译码前的符号软信息进行加权，得到待融合的软信息；将各个信道的待融合软信息进行相加融合，得到软信息的判决输出，完成融合；本发明在信道情况复杂多变的短波通信上，能更好地利用信号进行相互补偿、抵抗衰落，实现更优的分集增益，提升通信的可靠性。

Description

Technical Field

[0001] This invention relates to the field of shortwave communication, and in particular to a shortwave multi-channel diversity fusion method based on the triangular fuzzy number hierarchical analysis theory. Background Technology

[0002] Among all wireless channels, shortwave channels are the most typical. Shortwave communication has two modes: ground wave and sky wave. Ground wave is suitable for short-range communication, while sky wave uses one or more reflections from the ionosphere to achieve communication over thousands of kilometers. Shortwave communication has many advantages, including simple equipment, ease of use, high mobility, low cost, fewer terrain limitations, rapid circuit setup, and strong resilience. However, due to the extreme instability of the atmospheric ionosphere, shortwave channels are affected by more factors than traditional communication channels. These are mainly reflected in the following aspects:

[0003] (I) Multipath Propagation

[0004] The concentration of electrons in the ionosphere varies at different times and altitudes. Combined with the ground's reflection of shortwave waves, the radio wave signal from the transmitting source travels through different propagation paths to the receiving end, resulting in multipath propagation of the shortwave signal. The impact of this multipath propagation is called the multipath effect. The multipath effect refers to the phenomenon that signals arrive at the distant receiving end with different time delays as they travel along each path. These signals become out of phase due to the different time delays, and the attenuation loss varies along the different paths, leading to mutual interference between the signals at the receiving end. Therefore, multipath delay is one of the core parameters of shortwave channels and a significant cause of signal distortion.

[0005] (II) Doppler Extension

[0006] The time-varying nature of the ionosphere causes multipath signals to superimpose in phase or out of phase at the receiver, resulting in abnormal fluctuations in the signal envelope. Simultaneously, as the phase changes over time, frequency fluctuations inevitably occur, leading to spectral spread in the frequency domain. This phenomenon, known as Doppler spread, depends on changes in ionospheric characteristics, and the Doppler spread value reflects the severity of shortwave channel degradation.

[0007] (III) Noise and Interference

[0008] For every communication system, noise and interference, as limiting factors of the channel, determine whether the signal can be used for information transmission. Understanding the various interferences and noises encountered during signal transmission is crucial to transmitting signals with sufficient power to overcome these effects. Statistically, noise and interference in the shortwave band mainly include atmospheric noise, man-made noise, and interference from shortwave radio stations. Besides these external interferences, there are also internal interferences and noise generated by the receiving equipment itself, such as transmitter intermodulation interference and receiver intermodulation interference. External interference is the main impact on long-distance shortwave communication, and its noise impact can be described by the signal-to-noise ratio (SNR) parameter of the received signal.

[0009] Among numerous anti-interference technologies, diversity reception is the most widely used and easiest to implement. It utilizes the principle that multiple signals cannot simultaneously be in deep fading to achieve complementarity between signals, obtain diversity gain, and reduce the communication error rate. As diversity reception technology matures, it is gradually being applied to shortwave communication. Considering the characteristics of shortwave spectrum resource degradation and slow transmission rate, shortwave wide-area diversity reception technology uses each distributed station to receive and demodulate the signal individually, and then transmits its symbol information to the fusion center for merging. This can greatly reduce the amount of waveform information transmitted through the link and also reduce the processing complexity of the fusion center.

[0010] However, the weight allocation methods used in classic diversity fusion techniques become significantly less effective when directly applied to shortwave. Specifically, classic diversity fusion methods such as selection combining, equal-gain combining, and maximum ratio combining all rely solely on the signal-to-noise ratio (SNR) for weight allocation. This fails to adequately characterize the time-varying and fading characteristics of shortwave channels, resulting in weights that are not entirely proportional to channel quality. Consequently, the fusion process cannot fully utilize the information from all channels. Summary of the Invention

[0011] To address the problem of unreasonable weight allocation caused by the poor channel characteristics and differentiated channel features of shortwave, this invention proposes a shortwave multi-channel diversity fusion method based on triangular fuzzy number hierarchical analysis theory, which includes the following steps:

[0012] The symbol soft information is obtained after demodulation and before decoding at the receiver, and the multipath delay, Doppler spread and signal-to-noise ratio parameters of each channel are estimated.

[0013] The influence of multipath delay, Doppler spread, and signal-to-noise ratio parameters on each channel was determined using the triangular fuzzy hierarchical analysis method.

[0014] The discount factor is calculated based on the multipath delay, Doppler spread, and signal-to-noise ratio parameters of each channel and their corresponding impact values.

[0015] The soft information of the symbols before decoding is weighted using a discount factor to obtain the soft information to be fused;

[0016] The soft information to be fused from each channel is added together and fused to obtain the decision output of the soft information, thus completing the fusion.

[0017] Furthermore, the process of using triangular fuzzy hierarchical analysis to determine the degree of influence of multipath delay, Doppler spread, and signal-to-noise ratio parameters in each channel includes the following steps:

[0018] Construct a fuzzy judgment matrix, where the value in the i-th row and j-th column represents the importance of the i-th parameter relative to the j-th parameter. Based on the judgments of multiple decision-makers, construct a comprehensive triangular fuzzy matrix.

[0019] The degree of integration of each decision factor is calculated based on the comprehensive triangular fuzzy matrix;

[0020] A weight vector is constructed based on the degree values ​​between ambiguity numbers, and the weight vector is normalized to obtain the degree values ​​of the influence of multipath delay, Doppler spread, and signal-to-noise ratio parameters of each channel.

[0021] Furthermore, if there exist n parameters {a1, a2, ..., a...} that influence the final decision... n}, then the fuzzy judgment matrix of the p-th decision-maker is represented as:

[0022]

[0023] Among them, A p This represents the fuzzy judgment matrix for the p-th decision-maker. Represents the fuzzy judgment matrix A p The element in the i-th row and j-th column.

[0024] Furthermore, the fuzzy judgment matrix A p The element in the i-th row and j-th column is represented as in This represents the upper bound of the triangular fuzzy function between two evaluation criteria. A numerical measure representing the relative importance between two evaluation criteria. This represents the lower bound of the triangular fuzzy function between the two evaluation criteria.

[0025] Furthermore, if the i-th indicator and the j-th indicator are equally important, the numerical measure of the relative importance between the two evaluation criteria is 1; if the i-th indicator is slightly more important than the j-th indicator, the numerical measure of the relative importance between the two evaluation criteria is 3; if the importance between the i-th indicator and the j-th indicator is between equally important and slightly important, the numerical measure of the relative importance between the two evaluation criteria is 2.

[0026] If the i-th indicator is more important than the j-th indicator, the numerical measure of the relative importance between the two evaluation criteria is 5; if the importance between the i-th indicator and the j-th indicator is between slightly important and important, the numerical measure of the relative importance between the two evaluation criteria is 4.

[0027] If the i-th indicator is much more important than the j-th indicator, then the numerical measure of the relative importance between the two evaluation criteria is 7; if the importance between the i-th indicator and the j-th indicator is between important and much more important, then the numerical measure of the relative importance between the two evaluation criteria is 6.

[0028] If the i-th indicator is absolutely more important than the j-th indicator, then the numerical measure of relative importance between the two evaluation criteria is 9; if the importance between the i-th indicator and the j-th indicator is between much more important and much more important, then the numerical measure of relative importance between the two evaluation criteria is 8.

[0029] Furthermore, the degree of integration of each decision factor is calculated based on the comprehensive triangular fuzzy judgment matrix, including:

[0030]

[0031]

[0032]

[0033] Among them, b ij The element in the i-th row and j-th column of the combined triangular fuzzy matrix is ​​represented as b. ij =(l ij ,m ij ,u ij ).

[0034] Furthermore, constructing a weight vector based on the degree values ​​between fuzzy numbers includes:

[0035] W = [V(d1),V(d2),…,V(d...] n )] T

[0036]

[0037]

[0038] Where, m i The median value representing the degree of fuzziness of factor i, u i The lower bound of the fuzziness level of factor i, l k This represents the upper bound of the fuzziness level of factor i.

[0039] Compared to traditional diversity merging methods that only use the signal-to-noise ratio (SNR) as a weighting parameter, this invention considers multiple influencing parameters of the shortwave channel, treating them collectively as factors affecting the weighting. This allows for a more reasonable assessment of channel quality. Simultaneously, using triangular fuzzy hierarchical analysis (AHP), a weight score is obtained by comparing pairwise factors. This score is then multiplied by the corresponding channel parameter values ​​to obtain the weighted estimate β for each channel. Multiplying the soft information of each branch by this estimate allows for fusion and decoding. Therefore, this invention can better utilize signals for mutual compensation and fading resistance in shortwave communication with complex and variable channel conditions, achieving superior diversity gain and thus improving communication reliability. Attached Figure Description

[0040] Figure 1 This is a flowchart of the algorithm for a shortwave multi-channel diversity fusion method based on the triangular fuzzy number hierarchical analysis theory of the present invention.

[0041] Figure 2 This is a schematic diagram of the DSDFT process for receiving MFSK signals in this invention;

[0042] Figure 3 This is a schematic diagram of the time-frequency information of the received signal in this invention;

[0043] Figure 4 This is a schematic diagram of the coefficient matrix in this invention;

[0044] Figure 5 This is a schematic diagram of soft information extraction in this invention. Detailed Implementation

[0045] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. 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.

[0046] This invention proposes a shortwave multi-channel diversity fusion method based on triangular fuzzy number hierarchical analysis theory, which specifically includes the following steps:

[0047] The symbol soft information is obtained after demodulation and before decoding at the receiver, and the multipath delay, Doppler spread and signal-to-noise ratio parameters of each channel are estimated.

[0048] The influence of multipath delay, Doppler spread, and signal-to-noise ratio parameters on each channel was determined using the triangular fuzzy hierarchical analysis method.

[0049] The discount factor is calculated based on the multipath delay, Doppler spread, and signal-to-noise ratio parameters of each channel and their corresponding impact values.

[0050] The soft information of the symbols before decoding is weighted using a discount factor to obtain the soft information to be fused;

[0051] The soft information to be fused from each channel is added together and fused to obtain the decision output of the soft information, thus completing the fusion.

[0052] In this embodiment, as Figure 1 Assuming there are N channels transmitting information simultaneously, and all transmitters use the MFSK modulation method with strong anti-interference capabilities, the receiver processes the signal according to the steps described in this invention, specifically including the following steps:

[0053] (I) Determining the influence of channel parameters using TFN-AHP: W = [W1, W2, ... W N ]

[0054] There are many factors affecting the quality of shortwave signals. These factors need to be considered when fusing multipath information from shortwave signals. Therefore, determining the weights for signal fusion is a complex, multi-factor calculation process. The traditional Analytic Hierarchy Process (AHP) constructs a judgment matrix by comparing the relative importance of each pair of decision factors. It then uses mathematical theory to calculate the vector corresponding to the largest eigenvalue of the matrix. This vector represents the weight evaluation factor of that parameter, used to quantify the influence of factors on the decision. Its drawback is that it does not consider the absolute fuzziness of things when manually determining the importance of each pair of factors. In fuzzy decision-making or fuzzy optimization, fuzzy numbers are often used to handle similar boundary problems. Therefore, introducing the constraint method of fuzzy mathematics into the AHP method, forming TFN-AHP, can fully consider the fuzziness and uncertainty of individual judgments, thereby optimizing the construction of the judgment matrix. This embodiment uses the following steps to obtain the influence value of channel parameters, including:

[0055] A. Construction of the fuzzy judgment matrix

[0056] Construct a fuzzy judgment matrix based on the importance of factor i compared to factor j. ij Represented as a ij =[l ij,m ij ,u ij ], a ij Let l be a triangular fuzzy number. ij u ij Let m be the upper and lower bounds of the triangular fuzzy number. ij The numerical measure of the relative importance between the two evaluation criteria in the AHP method is shown in Table 1 below.

[0057] Table 1 Definition of AHP Evaluation Criteria

[0058]

[0059] According to the AHP evaluation scale definition in Table 1, if the i-th and j-th indicators are equally important, the numerical measure of relative importance between the two evaluation criteria is 1; if the i-th indicator is slightly more important than the j-th indicator, the numerical measure of relative importance between the two evaluation criteria is 3; if the importance of the i-th and j-th indicators is between equally important and slightly important, the numerical measure of relative importance between the two evaluation criteria is 2; if the i-th indicator is more important than the j-th indicator, the numerical measure of relative importance between the two evaluation criteria is 5; if the importance of the i-th and j-th indicators is between slightly important and important... If the i-th indicator is significantly more important than the j-th indicator, the numerical measure of relative importance between the two evaluation criteria is 4; if the i-th indicator is significantly more important than the j-th indicator, the numerical measure of relative importance between the two evaluation criteria is 7; if the importance of the i-th indicator and the j-th indicator is between important and significantly more important, the numerical measure of relative importance between the two evaluation criteria is 6; if the i-th indicator is absolutely more important than the j-th indicator, the numerical measure of relative importance between the two evaluation criteria is 9; if the importance of the i-th indicator and the j-th indicator is between significantly more important and significantly more important, the numerical measure of relative importance between the two evaluation criteria is 8.

[0060] Suppose there are n parameters {a1, a2, ..., a...} that influence the final decision. n The fuzzy decision matrix can then be obtained.

[0061]

[0062] in, Let a be the triangular fuzzy number representing the importance of the i-th and j-th indicator parameters given by the p-th expert, where a ii = (1,1,1), indicating that the influencing factors are equal to the influencing factors for itself. It is the reciprocal of the fuzzy number, where p represents the decision-maker's number, p = 1, 2, ..., k.

[0063] By combining the opinions of multiple decision-makers, a comprehensive triangular fuzzy matrix B can be obtained, where b ij The value of the triangular fuzzy number is given by the following formula:

[0064]

[0065] The comprehensive triangular fuzzy matrix, constructed by combining the judgments of multiple decision-makers, is as follows:

[0066]

[0067] Where (l,l,l) represents the importance of factor i to factor i on the diagonal, that is, the importance of oneself compared to oneself. In this embodiment, the importance of oneself is set to (1,1,1).

[0068] B. Calculate the comprehensive value of the indicators

[0069] The degree of integration of each decision factor is calculated based on the comprehensive triangular fuzzy judgment matrix, as shown below:

[0070]

[0071]

[0072]

[0073] C. Comparison of fuzziness levels and calculation of weight vectors

[0074] The degree of ambiguity between the numbers is given by the following formula:

[0075]

[0076] in The calculation method is given by the following formula:

[0077]

[0078] in, A triangular fuzzy number form of the k-th factor. m i The median value representing the degree of fuzziness of factor i, u i The lower bound of the fuzziness level of factor i, l k This represents the upper bound of the fuzziness level of factor i.

[0079] That is, the weight vector is:

[0080] W = [V(d1),V(d2),…,V(d...] n )] T (9)

[0081] Normalize the weight vector:

[0082]

[0083] in:

[0084]

[0085] Finally, the weight vector of the influencing factors is obtained. Used for subsequent calculations.

[0086] (ii) The receiver demodulates the signals of each channel individually to obtain symbol soft information; and obtains the multipath delay, Doppler spread and signal-to-noise ratio parameters of each channel through ideal channel estimation, and constructs the parameter matrix of N channels.

[0087] Shortwave transmission often uses MFSK signals for modulation. The following describes the soft information extraction process of MFSK signals. The illustration uses M=8 for demonstration, specifically including:

[0088] A. After receiving the signal x(t), perform a short-time Fourier transform on it.

[0089] like Figure 2 When the receiving end receives the received signal x(t), it samples x(t) to obtain the numerical signal x(n);

[0090] The numerical signal x(n) is windowed using a window function, resulting in M ​​signals, where the m-th signal is represented as x. m (t), m∈{1,2,…,M};

[0091] For signal x m (t) Perform Discrete Fourier Transform (DFT) and then calculate the magnitude;

[0092] After obtaining the modulus, the spectrum of each signal segment is calculated, where the m-th signal is represented as X. m (W), m∈{1,2,…,M};

[0093] The time spectrum is obtained by combining and outputting each spectral signal segment.

[0094] B. The time-domain signal is processed according to symbol time T. s Segmentation is performed to obtain the spectral information of the corresponding code elements.

[0095] like Figure 3 On the timeline, according to the symbolic time T s Perform segmentation, f s The vertical axis represents the sampling frequency, therefore each cell on the vertical axis represents the frequency resolution f. s / N, where each cell on the horizontal axis represents a symbol time Ts. Figure 3 In the first column of data, the energy values ​​of the eight frequencies of the first symbol Ts are taken, namely Sym_Inf1, Sym_Inf2, Sym_Inf3, Sym_Inf4, Sym_Inf5, Sym_Inf6, Sym_Inf7, and Sym_Inf8.

[0096] Where N represents the number of Fourier transform points in time Ts.

[0097] C. Given the frequency set, map 0 to -1 to construct the coefficient matrix.

[0098] like Figure 4 The eight frequencies f0 to f7 are each represented by eight binary symbols, namely {000, 001, 010, 011, 100, 101, 110, 111}. By mapping symbol 0 to -1 and symbol 1 to 1 in each symbol, an 8×3 matrix is ​​obtained. For example, symbol 000 is mapped to {-1, -1, -1} and symbol 001 is mapped to {-1, -1, 1}.

[0099] D. By multiplying the symbol energy information of M frequencies with the coefficient matrix using matrix multiplication, log2 M bits of soft information values ​​are obtained.

[0100] like Figure 4 Multiply the spectral information of the symbol obtained in step B with the coefficient matrix obtained in step C to obtain the soft information value.

[0101] Each channel of the received signal obtains the multipath delay τ, Doppler spread fd, and signal-to-noise ratio (SNR) values ​​through an ideal channel estimation technique, and constructs an actual parameter matrix, namely:

[0102]

[0103] In the actual parameter matrix, the multipath delay τ, Doppler spread fd and signal-to-noise ratio SNR values ​​of each row are all estimated values ​​based on the ideal values ​​of the corresponding channels. In equation (12), N represents the number of channels.

[0104] (III) Calculation of fusion weights.

[0105] Based on the index weight W obtained in (1) i The specific value R of the index obtained by (Equation 10) and step (2) N×3 (Equation 12, abbreviated as R), through equation β i =W i ·R T The fusion weight β of each signal is obtained. iAnd multiply it with the corresponding channel soft information as shown in the following formula (13) to obtain the soft information m(1)' and m(0)' to be fused for each channel.

[0106]

[0107] In this algorithm, A and B represent bits 1 and 0, respectively. i represents the i-th channel; m i (A), m i (A) represents the raw soft information demodulated from the i-th channel; m i (A)'、m i (B)' is the soft information to be fused after being described by weights.

[0108] (iv) The soft information values ​​of each channel are added together to obtain the fused soft information. Specifically, this includes:

[0109]

[0110] (v) Judgment output.

[0111] The decoding decision output is as follows (15):

[0112]

[0113] 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 shortwave multi-channel diversity fusion method based on triangular fuzzy number hierarchical analysis theory, characterized in that, Includes the following steps: The symbol soft information is obtained after demodulation and before decoding at the receiver, and the multipath delay, Doppler spread and signal-to-noise ratio parameters of each channel are estimated. Using the triangular fuzzy hierarchical analysis method, the influence of multipath delay, Doppler spread, and signal-to-noise ratio parameters on each channel is determined, including the following steps: Construct a fuzzy judgment matrix, where the value in the i-th row and j-th column represents the importance of the i-th parameter relative to the j-th parameter. Based on the judgments of multiple decision-makers, construct a comprehensive triangular fuzzy matrix. The degree of integration of each decision factor is calculated based on the comprehensive triangular fuzzy matrix; A weight vector is constructed based on the degree values ​​between ambiguity numbers, and the weight vector is normalized to obtain the degree values ​​of the influence of multipath delay, Doppler spread, and signal-to-noise ratio parameters of each channel. If there are n parameters that influence the final decision Then the fuzzy judgment matrix of the p-th decision-maker is represented as: ； in, This represents the fuzzy judgment matrix for the p-th decision-maker. Represents the fuzzy judgment matrix The element in the i-th row and j-th column is represented as ,in , This represents the upper bound of the triangular fuzzy function between two evaluation criteria. A numerical measure representing the relative importance between two evaluation criteria. This represents the lower bound of the triangular fuzzy function between two evaluation criteria, where p represents the decision-maker's ID. ; By combining the opinions of multiple decision-makers, a comprehensive triangular fuzzy matrix B can be obtained, where... The value of the triangular fuzzy number is given by the following formula: ； The comprehensive triangular fuzzy matrix, constructed by combining the judgments of multiple decision-makers, is as follows: ； in, The diagonal line represents the importance of factor i to factor i; The degree of integration of each decision factor is calculated based on the comprehensive triangular fuzzy judgment matrix, as shown below: ； ； ； in, To synthesize the elements in the i-th row and j-th column of the triangular fuzzy matrix, it is represented as: ; The weight vector is constructed based on the degree values ​​between fuzzy numbers, including: ； ； in, The median value representing the degree of fuzziness of factor i. The lower bound representing the degree of fuzziness of factor i. The upper bound representing the degree of fuzziness of factor i; The discount factor is calculated based on the multipath delay, Doppler spread, and signal-to-noise ratio parameters of each channel and their corresponding impact values. Specifically, this includes: The multipath delay for each channel of the received signal is obtained using ideal channel estimation techniques. The Doppler spread fd and signal-to-noise ratio (SNR) are used to construct the actual parameter matrix, which is expressed as follows: ； In the actual parameter matrix, each row has a multipath delay. The Doppler spread fd and signal-to-noise ratio (SNR) values ​​are both ideal estimates based on the corresponding channels, where N represents the number of channels. The discount factor for each channel is obtained by weighting the actual parameter matrix using the aforementioned influence value; The soft information of the symbols before decoding is weighted using a discount factor to obtain the soft information to be fused; The soft information to be fused from each channel is added together and fused to obtain the decision output of the soft information, thus completing the fusion.

2. The shortwave multi-channel diversity fusion method based on triangular fuzzy number hierarchical analysis theory according to claim 1, characterized in that, If the i-th indicator and the j-th indicator are equally important, then the numerical measure of the relative importance between the two evaluation criteria is 1. If the i-th indicator is slightly more important than the j-th indicator, then the numerical measure of the relative importance between the two evaluation criteria is 3. If the importance of the i-th indicator and the j-th indicator is between equally important and slightly important, then the numerical measure of the relative importance between the two evaluation criteria is 2. If the i-th indicator is more important than the j-th indicator, the numerical measure of the relative importance between the two evaluation criteria is 5; if the importance between the i-th indicator and the j-th indicator is between slightly important and important, the numerical measure of the relative importance between the two evaluation criteria is 4. If the i-th indicator is much more important than the j-th indicator, then the numerical measure of the relative importance between the two evaluation criteria is 7. If the importance of the i-th indicator and the j-th indicator is between important and much more important, then the numerical metric for the relative importance between the two evaluation criteria is 6. If the i-th indicator is absolutely more important than the j-th indicator, then the numerical metric for the relative importance between the two evaluation criteria is 9. If the importance of the i-th indicator to the j-th indicator is between much more important and much more important, then the numerical metric for the relative importance between the two evaluation criteria is 8.

Images