A soft demodulation method for low signal-to-noise ratio 188-110A signal under large frequency offset

Abstract

The application discloses a kind of soft demodulation methods for low signal-to-noise ratio 188-110A signal under large frequency offset, it is related to communication technical field, comprising: S1, preamble sequence detection and preamble sequence coarse synchronization based on autocorrelation;S2, prepare baseband preamble reference modulation signal;S3, preamble sequence fine synchronization based on differential cumulative correlation;S4, M&M and L&R joint frequency offset estimation and frequency offset correction based on preamble reference signal;S5, according to D1, D2 value table in 188-110A protocol, the demodulation of interleaving indicating D1, D2;S6, the frame search and frame demodulation of data are executed;S7, after the frame constellation of data is disturbed, the soft information of data frame is obtained;S8, based on 188-110A protocol, the deinterleaving of data frame soft information is completed by corresponding bit storage and reading mode, then it is decoded.After compared with prior art, signal detection anti-frequency offset is improved from ±0.4% to ±100%, signal detection performance is improved about 3dB gain, signal demodulation decoding performance is improved about 2.6dB gain.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of communication technology, in particular to a soft demodulation method for 188-110A signal with low signal-to-noise ratio under large frequency offset. BACKGROUND

[0002] In short wave communication, 188-110 signal is a kind of waveform standard with good compatibility, strong anti-interference ability and high efficient data transmission ability, which is widely used in short wave radio equipment in recent years. It has great practical significance to study 188-110 signal, which can be used for electronic countermeasure and provide reference for further development of short wave wireless communication.

[0003] So far, the existing 188-110A signal processing algorithm mainly considers how to find and demodulate information bits under small frequency offset. With the rise of software radio, it becomes more and more common to flexibly change the signal transmission frequency point. Since the hardware frequency conversion and channelization algorithm causes frequency offset more or less when digitally down-converting signals of different frequency bands, the original way of optimizing hardware radio frequency parameters to reduce system frequency offset error becomes infeasible. There is an urgent need for a flexible and practical signal processing algorithm to detect and correct frequency offset to complete the interpretation of 188-110A signal. Although some documents use maximum likelihood method to detect and interpret 188-110A signal, the high computational complexity makes it difficult to realize in engineering application. In addition, in the process of electronic countermeasure and radio reconnaissance, it is often desired that the lower the lower limit of signal-to-noise ratio that the signal processing algorithm can handle, the better, which is more conducive to long-distance signal reconnaissance. SUMMARY

[0004] The present application aims to overcome the shortcomings of the prior art and provide a soft demodulation method for 188-110A signal with low signal-to-noise ratio under large frequency offset.

[0005] The purpose of the present application is achieved by the following technical solutions:

[0006] The present application discloses a soft demodulation method for 188-110A signal with low signal-to-noise ratio under large frequency offset, comprising the following steps:

[0007] S1, preamble sequence detection and preamble sequence coarse synchronization based on autocorrelation;

[0008] S2, preparing a baseband preamble reference modulation signal;

[0009] S3, preamble sequence fine synchronization based on differential cumulative correlation;

[0010] S4, M&M and L&R joint frequency offset estimation and frequency offset correction based on preamble reference signal;

[0011] S5, demodulate according to the D1, D2 value table in the 188-110A protocol, and interleave the indication of D1, D2;

[0012] S6, perform frame search and frame demodulation of the data;

[0013] S7, after the frame constellation of the data is descrambled, obtain the soft information of the data frame;

[0014] S8, based on the 188-110A protocol, complete the deinterleaving of the soft information of the data frame through the corresponding bit storage and reading mode, obtain the deinterleaved soft information of the data frame, and then code it.

[0015] Further, step S1 specifically comprises: setting the single-carrier 188-110A signal as a channelized digital signal For , indicating symbol-to-constellation mapping, indicating Gaussian white noise, indicating a root raised cosine filter, indicating residual frequency offset after digital down-conversion and channelization, indicating the sampling rate of the narrowband signal after channelization, and N' indicating the symbol period, indicating symbol count; performing normalized autocorrelation on , m indicating the first summation variable, indicating the complex conjugate of , indicating the complex conjugate of , to obtain the first autocorrelation result ; if there are consecutive correlation values that make , then the 188-110A signal is found, wherein indicates the correlation threshold, ; the starting sample position of the consecutive correlation values is taken as the coarse starting position of the signal ; if the 188-110A signal is not found, repeat step S1 until the 188-110A signal is found.

[0016] Preferably, step S2 specifically comprises: taking the first 9 of the 15 preamble sequence channel symbols, mapping the 9 channel symbols to modulation symbols according to the Hadamard sequence mapping table of the preamble channel symbols and modulation symbols in the 188-110A protocol, and repeating 4 times to obtain the preamble reference modulation symbols, wherein the preamble reference modulation symbols are first scrambled according to the corresponding positions of the preamble sequence, and then constellation-mapped, 4 times up-sampled, and shaped filtered according to the 8PSK constellation in the protocol to obtain the baseband preamble reference modulation signal. , the length of which is .

[0017] Preferably, step S3 specifically comprises: taking starting from first sample points to delay 2, 4, 8, 16, 32 pairs of normalization differential autocorrelation to obtain second autocorrelation results , wherein denotes the delay, k denotes the second summation independent variable; the differential conjugate product of delay i is The calculation formula of the differential conjugate product of delay is , is the complex conjugate of delay , i is the complex conjugate of delay , denotes the complex conjugate of delay i , is the differential conjugate product of delay The calculation formula of the differential conjugate product of delay is , is the complex conjugate of delay i , is the complex conjugate of delay If the maximum value of delay is greater than , the position of the maximum value is recorded as the accurate starting position of the signal preamble sequence ; otherwise, it is considered that the 188-110A signal is not found, and step S1 is returned.

[0018] Preferably, step S4 specifically comprises: taking starting from second sample points , performing M&M and L&R joint frequency offset estimation based on to calculate the residual frequency offset contained in the current signal , and the calculation formula is as follows:

[0019] ;

[0020] ;

[0021] ;

[0022] ;

[0023] ;

[0024] ;

[0025] ;

[0026] ;

[0027] wherein, , , N is half of, denotes the preamble signal after demodulation, denotes the complex conjugate of the value at position ; is the autocorrelation function of, denotes the complex conjugate of the value at position , denotes a smoothing factor, is the coarse frequency offset estimated by the M&M algorithm, is the result of a frequency offset correction of on , denotes the complex conjugate of the value at position , is the autocorrelation function of, is the fine frequency offset obtained by the L&R algorithm; using the residual frequency offset contained in the current signal to digitally down-convert to obtain the frequency offset corrected baseband signal .

[0028] Preferably, step S5 specifically comprises: based on and , taking the baseband signal corresponding to the 10th and 11th preamble sequence channel symbols, whose length is , mapping D1 and D2 according to the D1 and D2 value table in the 188-110A protocol, and according to the Hadamard sequence mapping table, to obtain interleaving indication modulation symbols, scrambling the interleaving indication modulation symbols according to the positions corresponding to the preamble sequence, and performing constellation mapping, 4 times up-sampling and shaping filtering according to the 8PSK constellation diagram specified in the protocol, to obtain 14 groups of interleaving indication reference modulation signals , correlating and to obtain a third autocorrelation result , wherein is the complex conjugate of ,q representing the D1, D2 value set count, , taking the maximum correlation value is the decision value of D1, D2.

[0029] Preferably, step S6 specifically includes: the frame search of the data includes, based on and , the period of the 15 channel symbol preamble sequence is taken as a step size, and at most 24 groups of the corresponding baseband signal are continuously obtained , respectively, and are autocorrelated with , respectively , to obtain a fourth autocorrelation result , wherein is the complex conjugate of , based on a correlation threshold , sequentially search , until the fourth autocorrelation result g , the starting position of the first group is taken as the starting position of the data frame , , and then the frame demodulation of the data is performed; otherwise, it is considered that the starting position of the data frame is not found, and the step S1 is returned;

[0030] The frame demodulation of the data includes: a matched filter with a tap number of 33, a 4 times oversampling multiple, and a roll-off coefficient of 0.35 is constructed, and is matched filtered to obtain a filtered signal starting from , and the best sampling point is extracted, wherein represents the oversampling multiple, and a data frame demodulation constellation diagram is obtained , which is power normalized to obtain a power normalized demodulation constellation diagram , wherein represents the power normalization selection length.

[0031] Preferably, step S7 specifically includes: based on the 188-110A protocol data scrambling rule, a scrambling sequence with a period of 160 symbols is generated, based on the demodulation constellation diagram of the current data position j , a descrambled symbol and a descrambled constellation diagram are obtained;

[0032] Based on the 188-110A protocol, the log likelihood ratio of the descrambled constellation diagram is obtained.

[0033] Disturbance to the constellation chart In data rate During modulation, 8PSK modulation is used to obtain the first 1-bit log likelihood ratio. First 2-bit log likelihood ratio Compared with the first 3-bit log likelihood ;

[0034] Disturbance to the constellation chart In data rate During modulation, QPSK modulation is used to obtain the second 1-bit log likelihood ratio. Second 2-bit log likelihood ratio ;

[0035] Disturbance to the constellation chart In data rate During modulation, BPSK modulation is used to obtain the third 1-bit log likelihood ratio. ;

[0036] All the obtained log likelihood ratios are arranged and combined in order to obtain the soft information of the data frame. .

[0037] Preferably, step S8 specifically includes: firstly, in the 188-110A protocol interleaving information lookup table, using the parsed interleaving information D1 and D2, querying whether short interleaving or long interleaving is used; then, in the 188-110A protocol interleaving matrix table, using the data rate... The query retrieves the number of interleaved rows and columns; based on the bit storage and retrieval method corresponding to the 188-110A protocol, the interleaving of the soft information of the data frame is completed, resulting in the deinterleaved soft information of the data frame. Then, it is processed according to the maximum amplitude of 1. Bit quantization The quantitative results obtained are Through formula Calculate the branch metric of a two-bit convolutional code decoding ,in , Indicates branch output, binary high-order polarity The calculation formula is , binary low-order polarity The calculation formula is ,use By performing a Viterbi decoding operation on the Trellis trellis diagram of the convolutional code, soft information decoding is completed, and the final decoded information bits are obtained. .

[0038] The beneficial effects of the present application are:

[0039] 1) The present application is suitable for detecting and interpreting 188-110A signals under large frequency offset and low signal-to-noise ratio conditions. Compared with the traditional hard decision 188-110A signal processing, the signal detection anti-frequency offset is improved from ±0.4% of the symbol rate to ±100%, the signal detection performance is improved by about 3dB gain, and the signal demodulation decoding performance is improved by about 2.6dB gain. It is very suitable for deploying the algorithm in the single-chip microcomputer in the electronic countermeasure receiver. BRIEF DESCRIPTION OF DRAWINGS

[0040] Figure 1 The figure is a schematic diagram of the autocorrelation detection result of the preamble signal of the embodiment of the present application.

[0041] Figure 2 The figure is a schematic diagram of the differential cumulative correlation result of the embodiment of the present application.

[0042] Figure 3 The figure is a schematic diagram of the demodulation constellation of the embodiment of the present application.

[0043] Figure 4 The figure is a schematic diagram of the constellation after descrambling of the embodiment of the present application.

[0044] Figure 5 The figure is a schematic diagram of the comparison between the decoding result and the transmitted information bit result of the embodiment of the present application.

[0045] Figure 6 The figure is a flowchart of a soft demodulation method for 188-110A signals with large frequency offset and low signal-to-noise ratio according to an embodiment of the present application.

[0046] Figure 7 The figure is a schematic diagram of the detection accuracy of 188-110A signals with a frequency offset of 0 changing with signal-to-noise ratio according to an embodiment of the present application.

[0047] Figure 8 The figure is a schematic diagram of the comparison between the results of the present application and the traditional method when EbN0=0 according to an embodiment of the present application.

[0048] Figure 9 The figure is a schematic diagram of the comparison between the decoding bit error rate curves of the present application and the traditional hard decision method according to an embodiment of the present application. DETAILED DESCRIPTION

[0049] The technical solutions of the present application will be described in detail below with reference to the embodiments. Obviously, the described embodiments are only a part of the embodiments of the present application, not all. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor are within the scope of protection of the present application.

[0050] This application discloses a soft demodulation method for 188-110A signals with large frequency offset and low signal-to-noise ratio (SNR). It can quickly detect 188-110A signals even with large frequency offset, and is characterized by its ability to detect and decode 188-110A signals under low SNR conditions. Compared to hard-decision 188-110A signal decoding, it improves the frequency offset resistance from ±0.4% of the symbol rate to ±100%, and enhances demodulation and decoding performance by approximately 2.6 dB. The method is simple, fast, and has excellent decoding performance, making it very suitable for processing 188-110A signals in low-cost microcontrollers. A flowchart of the method is shown below. Figure 6 As shown, the specific steps include: S1, leader sequence detection and coarse synchronization of leader sequence based on autocorrelation;

[0051] S2. Prepare the baseband preamble reference modulation signal;

[0052] S3. Precise synchronization of leader sequences based on differential cumulative correlation;

[0053] S4. Joint frequency offset estimation and correction based on M&M and L&R based on the preamble reference signal;

[0054] S5. Demodulate the interleaving instructions D1 and D2 according to the D1 and D2 value table in the 188-110A protocol;

[0055] S6. Perform frame search and frame demodulation of the execution data;

[0056] S7. After descrambling the data frame constellation diagram, obtain the soft information of the data frame;

[0057] S8. Based on the 188-110A protocol, the deinterleaving of the soft information of the data frame is completed through the corresponding bit storage and reading method to obtain the deinterleaved soft information of the data frame, and then it is decoded.

[0058] For example, step S1 specifically includes: assuming the single-carrier 188-110A signal is processed by a channelized digital signal for , This represents the mapping from symbol to constellation diagram. This represents Gaussian white noise. This represents a root-raised cosine filter. This represents the residual frequency offset after digital down-conversion and channelization. N represents the sampling rate of the narrowband signal after channelization, and N' represents the symbol period. The symbol count is represented by the sampling rate in this embodiment. This will be illustrated using an example; the autocorrelation data will be normalized using the autocorrelation window length L. , m represents the first independent variable for summation, express Complex conjugate, express The complex conjugate of the autocorrelation data window length L in this embodiment is 200ms, i.e. The first autocorrelation result was obtained from 1 sampling point. In the normalized autocorrelation formula of this embodiment, the denominator is the normalization factor, and both the numerator and denominator can be quickly implemented using the FFT algorithm; if there is continuous The correlation values ​​make Then the 188-110A signal was found, where This represents the relevant threshold; in this embodiment, the relevant threshold is... , ; this The starting sample location of each consecutive correlation value is used as the rough starting location of the signal. If the 188-110A signal is not found, repeat step S1 until the 188-110A signal is found.

[0059] For example, step S2 specifically includes: taking the first 9 preamble channel symbols from the 15 preamble sequence (i.e., the non-interleaving indicator and non-counting indicator parts, channel symbols 0, 1, 3, 0, 1, 3, 1, 2, 0), mapping the 9 channel symbols to modulation symbols according to the Hadamard sequence mapping table of preamble channel symbols and modulation symbols in the 188-110A protocol, and repeating this process 4 times to obtain the preamble reference modulation symbol. The preamble reference modulation symbol is first scrambled according to the corresponding positions in the preamble sequence, and then constellation mapped, upsampled by 4 times, and shaped filtered according to the 8PSK constellation diagram specified in the protocol to obtain the baseband preamble reference modulation signal. Its length is .

[0060] For example, step S3 specifically includes: from Begin, take The first point is the same With delays of 2, 4, 8, 16, and 32, for Perform normalized difference autocorrelation Obtain the second autocorrelation result. In the normalized difference autocorrelation formula of this embodiment, the denominator is the normalization factor, and both the numerator and denominator can be quickly implemented using the FFT algorithm. Denotes delay, and k represents the second summation variable; delay i hour Difference conjugate product The calculation formula is , for Delay i Complex conjugate, express Complex conjugate, delay i hour Difference conjugate product The calculation formula is , for Delay i Complex conjugate, for The complex conjugate, if The maximum value is greater than Record the position of the maximum value as the precise starting position of the signal preamble sequence. Otherwise, it is assumed that the 188-110A signal was not found, and the process returns to step S1.

[0061] For example, step S4 specifically includes: from Begin, take The second sample point ,based on Perform joint M&M and L&R frequency offset estimation to calculate the residual frequency offset of the current signal. The calculation formula is as follows:

[0062] ;

[0063] ;

[0064] ;

[0065] ;

[0066] ;

[0067] ;

[0068] ;

[0069] ;

[0070] in, , N is Half of This indicates the demodulated preamble signal, which is a single tone. express In position The complex conjugate of the values, yes The autocorrelation function, express In position The complex conjugate of the values, Represents the smoothing factor. The coarse frequency offset is estimated by the M&M algorithm, and the estimated coarse frequency offset is used for correction. , Through coarse frequency offset right The result of frequency offset correction, express In position The complex conjugate of the values, yes The autocorrelation function was then used to obtain the fine frequency offset using the L&R algorithm. , and The sum of these is the final estimated residual frequency offset. Its estimation range can reach Use the residual frequency offset contained in the current signal. right Digital downconversion The frequency offset corrected baseband signal is obtained. .

[0071] For example, step S5 specifically includes: based on and Take the baseband signal corresponding to the 10th and 11th preamble sequence channel symbols. (i.e., from) Initially, 256 sampling points were taken, with a length of... According to the D1 and D2 value table in the 188-110A protocol, D1 and D2 are mapped according to the Hadamard sequence mapping table to obtain interleaving indicator modulation symbols. The interleaving indicator modulation symbols are scrambled according to the corresponding positions of the preamble sequence, and constellation mapping, 4x upsampling, and shaping filtering are performed according to the 8PSK constellation diagram specified in the protocol to obtain 14 sets of interleaving indicator reference modulation signals. to Conduct relevant Obtain the third autocorrelation result ,in for Complex conjugate, q This represents the count of the sets of values ​​that D1 and D2 can take. Take the maximum correlation value These are the decision values ​​for D1 and D2.

[0072] For example, step S6 specifically includes: the frame search of the data includes, based on and The period of the 15 channel symbol preamble sequence As step size (in this embodiment) =1920), obtain a maximum of 24 consecutive sets. Corresponding baseband signal (i.e., from) (Starting with a step size of 1920, 1152 sampling points were continuously taken, respectively with...) Perform autocorrelation The fourth autocorrelation result was obtained. ,in for Complex conjugate, based on relevant threshold Search sequentially Until the fourth autocorrelation result Then the first g Group The starting position is used as the starting position of the data frame. , , Then, perform frame demodulation of the data; otherwise, it is assumed that the start position of the data frame has not been found, and return to step S1.

[0073] The frame demodulation of the data includes a matched filter constructed with 33 taps, a 4x oversampling factor, and a roll-off factor of 0.35. Perform matched filtering to obtain from Initial filtered signal Extract the optimal sampling point ,in This indicates the oversampling factor; in this embodiment, the oversampling factor is... Equal to 4, obtain the data frame demodulation constellation diagram. Power normalization The demodulation constellation diagram with power normalization is obtained. ,in This indicates the length selected for power normalization.

[0074] For example, step S7 specifically includes: for PSK modulation, scrambling the modulation symbols can be seen as rotating the constellation diagram mapped by the modulation symbols, and descrambling can be performed by performing the corresponding inverse rotation, specifically including: generating a scrambling sequence with a period of 160 symbols based on the 188-110A protocol data scrambling rules. Demodulation constellation diagram based on current data position j Obtain the descrambling symbol And the constellation chart ;

[0075] Obtain the de-scratched constellation diagram based on the 188-110A protocol. The log likelihood ratio; the 188-110A protocol stipulates that the scrambled decision bits should pass through a Gray code encoder to obtain the bits before the deinterleaver. However, it is not easy to obtain the log likelihood ratio of the Gray code encoded bits by directly using the soft information (i.e., the log likelihood ratio of the decision bits of each scrambled constellation) obtained from the scrambled constellation diagram. Therefore, this application uses the scrambled constellation diagram to directly obtain the log likelihood ratio of the Gray code encoded bits. This step is the key to demodulation and decoding using soft information.

[0076] Disturbance to the constellation chart In data rate During modulation, 8PSK modulation is used to obtain the first 1-bit log likelihood ratio. First 2-bit log likelihood ratio Compared with the first 3-bit log likelihood ;

[0077] Disturbance to the constellation chart In data rate During modulation, QPSK modulation is used to obtain the second 1-bit log likelihood ratio. Second 2-bit log likelihood ratio ;

[0078] Disturbance to the constellation chart In data rate During modulation, BPSK modulation is used to obtain the third 1-bit log likelihood ratio. ;

[0079] The calculation formula is as follows:

[0080] ;

[0081] ;

[0082] ;

[0083] ;

[0084] ;

[0085] ;

[0086] ;

[0087]

[0088] in, The demodulation that can be determined by the constellation diagram has the following set of values: , , express MPSK constellation mapping, express and The distance of the constellation chart, Expressing the request The minimum constellation distance of the set of values;

[0089] All the obtained log likelihood ratios are arranged and combined in order to obtain the soft information of the data frame. .

[0090] For example, step S8 specifically includes: firstly, in the 188-110A protocol interleaving information lookup table, using the parsed interleaving information D1 and D2, querying to determine whether short interleaving or long interleaving is used; then, in the 188-110A protocol interleaving matrix table, using the data rate... The query retrieves the number of interleaved rows and columns; based on the bit storage and retrieval method corresponding to the 188-110A protocol, the interleaving of the soft information of the data frame is completed, resulting in the deinterleaved soft information of the data frame. Then, it is processed according to the maximum amplitude of 1. Bit quantization In this embodiment The quantitative results obtained are The maximum value of quantization does not exceed The minimum value is not lower than Through formula Calculate the branch metric of a two-bit convolutional code decoding ,in , Indicates branch output, binary high-order polarity The calculation formula is , binary low-order polarity The calculation formula is ,use By performing a Viterbi decoding operation on the Trellis trellis diagram of the convolutional code, soft information decoding is completed, and the final decoded information bits are obtained. .

[0091] For example, in this embodiment, the data rate value is set to Other signal parameter values ​​are set as follows: the sampling rate of the DDC signal. symbol rate carrier Information bits The quantity is 7200, with residual frequency offset. The signal is preceded by noise with 4800 samples, and the ratio of energy per bit to noise power spectral density is... The schematic diagram of the obtained preamble signal autocorrelation detection results is shown below. Figure 1 As shown, the approximate starting position of the signal can be obtained. The diagram illustrating the differential cumulative correlation results is shown below. Figure 2 As shown, the precise start position of the signal preamble sequence can be obtained. After correcting the frequency offset by an estimated 2400.013Hz, the demodulation constellation diagram is obtained as follows. Figure 3 As shown. The constellation chart is de-scrambled, and the resulting constellation diagram is shown below. Figure 4 As shown, soft information decoding is then performed to obtain the final decoding result. The decoding result and transmit information bits A comparison diagram showing the decoding result and the transmitted information bit result is shown below. Figure 5 As shown, the received result contains no errors, thus confirming the effectiveness of the present invention.

[0092] In summary, the soft demodulation method disclosed in this application for low signal-to-noise ratio (SNR) 188-110A signals with large frequency offset can quickly detect 188-110A signals even with large frequency offset, and has the characteristic of detecting and decoding 188-110A signals under low SNR conditions; when residual frequency offset The following is a schematic diagram showing the variation of the 188-110A signal detection accuracy with the signal-to-noise ratio: Figure 7 As shown in the figure, the autocorrelation and differential cumulative correlation detection proposed in this application can improve the gain by approximately 3dB. It can complete a 100% accuracy test in a short time; The results of this application compared with those of traditional methods are shown in the following diagram. Figure 8 As shown, by Figure 8 It can be seen that traditional methods are... The method would become completely ineffective at that time, while the method proposed in this application can cover [the following]. The entire range; a schematic diagram comparing the decoding error rate curves of this application and traditional hard-decision methods is shown in the figure. Figure 9 As shown, this application improves the signal by approximately 2.6 dB compared to traditional hard demodulation, demonstrating the processing advantage of this application at low signal-to-noise ratios.

[0093] The above description is merely a preferred embodiment of the present invention. It should be understood that the present invention is not limited to the forms disclosed herein and should not be construed as excluding other embodiments. It can be used in various other combinations, modifications, and environments, and can be altered within the scope of the concept described herein through the above teachings or related technologies or knowledge. Modifications and variations made by those skilled in the art that do not depart from the spirit and scope of the present invention should be within the protection scope of the appended claims.

Claims

1. A soft demodulation method for low signal-to-noise ratio (SNR) signals with large frequency offset (188-110A), characterized in that, Includes the following steps: S1. Autocorrelation-based leader sequence detection and coarse synchronization of leader sequences; S2. Prepare the baseband preamble reference modulation signal; S3. Precise synchronization of leader sequences based on differential cumulative correlation; S4. Joint frequency offset estimation and correction based on M&M and L&R based on preamble reference signal; S5. Demodulate the interleaving instructions D1 and D2 according to the D1 and D2 value table in the 188-110A protocol; S6. Perform frame search and frame demodulation of the execution data; S7. After descrambling the data frame constellation diagram, obtain the soft information of the data frame; S8. Based on the 188-110A protocol, the data frame soft information is deinterleaved through the corresponding bit storage and reading methods to obtain the deinterleaved data frame soft information, which is then decoded. Step S3 specifically includes: starting from the approximate starting position of the signal Begin, take The first point is the same L represents the autocorrelation data window length, with delays of 2, 4, 8, 16, and 32. Perform normalized difference autocorrelation Obtain the second autocorrelation result. ,in Let m represent the first summation variable and k represent the second summation variable; Delay i hour Difference conjugate product The calculation formula is , for Delay i Complex conjugate, express Complex conjugate, delay i Time baseband preamble reference modulation signal Difference conjugate product The calculation formula is , for Delay i Complex conjugate, for The complex conjugate, if The maximum value is greater than the relevant threshold. Record the position of the maximum value as the precise starting position of the signal preamble sequence. Otherwise, it is assumed that the 188-110A signal was not found, and the process returns to step S1. Step S4 specifically includes: from Begin, take The second sample point , Indicates the baseband preamble reference modulation signal The length, based on Perform joint M&M and L&R frequency offset estimation to calculate the residual frequency offset of the current signal. The calculation formula is as follows: ； ； ； ； ； ； ； ； in, , N is Half of This represents the demodulated preamble signal. express In position The complex conjugate of the values; yes The autocorrelation function, express In position The complex conjugate of the values, Represents the smoothing factor. It is the coarse frequency offset estimated by the M&M algorithm. Through coarse frequency offset right The result of frequency offset correction, express In position The complex conjugate of the values, yes The autocorrelation function, It is the fine frequency offset obtained by the L&R algorithm; it uses the residual frequency offset contained in the current signal. right Digital downconversion The frequency offset corrected baseband signal is obtained. .

2. The soft demodulation method for low signal-to-noise ratio (SNR) signals of 188-110A with large frequency offset according to claim 1, characterized in that, Step S1 specifically includes: assuming the single-carrier 188-110A signal is processed by a channelized digital signal for , This represents the mapping from symbol to constellation diagram. This represents Gaussian white noise. This represents a root-raised cosine filter. This indicates the residual frequency offset after digital down-conversion and channelization. N represents the sampling rate of the narrowband signal after channelization, and N' represents the symbol period. Represents symbol count; through autocorrelation data window length L... Perform normalized autocorrelation , express Complex conjugate, express The complex conjugate of the first autocorrelation result is obtained. If there is a sequence The correlation values ​​make Then the 188-110A signal was found. ; this The starting sample location of each consecutive correlation value is used as the rough starting location of the signal. If the 188-110A signal is not found, repeat step S1 until the 188-110A signal is found.

3. The soft demodulation method for low signal-to-noise ratio (SNR) signals of 188-110A with large frequency offset according to claim 2, characterized in that, Step S2 specifically includes: selecting the first 9 preamble channel symbols from the 15 preamble sequence channel symbols; mapping the 9 channel symbols to modulation symbols according to the Hadamard sequence mapping table of preamble channel symbols and modulation symbols in the 188-110A protocol; repeating this process 4 times to obtain the preamble reference modulation symbol. The preamble reference modulation symbol is first scrambled according to the corresponding positions in the preamble sequence, and then constellation mapped, upsampled by 4 times, and shaped filtered according to the 8PSK constellation diagram specified in the protocol to obtain the baseband preamble reference modulation signal. .

4. The soft demodulation method for low signal-to-noise ratio (SNR) signals of 188-110A with large frequency offset according to claim 3, characterized in that, Step S5 specifically includes: based on and Take the baseband signal corresponding to the 10th and 11th preamble sequence channel symbols. Its length is According to the D1 and D2 value table in the 188-110A protocol, D1 and D2 are mapped according to the Hadamard sequence mapping table to obtain interleaving indicator modulation symbols. The interleaving indicator modulation symbols are scrambled according to the corresponding positions of the preamble sequence, and constellation mapping, 4x upsampling, and shaping filtering are performed according to the 8PSK constellation diagram specified in the protocol to obtain 14 sets of interleaving indicator reference modulation signals. , and Conduct relevant Obtain the third autocorrelation result ,in for Complex conjugate, q This represents the count of the sets of values ​​that D1 and D2 can take. Take the maximum correlation value These are the decision values ​​for D1 and D2.

5. A soft demodulation method for low signal-to-noise ratio (SNR) signals of 188-110A with large frequency offset according to claim 4, characterized in that, Step S6 specifically includes: the frame search of the data includes, based on and The period of the 15 channel symbol preamble sequence As the step size, a maximum of 24 sets can be continuously acquired. Corresponding baseband signal , respectively with Perform autocorrelation The fourth autocorrelation result was obtained. ,in for Complex conjugate, based on relevant threshold Search sequentially Until the fourth autocorrelation result Then the first g Group The starting position is used as the starting position of the data frame. , , Then, perform frame demodulation of the data; otherwise, it is assumed that the start position of the data frame has not been found, and return to step S1. The demodulation of the data frame includes a matched filter constructed with 33 taps, a 4x oversampling factor, and a roll-off factor of 0.

35. Perform matched filtering to obtain from Initial filtered signal Extract the optimal sampling point ,in Indicates the oversampling factor, and obtains the demodulation constellation diagram of the data frame. Power normalization The demodulation constellation diagram with power normalization is obtained. ,in This indicates the length selected for power normalization.

6. A soft demodulation method for low signal-to-noise ratio (SNR) signals of 188-110A with large frequency offset according to claim 5, characterized in that, Step S7 specifically includes: generating a scrambling sequence with a period of 160 symbols based on the 188-110A protocol data scrambling rules. Demodulation constellation diagram based on current data position j Obtain the descrambling symbol And the constellation chart ; Obtain the de-scratched constellation diagram based on the 188-110A protocol. The log likelihood ratio; Disturbance to the constellation chart In data rate During modulation, 8PSK modulation is used to obtain the first 1-bit log likelihood ratio. First 2-bit log likelihood ratio Compared with the first 3-bit log likelihood ; Disturbance to the constellation chart In data rate During modulation, QPSK modulation is used to obtain the second 1-bit log likelihood ratio. Second 2-bit log likelihood ratio ; Disturbance to the constellation chart In data rate During modulation, BPSK modulation is used to obtain the third 1-bit log likelihood ratio. ; All the obtained log likelihood ratios are arranged and combined in order to obtain the soft information of the data frame. .

7. The soft demodulation method for a low signal-to-noise ratio (SNR) signal of 188-110A with large frequency offset according to claim 6, characterized in that, Step S8 specifically includes: First, in the 188-110A protocol interleaving information lookup table, using the parsed interleaving information D1 and D2, querying to determine whether short interleaving or long interleaving is used; then, in the 188-110A protocol interleaving matrix table, using the data rate... The query retrieves the number of interleaved rows and columns; based on the bit storage and retrieval method corresponding to the 188-110A protocol, the interleaving of the soft information of the data frame is completed, resulting in the deinterleaved soft information of the data frame. Then, it is processed according to the maximum amplitude of 1. Bit quantization The quantitative results obtained are Through formula Calculate the branch metric of a two-bit convolutional code decoding ,in , Indicates branch output, binary high-order polarity The calculation formula is , binary low-order polarity The calculation formula is ,use By performing a Viterbi decoding operation on the Trellis trellis diagram of the convolutional code, soft information decoding is completed, and the final decoded information bits are obtained. .

Images