Method and system for structured blind identification of MPSK signal structure

Abstract

The application discloses a kind of MPSK signal structured full-blind identification, i.e.method and system for detecting, mainly solve the problem of low signal-to-noise ratio under the traditional "first detection and then identification" serial architecture link failure, rely on signal priori, long processing delay, high demand for computing power.The pure open-loop feedforward architecture of "cold start blind configuration + steady-state identification, i.e.detection" is adopted, the carrier dynamic structural annihilation is realized based on phase-preserving difference operator, the optimal differential interval blind configuration without priori is completed through the coefficient of variation of differential modulus, and the second-order differential sample variance of modulation phase is used as the core statistic quantity, and the MPSK modulation identification and signal existence judgment are completed at the same time, to realize the "identification and detection" closed loop of one-step completion.The application does not rely on any signal priori throughout the process, and the identification is accurate, the false alarm rate is stable and controllable under low signal-to-noise ratio, the end-to-end delay meets the requirements of vehicle level, and can be used in 6G vehicle-mounted high dynamic, low-power non-cooperative communication scene.

Description

Technical Field

[0001] This invention belongs to the field of signal processing technology, and in particular relates to a structured blind identification and detection method and system for MPSK signals, which can be used for non-cooperative signal blind processing in scenarios such as 6G vehicle communication, cognitive radio, broadband spectrum monitoring, and emergency communication. Background Technology

[0002] In fields such as 6G vehicular non-cooperative communication and broadband spectrum monitoring, receivers often face complex receiving environments with unknown frequency bands, extremely low signal-to-noise ratios, and high dynamic carrier interference. They must determine the existence of valid signals and identify modulation types without any prior signal information. Existing technologies for blind processing of non-cooperative signals mainly employ schemes such as serial closed-loop architectures, closed-loop synchronous architectures, and architectures based on cyclostationary characteristics and high-order cumulants.

[0003] The serial closed-loop architecture achieves blind processing by first detecting energy and then identifying modulation. However, the theoretical minimum working signal-to-noise ratio of energy detection is much higher than that of modulation identification. There is a significant "detection-identification performance inversion" phenomenon between the two. This results in the signal being stably identified in the low signal-to-noise ratio range but not being judged by traditional energy detection. The serial architecture has a fatal flaw of complete link collapse in the low signal-to-noise ratio range, making it unsuitable for extremely low signal-to-noise ratio automotive scenarios.

[0004] The closed-loop synchronous architecture uses feedback tracking modules such as phase-locked loops and Kalman filters to achieve carrier synchronization and bit synchronization for modulation recognition. However, the closed-loop architecture has a long convergence time and high hardware computing power requirements, making it unsuitable for the real-time requirements of low-computing-power automotive terminals.

[0005] The blind detection and recognition scheme based on cyclostationary features and high-order cumulants mainly achieves blind detection and recognition by weighted fusion of cyclostationary features and high-order cumulants. However, this method requires preset prior information such as symbol rate and carrier frequency, and needs to perform large-point FFT and high-order moment operations, resulting in high computational complexity and long processing latency, which cannot meet the core requirements of no prior information and low latency in non-cooperative vehicle scenarios.

[0006] Therefore, there is an urgent need for a fully blind identification and detection solution that abandons the traditional serial architecture, requires no prior signal information, and balances the ability to operate with extremely low signal-to-noise ratio with automotive-grade low latency requirements, so as to fundamentally solve the core defects of existing technologies. Summary of the Invention

[0007] The purpose of this invention is to propose a structured, fully blind identification and detection method and system for MPSK signals in 6G vehicular non-cooperative communication, in order to solve the core problems of the traditional "detect first, identify later" serial architecture, such as link failure under low signal-to-noise ratio, reliance on prior signal information, extended processing time, high hardware computing power requirements, and inability to adapt to the high dynamic and low computing power scenarios of 6G vehicular traffic.

[0008] The technical approach of this invention is as follows: Based on the newly revealed core physical phenomenon of "detection-recognition performance reversal," the traditional serial closed-loop architecture of "detection first, recognition later" is completely abandoned, and a pure open-loop feedforward architecture of "cold-start blind configuration + steady-state recognition as detection" is constructed. Based on the phase structure-preserving differential operator with group homomorphic mapping characteristics, the dynamic structure of typical vehicle carriers of second order and below is structurally annihilated, while the phase structure characteristics of MPSK modulation are fully preserved. Through the global minimum value characteristic of the differential modulus variation coefficient, the optimal differential interval blind configuration without any prior information is achieved. Using the sample variance of the second-order differential of the modulation phase as the only core statistic, the multi-class identification of MPSK modulation order and the binary decision of signal existence are completed simultaneously, realizing the theoretical closed loop of "recognition as detection" in which identification and detection are completed in one step.

[0009] Based on the above ideas, the technical solution of the present invention is as follows:

[0010] 1. A method for structured, fully blind identification and detection of MPSK signals, characterized in that it comprises:

[0011] (1) Set input constraints and global parameters to obtain the broadband complex baseband observation sequence of the unknown frequency band of 6G vehicle;

[0012] (2) Blind configuration of optimal difference interval for cold start: Traverse the preset difference interval range, calculate the coefficient of variation of the corresponding difference modulus, and blindly determine the optimal difference interval based on the global minimum value characteristic of the coefficient of variation, without the need for symbolic a prior information;

[0013] (3) The differential operation structure of the operator is determined by the optimal differential interval. The structural annihilation of second-order and lower-order carrier dynamics is realized through the group homomorphic mapping characteristics of the operator, and the output is a differential phase sequence that retains only the modulation phase characteristics.

[0014] (4) Perform mean removal processing on the differential phase sequence to obtain the core modulation feature statistics;

[0015] (5) Based on the modulation feature statistics, the multi-class identification of MPSK modulation order and the binary hypothesis test decision of signal existence are completed simultaneously, realizing the integration of identification and detection, and outputting the modulation type and signal existence flag.

[0016] Furthermore, in step (2), the range of positive integers of candidate differential intervals preset based on the system sampling rate is traversed; when the differential interval is a positive integer multiple of the oversampling rate, the coefficient of variation reaches a global minimum value, and when the coefficient of variation is less than or equal to the preset threshold, it is determined to be a valid differential interval.

[0017] Furthermore, in (3), the structured carrier annihilation is achieved by using a phase-preserving differential operator based on the characteristics of group homomorphic mapping. This operator performs differential operations on the input sequence, mapping the dynamic phase of the second-order and lower carrier polynomials to a constant or zero, while preserving the discrete structural characteristics of the MPSK modulation phase without loss. The differential interval of the differential operator is uniquely specified by the optimal differential interval blindly determined in (2).

[0018] Furthermore, in step (4), the mean-removal process is performed on the differential phase sequence to obtain the core modulation feature statistics, including: eliminating the constant phase offset corresponding to the linear frequency modulation slope, and calculating the sample variance of the differential sequence to obtain the sample variance of the second-order difference of the modulation phase, which is used as the core modulation feature statistics.

[0019] Furthermore, in step (5), the benchmark threshold for signal existence determination is determined based on a preset weighted interval of the theoretical variance of the second-order difference of the modulation phase corresponding to the highest modulation order and the theoretical variance of the second-order difference of the pure noise phase; the theoretical variance of the pure noise is... The modulation theory variance is given by the formula The only certainty is that among them The system presets the highest modulation order to be detected; modulation order identification is completed based on the unique identification characteristics of the variance of the second-order difference theory of the modulation phase of MPSK signals of different orders.

[0020] 2. A structured, fully blind identification and detection system for MPSK signals, characterized in that it comprises:

[0021] The input configuration module is used to complete the reception of input sequences, the definition of global constraint parameters, and the configuration of fully blind criteria;

[0022] The cold-start blind configuration module is used to determine the optimal difference interval without prior blinding by using the coefficient of variation of the difference modulus.

[0023] The phase-preserving differential feature extraction module is used to perform phase-preserving differential operations, realize carrier dynamic structural annihilation, and output differential phase sequence;

[0024] The phase feature statistics module is used to calculate the variance of differential phase samples and construct core modulation feature statistics.

[0025] The integrated identification and detection decision module is used to simultaneously perform binary hypothesis testing and multi-class identification based on the same modulation feature statistics, replacing the traditional serial operation of energy detection and feature extraction, and simultaneously completing modulation order identification and signal existence determination.

[0026] The result output module is used to output the modulation type and signal presence flag.

[0027] Compared with the prior art, the present invention has the following advantages:

[0028] Firstly, this invention abandons the traditional serial architecture of "detect first, then identify," and is based on "detection-identification performance inversion."

[0029] The phenomenon constructs an integrated architecture of "identification as detection" that completes identification and detection in one step. It can effectively operate with a threshold as low as 4.2dB, which is more than 10dB higher than traditional energy detection. It completely solves the problem of link collapse caused by detection failure under low signal-to-noise ratio.

[0030] Secondly, this invention requires no prior information such as signal modulation method, symbol rate, carrier frequency offset, or channel state. It achieves blind configuration of the optimal differential interval through the differential modulus variation coefficient and realizes dynamic structural annihilation of the carrier through the phase-preserving differential operator. It has no closed-loop tracking and no parameter iterative updates, making it perfectly suitable for non-cooperative communication scenarios without prior information.

[0031] Third, this invention adopts a pure open-loop feedforward architecture, with no loop convergence delay, low computational load, and end-to-end processing latency far lower than the automotive-grade 1ms latency constraint. It also consumes less hardware resources and can be stably implemented on a low-cost automotive-grade FPGA platform, making it suitable for low-computing-power automotive terminals.

[0032] Fourth, the operator differential operation structure constructed by the present invention using the optimal differential interval has a strong tolerance to typical static frequency offset, linear frequency modulation slope, multipath interference and timing error in vehicle scenarios. It can work stably in quasi-constant envelope scenarios and fully cover the complex communication scenarios of 6G vehicle high-speed movement and high dynamic interference. Attached Figure Description

[0033] Figure 1 This is a flowchart illustrating the overall implementation of the MPSK signal structured fully blind identification and detection method of the present invention.

[0034] Figure 2 This is a sub-flowchart of the cold start differential interval blind configuration in the method of the present invention;

[0035] Figure 3 This is a schematic diagram illustrating the principle of the integrated identification and detection decision in the method of the present invention;

[0036] Figure 4 This is a block diagram of the MPSK signal structured blind identification and detection system of the present invention. Detailed Implementation

[0037] The embodiments of the present invention will now be described in detail and completely with reference to the accompanying drawings.

[0038] Example 1: A structured, fully blind identification and detection method for MPSK signals in 6G vehicular non-cooperative communication.

[0039] This embodiment is designed based on a 6G vehicular broadband non-cooperative communication scenario. It requires no prior signal information throughout the process and achieves a QPSK signal recognition accuracy of ≥90% at an extremely low input signal-to-noise ratio of 4.2 dB. The system's false alarm rate is stably controlled within [specific parameters]. The end-to-end total processing latency is ≤164μs, which fully meets automotive-grade engineering requirements.

[0040] Reference Figure 1 The implementation steps of this embodiment include the following:

[0041] Step 1: Set input constraints and global parameters, and clarify the design criteria for full blindness.

[0042] To adapt to the false alarm rate tolerance requirements of vehicle communication scenarios and provide unified benchmark parameters for subsequent full-process processing, it is necessary to set input constraints and global parameters, and clarify the fully blind design principle, that is, the detection method does not require any prior information throughout the entire process. Its implementation includes the following:

[0043] 1.1) Set a constant false alarm rate baseline value for the system. The scope of engineering compatibility is To adapt to the false alarm rate tolerance requirements of vehicle communication scenarios;

[0044] 1.2) Based on the engineering adaptation range of 1024 to 8192, define the reference length of the single observation window. =4096, this reference length The value is the theoretical optimal value that combines statistical convergence accuracy, processing delay, and vehicle channel coherence time to ensure channel stability within a single window while meeting automotive-grade delay requirements.

[0045] 1.3) Set a fully blind design criterion in which all parameters are adaptively determined by the statistical characteristics of the input sequence, so as to avoid dependence on prior information such as signal modulation method, symbol rate, carrier parameters, and channel state throughout the process.

[0046] Step 2: Construct a difference operator using broadband complex baseband observation sequences.

[0047] To eliminate carrier dynamics in the observation sequence and retain only modulation features, a differential operator needs to be constructed, the implementation of which includes:

[0048] 2.1) The broadband complex baseband observation sequence of the unknown frequency band of the 6G vehicle-mounted system is obtained and filtered to obtain the operator input signal. :

[0049] ,

[0050] in, For discrete-time indexing, It is a second-order polynomial dynamic carrier phase. This is the MPSK modulation phase, the composition of which is only related to the modulation order and does not include carrier dynamics;

[0051] 2.2) Input signal to the operator By performing the core operational logic of "complex conjugate → complex square → two complex multiplications", the second-order complex conjugate difference operator is obtained. :

[0052] ,

[0053] in, It is the difference interval. This is a complex conjugate operation.

[0054] Step 3: Traverse the preset difference interval range of the difference operator, calculate the corresponding difference modulus variation coefficient, and blindly determine the optimal difference interval.

[0055] Reference Figure 2 The implementation of this step includes the following:

[0056] 3.1) Set the difference interval The range is 4 to 25, and it is a positive integer;

[0057] 3.2) Traverse the difference interval range D=4 to 25, and for each candidate difference interval D, calculate its difference operator. The mean of the difference modulus sequence and standard deviation :

[0058] ,

[0059] ,

[0060] Where p is the number of discrete-time indices n;

[0061] 3.3) Based on each candidate difference interval The coefficient of variation of the difference modulus is calculated from the corresponding mean and standard deviation. :

[0062] ;

[0063] 3.4) Based on the coefficient of variation of the difference modulus Generate the coefficient of variation sequence :

[0064] = ,... ,... ];

[0065] in, The variables are in the range of difference interval D = 4 to 25, and are positive integers;

[0066] 3.5) In the coefficient of variation sequence In the middle, take the difference interval corresponding to the global minimum. As the optimal difference interval ;

[0067] 3.6) Based on the analytical derivation of the difference modulus coefficient of variation moment characteristic theory, a judgment threshold of 0.39 is obtained. The judgment threshold is set to 0.39 to judge the coefficient of variation. Determine the optimal difference interval based on the size of the judgment threshold. Is this a valid configuration?

[0068] like If the difference is ≤0.39, then the optimal difference interval is... For effective configuration, proceed to step 4;

[0069] Otherwise, the configuration is invalid, the input sequence is determined to be pure noise, and the process ends.

[0070] Step 4: Determine the differential operation structure of the operator by the optimal differential interval, and realize the structural annihilation of second-order and lower-order carrier dynamics through the group homomorphic mapping characteristics of the operator.

[0071] 4.1) Based on the operator input signal initial phase Static frequency offset linear frequency modulation slope and sampling period ,Establish carrier phase second-order polynomial :

[0072] ;

[0073] 4.2) Based on the optimal difference interval The optimized difference operator is obtained. :

[0074] ;

[0075] 4.3) Based on the operator input signal The expansion yields the optimized difference operator. Difference operation structure:

[0076] ;

[0077] 4.4) Based on the principle that complex multiplication is equivalent to phase addition, the optimized operator is obtained. Output phase :

[0078] ,

[0079] make The second-order difference of the MPSK signal modulation phase is used to characterize the inherent phase change of the modulation feature. Its value is determined by the MPSK modulation order and does not include carrier dynamics. Output phase Simplified representation:

[0080] ;

[0081] 4.5) Phase the second-order polynomial carrier. Substitution To achieve structural annihilation of second-order and lower-order carrier dynamics, differential phase sequences are obtained. :

[0082]

[0083]

[0084]

[0085] ;

[0086] in, It is the constant phase shift corresponding to the linear frequency modulation slope.

[0087] As can be seen from the above calculation results, after performing the difference operation on the input sequence by the operator, the zero-order initial phase and the first-order static frequency offset in the input sequence can be completely annihilated, and the second-order linear frequency modulation slope can only retain a constant phase offset independent of the sampling point. At the same time, the discrete structural characteristics of the MPSK modulation phase are preserved without loss, and the final result only contains constant terms. It achieves structural annihilation of typical second-order and lower-order carrier dynamics in vehicles.

[0088] Step 5: Perform mean removal processing on the differential phase sequence to obtain the core modulation feature statistics.

[0089] 5.1) Based on differential phase sequence From the constant term The structure consists of constant terms whose mean is itself. Since the modulated signal usually satisfies the characteristic of zero mean, we can obtain The mean is 0, therefore for the differential phase sequence Performing mean removal eliminates the constant phase shift corresponding to the linear frequency modulation slope, resulting in a pure modulation phase difference sequence. :

[0090] ;

[0091] 5.2) Based on the difference sequence The modulation characteristics and variance were used to obtain the core modulation feature statistics. Modulation theory variance for:

[0092] ,

[0093] in It is the MPSK modulation order, which is completely independent of carrier dynamics and channel parameters.

[0094] Step 6: Based on modulation feature statistics, simultaneously complete the determination of signal existence and the classification and identification of MPSK modulation order.

[0095] Reference Figure 3 The implementation of this step includes the following:

[0096] 6.1) Determine the decision threshold A1 for the existence of the signal based on the midpoint between the theoretical variance of the highest modulation order to be detected and the theoretical variance of pure noise, which is preset by the system.

[0097] In this embodiment, the highest modulation order signal to be detected by the system is QPSK modulation, and the decision threshold A1 is preferably 19.12.

[0098] 6.2) Comparison of characteristic statistics The value of the decision threshold A1 is used to determine whether a valid signal exists within the target frequency band.

[0099] like If the value is less than A1, then a valid signal is determined to exist within the target frequency band, and step 6.3 is executed.

[0100] like If the result is greater than or equal to A1, it is determined to be pure noise, and an invalid detection result is output directly.

[0101] 6.3) Set the midpoint between the theoretical variance of BPSK modulation and the theoretical variance of QPSK modulation as the discrimination threshold A2;

[0102] In this embodiment, the preferred theoretical variance of BPSK modulation is 14.804, the preferred theoretical variance of QPSK modulation is 18.504, and the midpoint of the theoretical variances of the two is 16.655. Therefore, the distinction threshold is set to 16.655.

[0103] 6.4) Comparative characteristic statistics By comparing the magnitude of the discrimination threshold A2, the signal modulation order can be identified.

[0104] like If the value is less than A2, then the signal within the target frequency band is determined to be a second-order BPSK modulated signal;

[0105] like If the value is greater than or equal to A2, then the signal in the target frequency band is determined to be a fourth-order QPSK modulated signal;

[0106] 6.5) Based on the decisions made in steps 3.6), 6.2), and 6.4), output the final detection result:

[0107] If step 3.6) or step 6.2) is determined to be pure noise, then output the detection invalid flag.

[0108] If step 6.2) determines that a valid signal exists, then output the valid detection flag bit and the corresponding modulation type result.

[0109] The above output results realize a "recognition equals detection" closed loop that completes recognition and detection in one step.

[0110] Example 2: MPSK signal structured blind identification and detection system for 6G vehicular non-cooperative communication.

[0111] Reference Figure 4 This embodiment includes an input configuration module 1, a cold-start blind configuration module 2, a phase-structure-preserving differential feature extraction module 3, a phase feature statistics module 4, an integrated identification and detection decision module 5, and a result output module 6. The cold-start blind configuration module 2 includes: an initialization submodule 21, a mean and standard deviation submodule 22, a coefficient of variation submodule 23, a configuration submodule 24, and a verification submodule 25; the integrated identification and detection decision module 5 includes: an input submodule 51, a detection submodule 52, and an identification submodule 53.

[0112] The working principle of the entire system is as follows:

[0113] The input configuration module 1 is used to receive the input sequence, define the global constraint parameters and configure the full-blind criterion, and provide the received input sequence, global constraint parameter definition and full-blind criterion configuration to the cold start blind configuration module 2.

[0114] The cold-start blind configuration module 2 is used to construct the differential modulus variation coefficient based on the received input sequence, global constraint parameter definition, and full-blind criterion configuration, thereby completing the prior-free blind determination of the optimal differential interval. The initialization submodule 21 initializes the differential interval range based on the received input sequence and global constraints, and provides this range to the mean and standard deviation submodule 22. The mean and standard deviation submodule 22 calculates the mean and standard deviation of the differential modulus sequence based on the differential interval, and provides the obtained mean and standard deviation to the variation coefficient submodule 23. The variation coefficient submodule 23 calculates the differential modulus variation coefficient based on the mean and standard deviation, and provides the variation coefficient to the configuration submodule 24. The configuration submodule 24 extracts the global minimum of the variation coefficient sequence, configures the optimal differential interval, and provides the configured optimal differential interval to the verification submodule 25. The verification submodule 25 verifies whether the configured optimal differential interval is a valid configuration parameter and provides the verification result to the phase-preserving differential feature extraction module 3.

[0115] The phase-structure-preserving differential feature extraction module 3 is used to perform phase-structure-preserving differential operation according to the effective configuration parameters, realize carrier dynamic structural annihilation, output differential phase sequence, and provide the differential phase sequence to the phase feature statistics module 4.

[0116] The phase feature statistics module 4 is used to calculate the variance of the differential phase sequence samples, construct the core modulation feature statistics, and provide the core modulation feature statistics to the integrated identification and detection decision module 5.

[0117] The integrated identification and detection decision module 5 is used to simultaneously perform binary hypothesis testing and multi-class identification based on core modulation feature statistics. The input submodule 51 receives the modulation feature statistics and provides them to the detection submodule 52. The detection submodule 52 determines whether a valid signal exists within the target frequency band based on the received modulation feature statistics and provides the decision result to the identification submodule 53 and the result output module 6. The identification submodule 53 identifies the modulation type of the signal based on the decision result and provides the identification result to the result output module 6.

[0118] The result output module 6 is used to output the modulation type and signal presence flag based on the decision result.

[0119] It should be noted that in this embodiment, the direct coupling or communication connection between the modules can be achieved through some interfaces, or through indirect coupling or communication connections between modules. The functional modules and sub-modules in this embodiment can dynamically reside within a single processing unit, or each module can exist physically independently, or two or more modules can dynamically reside within a single processing unit. When the aforementioned dynamic components are implemented as software functional modules and sold or used as independent products, they can also be stored in a computer-readable and writable storage medium. This storage medium can be a memory, disk, or optical disc, etc.

[0120] The above description is merely a specific example of the present invention and does not constitute any limitation on the present invention. Obviously, those skilled in the art, after understanding the content and principles of the present invention, may make various modifications and changes in form and details without departing from the principles and structure of the present invention. For example, the configuration of the optimal differential interval and the construction of the core modulation feature statistics can be adjusted and changed in addition to those used in this example. These modifications and changes based on the ideas of the present invention are still within the scope of protection of the claims of the present invention.

Claims

1. A structured, fully blind identification and detection method for MPSK signals, characterized in that, include: (1) Set input constraints and global parameters to obtain the broadband complex baseband observation sequence of the unknown frequency band of 6G vehicle; (2) Blind configuration of optimal difference interval for cold start: Traverse the preset difference interval range, calculate the coefficient of variation of the corresponding difference modulus, and blindly determine the optimal difference interval based on the global minimum value characteristic of the coefficient of variation, without the need for symbolic a prior information; (3) The differential operation structure of the operator is determined by the optimal differential interval. The structural annihilation of second-order and lower-order carrier dynamics is realized through the group homomorphic mapping characteristics of the operator, and the output is a differential phase sequence that retains only the modulation phase characteristics. (4) Perform mean removal processing on the differential phase sequence to obtain the core modulation feature statistics; (5) Based on the modulation feature statistics, the multi-class identification of MPSK modulation order and the binary hypothesis test decision of signal existence are completed simultaneously, realizing the integration of identification and detection, and outputting the modulation type and signal existence flag.

2. The method according to claim 1, characterized in that, In step (2), the positive integer range of candidate differential intervals preset based on the system sampling rate is traversed; when the differential interval is a positive integer multiple of the oversampling rate, the coefficient of variation reaches a global minimum value, and when the coefficient of variation is less than or equal to the preset threshold, it is determined to be a valid differential interval.

3. The method according to claim 1, characterized in that, In (3), the structured carrier annihilation is achieved by using a phase-preserving differential operator based on the characteristics of group homomorphic mapping. This operator performs differential operations on the input sequence, mapping the dynamic phase of the second-order and lower carrier polynomials to a constant or zero, while preserving the discrete structural characteristics of the MPSK modulation phase without loss. The differential interval of the differential operator is uniquely specified by the optimal differential interval blindly determined in (2).

4. The method according to claim 1, characterized in that, In step (4), the mean-removal process is performed on the differential phase sequence to obtain the core modulation feature statistics, including: eliminating the constant phase offset corresponding to the linear frequency modulation slope, and calculating the sample variance of the differential sequence to obtain the sample variance of the second-order difference of the modulation phase, which is used as the core modulation feature statistics.

5. The method according to claim 1, characterized in that, In (5), the reference threshold for signal existence determination is determined based on a preset weighted interval of the theoretical variance of the second-order difference of the modulation phase corresponding to the highest modulation order and the theoretical variance of the second-order difference of the pure noise phase; the theoretical variance of the pure noise is... The modulation theory variance is given by the formula The only certainty is that among them The highest modulation order to be detected is preset by the system; Modulation order identification is achieved based on the unique identification characteristic of the variance of the second-order difference theory of the modulation phase of MPSK signals of different orders.

6. A structured, fully blind identification and detection system for MPSK signals, characterized in that, include: The input configuration module is used to complete the reception of input sequences, the definition of global constraint parameters, and the configuration of fully blind criteria; The cold-start blind configuration module is used to determine the optimal difference interval without prior blinding by using the coefficient of variation of the difference modulus. The phase-preserving differential feature extraction module is used to perform phase-preserving differential operations, realize carrier dynamic structural annihilation, and output differential phase sequence; The phase feature statistics module is used to calculate the variance of differential phase samples and construct core modulation feature statistics. The integrated identification and detection decision module is used to simultaneously perform binary hypothesis testing and multi-class identification based on the same modulation feature statistics. It replaces the traditional serial operation of energy detection and feature extraction with a single statistical comparison, and simultaneously completes modulation order identification and signal existence determination. The result output module is used to output the modulation type and signal presence flag.

7. The system according to claim 6, characterized in that, The cold start blind configuration module traverses the preset range of positive integer difference intervals, calculates the coefficient of variation of the corresponding difference modulus, takes the difference interval corresponding to the global minimum as the optimal difference interval, and verifies its effectiveness.

8. The system according to claim 6, characterized in that, The integrated identification and detection decision module is based on the sample variance of the second-order difference of the modulation phase, and simultaneously performs binary decision on signal existence and multi-class identification of modulation order.

Images