Weak signal envelope edge detection method
By using decision statistics derived from noise floor sampling and probability density function estimation, the problem of weak signal envelope edge detection is solved, achieving high-precision secondary radar interrogation signal detection and recognition, which is suitable for embedded systems.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-11-26
- Publication Date
- 2026-04-03
AI Technical Summary
Existing weak signal envelope edge detection methods cannot meet the high real-time requirements of secondary radar interrogation signals, and traditional methods have high computational and spatial complexity, making them difficult to apply effectively in embedded systems.
A decision-making method based on the maximum edge change energy statistic is adopted. By sampling the noise floor, estimating the probability density function, and setting a threshold for the false alarm rate, the maximum edge change energy of the received signal is calculated in real time to determine the existence of the weak signal envelope edge.
It improves the accuracy of secondary radar interrogation signal detection and recognition, reduces computational and spatial complexity, and is suitable for embedded systems with high real-time requirements.
Smart Images
Figure CN121784682A_ABST
Abstract
Description
Technical Field
[0001] This application belongs to the field of signal processing technology, and in particular relates to a method for detecting the envelope edge of a weak signal. Background Technology
[0002] Signal detection is a crucial aspect of signal processing, aiming to determine the presence of a specific signal in observed data. It has significant applications in systems such as communication, radar, and sonar. Due to the randomness of noise, signal detection is abstracted as a statistical hypothesis testing problem. By setting decision statistics, the probability of signal detection is maximized under certain criteria. The decision statistics and corresponding criteria form the core of signal detection. Classically, signal detection objects are categorized into three types: the presence or absence of a signal in noise, the detection of two signals of opposite polarity, and the detection of binary signals. These three types are judged using the Neyman-Pearson criterion, the uniform maximum power criterion, and the Bayes criterion, respectively. However, the detection of the envelope edge of a weak signal does not conform to the characteristics of these three types of detection objects, and these three decision criteria cannot be directly applied.
[0003] Signal detection is essentially a feature extraction process. The core of the research lies in what features to extract and how to extract them to distinguish signals from noise. Conventional, intuitive methods use energy, correlation, and cyclostationarity to differentiate noise from signals, or frequency characteristics in the frequency and time-frequency domains. To further improve the detection capability of weak signals, nonlinear feature extraction is increasingly being studied. Chaos theory, based on the characteristic that nonlinear systems are sensitive to initial conditions but immune to noise, detects weak signals through changes in the system's state. Stochastic resonance methods use nonlinear systems to convert some noise energy into signal energy, improving the signal-to-noise ratio to detect weak signals. Higher-order spectral analysis uses higher-order statistics to detect weak signals. These methods are all based on a physical understanding of signals, and the work mainly focuses on feature design. Multilayer perceptrons can automatically perceive signal features through learning, eliminating the need for feature design and shifting the focus to deep model design, thus effectively solving the problems of feature extraction and signal detection. However, neural network algorithms have high computational and space complexity, making it difficult to meet real-time requirements, and their deployment in embedded systems is costly and challenging.
[0004] The problem of weak signal envelope edge detection stems from the detection and identification of secondary radar interrogation signals. After envelope detection, long-duration differential phase-shift keying (DPSK) modulated signals exhibit small envelope changes at the points of phase reversal within the DC portion of the long pulse, resembling pulse edges. These weak signal envelope edges can affect the detection and identification of interrogation signals. Due to the system's high real-time requirements for signal detection, establishing a decision statistic that can effectively extract the features of weak signal envelope edges is a technical approach that meets the system requirements. Summary of the Invention
[0005] The purpose of this invention is to solve the problem of weak signal envelope edges affecting the detection and identification of secondary radar interrogation signals. This application provides a method for detecting the envelope edge of a weak signal, the method comprising: Step 1: Sample the noise floor and calculate the maximum edge change energy statistic of the noise floor; Step 2: Estimate the probability density function of the statistic; Step 3: Set the false alarm rate based on the probability density function, and obtain the threshold under the false alarm rate; Step 4: Calculate the maximum edge change energy of the received signal in real time, and determine whether the weak signal envelope edge exists based on the maximum edge change energy and the threshold.
[0006] Preferably, in step 1, the noise floor of the system received signal is sampled, and the maximum edge change energy statistic of the noise floor is calculated.
[0007] Preferably, the maximum edge change energy statistic is: (1) in, Indicates the received number A discrete signal Represents a set The number of elements, It is the first calculation The maximum edge change energy.
[0008] Preferably, in step 2, the probability density function of the maximum edge change energy of the noise floor is estimated using a nonparametric method of kernel density estimation.
[0009] Preferably, in step 3, the false alarm rate is set to 1%, and the threshold is obtained according to the probability density function.
[0010] Preferably, in step 4, the maximum edge change energy statistic of the received signal is calculated in real time. When the maximum edge change energy statistic is greater than the threshold, it is determined that the small signal envelope edge exists.
[0011] Preferably, the method is used to detect the presence of weak signals in the DC portion of a long pulse.
[0012] Preferably, the nonparametric method for kernel density estimation is the kdensity() function in Matlab.
[0013] Beneficial technical effects of the present invention: Compared to traditional weak signal detection methods, this invention is based on the Monte Carlo method. It estimates the probability density function of the decision statistic when the small signal is absent and present through sampling. Then, a threshold is obtained under a certain false alarm rate. When the decision statistic exceeds the threshold, the small signal envelope edge is determined to exist. This method solves the problem of small signal envelope edge detection, improves the accuracy of secondary radar interrogation signal detection and recognition, is simple to implement, and has low computational and spatial complexity, making it suitable for embedded systems with high real-time requirements. Attached Figure Description
[0014] Figure 1 A flowchart illustrating a weak signal envelope edge detection method provided in this application embodiment; Figure 2 A schematic diagram showing that the actual received signal provided in the embodiments of this application was not applied and the weak signal envelope edge was incorrectly identified as a pulse DC; Figure 3 is a histogram of the maximum edge change energy distribution provided in the embodiments of this application; Figure 4 The probability density map of maximum edge change energy provided for embodiments of this application; Figure 5 Provided for the embodiments of this application Figure 2 A schematic diagram of the maximum edge change energy calculated in real time for a signal; Figure 6 A schematic diagram of the detection results of the weak signal envelope edge provided in the embodiments of this application; Figure 7 This is a schematic diagram of the actual received signal provided in the embodiments of this application; Figure 8 The probability density map of maximum edge change energy provided for embodiments of this application; Figure 9 Provided for the embodiments of this application Figure 7 A schematic diagram illustrating the maximum edge change energy in real-time calculation of the signal. Figure 10 This is a schematic diagram of the detection results of the weak signal envelope edge provided in the embodiments of this application. Detailed Implementation
[0015] This invention provides a decision statistic for detecting small signal envelope edges, called the maximum edge change energy statistic. The maximum edge change energy has a high detection probability for small signal envelope edges within the DC portion of a long pulse. In implementation, it only requires sampling the noise floor and calculating the threshold, resulting in low computational and spatial complexity, making it suitable for embedded systems with high real-time requirements.
[0016] Please see Figures 1-10This invention provides a method and technique for detecting small signal envelope edges, which can be achieved through the following technical measures: Step 1: Sample the noise floor and calculate the maximum edge change energy statistic of the noise floor.
[0017] Step 2: Estimate the probability density function of the statistic.
[0018] Step 3: Set a certain false alarm rate and obtain the threshold under that false alarm rate.
[0019] Step 4: Calculate the maximum edge change energy of the received signal in real time. When it is greater than the threshold, it is determined that a small signal envelope edge exists.
[0020] Compared to traditional weak signal detection methods, this invention is based on the Monte Carlo method. It estimates the probability density function of the decision statistic when the small signal is absent and present through sampling. Then, a threshold is obtained under a certain false alarm rate. When the decision statistic exceeds the threshold, the small signal envelope edge is determined to exist. This method solves the problem of small signal envelope edge detection, improves the accuracy of secondary radar interrogation signal detection and recognition, is simple to implement, and has low computational and spatial complexity, making it suitable for embedded systems with high real-time requirements.
[0021] It should be noted that, Figure 2 In step 1 of the specific implementation example in this specification, the actual received signal is sampled for the noise floor of 0-8us; the long pulse DC part contains a small signal, and when this invention is not used, the envelope edge of the small signal is incorrectly identified as pulse DC, which is indicated by a red circle.
[0022] Figure 3. Histogram of the distribution of maximum edge change energy. The left histogram is the histogram of the decision statistics with noisy background, and the right histogram is the histogram of the decision statistics with long pulses containing small signals.
[0023] Figure 4 In step 2 of the specific implementation example in this manual, the probability density function (PDF) of the estimated maximum edge change energy is used. The blue line represents the PDF of the noisy floor decision statistic, and the red line represents the PDF of the decision statistic containing small signals and long pulses. The red line is only for comparison and does not need to be estimated in practice. The red dashed line represents the threshold obtained from the false alarm rate in step 3 of the specific implementation example in this manual. The portion of the red line above the threshold represents the detection probability of the small signal envelope edge, which is 45.17%.
[0024] Figure 5 In step 4 of the specific implementation example in this manual, real-time calculation... Figure 2 The maximum edge change energy statistic is defined as follows: the interval within the black dashed line represents the statistic corresponding to long pulses containing small signals; the red dashed line represents the threshold obtained in step 3; values greater than the threshold are considered to indicate the existence of a small signal envelope edge.
[0025] Figure 6 The detection results for the small signal envelope edge are shown in black circles, indicating no detected DC pulse. Figure 2 In comparison, the problem of small signals being incorrectly identified as pulsed DC signals is solved after using this invention.
[0026] Figure 7 In step 1 of the specific implementation example in this manual, the actual received signal, the area within the red circle is the noise floor, and the noise floor is sampled.
[0027] Figure 8 In step 2 of the specific implementation example in this manual, the estimated maximum edge change energy PDF is shown in red. The red line is only for comparison and does not need to be estimated in practice. The red dashed line represents the threshold obtained from the false alarm rate in step 3 of the specific implementation example in this manual.
[0028] Figure 9 In step 4 of the specific implementation example in this manual, real-time calculation... Figure 7 The maximum edge change energy statistic is shown in red. The red dashed line is the threshold obtained in step 3. If it is greater than the threshold, it is determined that the small signal envelope edge exists.
[0029] Figure 10 The detection results of the small signal envelope edge show that no pulse DC was detected by the black circle, which solves the problem of small signals being incorrectly identified as pulse DC.
[0030] This invention discloses a small-signal envelope edge detection method, primarily used to detect the presence of weak signals in the DC portion of long pulses. This method proposes a decision statistic based on the maximum edge change energy. Using Monte Carlo methods, it estimates the probability density function of the noise floor decision statistic through sampling, and then obtains a threshold at a certain false alarm rate. When the decision statistic exceeds the threshold, the presence of a small signal is determined.
[0031] In other embodiments of this application, a method for detecting the envelope edge of a weak signal (small signal) specifically includes the following steps: In step 1, the noise floor of the received signal is sampled, and the maximum edge change energy of the noise floor is calculated. The maximum edge change energy is... (1) in, Indicates the received number A discrete signal Represents a set The number of elements, It is the first calculation The maximum edge change energy.
[0032] In step 2, the probability density function of the maximum edge change energy of the noise floor is estimated using a nonparametric method of kernel density estimation, namely the kdensity() function in Matlab.
[0033] The probability density function (PDF) of the estimated maximum edge change energy statistic of the noise floor is attached. Figure 4 The blue line in the middle, Figure 4 The horizontal axis represents the maximum edge change energy statistic, and the vertical axis represents the probability. Figure 4 The red line in the diagram is the PDF of the maximum edge change energy statistics of long pulses containing small signals. It is only used for comparison and does not need to be estimated in practice.
[0034] In step 3, the false alarm rate is set to 1%, and the threshold is obtained based on the probability density function from the previous step.
[0035] Among them, the appendix Figure 4 The red dashed line in the diagram represents the threshold obtained based on the false alarm rate in step 3 of the specific implementation example in this manual. Figure 4 The portion of the red line above the threshold represents the detection probability of the small signal envelope edge, which is 45.17%.
[0036] In step 4, the maximum edge change energy statistic of the received signal is calculated in real time. When the maximum edge change energy statistic is greater than the threshold in the previous step, it is determined that the small signal envelope edge exists.
[0037] Among them, the appendix Figure 5 The blue line represents the maximum edge change energy statistic calculated in real time, and the red dashed line represents the threshold obtained in step 3. Values greater than the threshold are considered to indicate the existence of a small signal envelope edge. The interval within the black dashed line represents the maximum edge change energy statistic corresponding to long pulses containing small signals. As can be seen, this method can effectively detect the edges of weak signal envelopes.
Claims
1. A method for detecting the envelope edge of a weak signal, characterized in that, The method includes: Step 1: Sample the noise floor and calculate the maximum edge change energy statistic of the noise floor; Step 2: Estimate the probability density function of the statistic; Step 3: Set the false alarm rate based on the probability density function, and obtain the threshold under the false alarm rate; Step 4: Calculate the maximum edge change energy of the received signal in real time, and determine whether the weak signal envelope edge exists based on the maximum edge change energy and the threshold.
2. The method according to claim 1, characterized in that, In step 1, the noise floor of the received signal is sampled, and the maximum edge change energy statistic of the noise floor is calculated.
3. The method according to claim 2, characterized in that, The maximum edge change energy statistic is: (1) in, Indicates the received number A discrete signal Represents a set The number of elements, It is the first calculation The maximum edge change energy.
4. The method according to claim 1, characterized in that, In step 2, the probability density function of the maximum edge change energy of the noise floor is estimated using a nonparametric method of kernel density estimation.
5. The method according to claim 1, characterized in that, In step 3, the false alarm rate is set to 1%, and the threshold is obtained according to the probability density function.
6. The method according to claim 1, characterized in that, In step 4, the maximum edge change energy statistic of the received signal is calculated in real time. When the maximum edge change energy statistic is greater than the threshold, it is determined that the small signal envelope edge exists.
7. The method according to claim 1, characterized in that, The method is used to detect the presence of weak signals in the DC portion of a long pulse.
8. The method according to claim 4, characterized in that, The nonparametric method for kernel density estimation is the kdensity() function in Matlab.