Weak signal detection method and circuit based on run test

Abstract

The invention discloses a weak signal detection method and circuit based on run test, and belongs to the field of weak signal detection. The method comprises the following steps: isolating direct current of an analog signal to be detected through a capacitor C1, then superposing the analog signal to be detected with Gaussian white noise after direct current isolation through a capacitor C2, and mixing the superposed signal with a direct current voltage signal to obtain a mixed signal; the mixed signal is output to a comparator to be compared and quantized with a threshold voltage, and the signal quantized by the comparator is sent to a digital circuit to be sampled to obtain single-bit data; in a digital circuit, data analysis is carried out on single-bit data by utilizing a method based on run test, and whether weak signals exist or not is detected. The method can detect weak signals in real time, and has the advantages of being simple in circuit structure and low in implementation cost.

Description

Technical Field

[0001] The present invention belongs to the field of weak signal detection, and in particular relates to a weak signal detection method and circuit based on run-length test. Background Art

[0002] Weak signal detection has important applications in wireless communications, radar detection, and biomedical signal processing. Weak signals are often buried in a strong background of noise, making their detection challenging. Accurately and reliably detecting weak signals under low signal-to-noise ratio (SNR) conditions has become a key issue in system design and application. To address this issue, researchers have proposed a variety of detection methods, among which spectral analysis based on frequency domain analysis is a commonly used approach.

[0003] Spectral analysis transforms the signal into the frequency domain and analyzes the signal's spectral characteristics to identify the difference between noise and weak signals. In its implementation, the method first performs a Fourier transform on the signal and calculates its power spectral density. The presence of a signal is then determined by analyzing the characteristics of the power spectrum. However, this method has certain limitations when dealing with low signal-to-noise ratios: when there is a significant overlap between the noise power spectrum and the signal power spectrum, spectral analysis struggles to effectively separate weak signals from noise. Furthermore, frequency-domain analysis methods suffer from poor real-time performance and often require significant computing resources, making them unsuitable for resource-constrained embedded systems and miniaturized devices. Summary of the Invention

[0004] In view of the shortcomings of the existing technology, the present invention provides a weak signal detection method and circuit based on run-length test, which can realize real-time detection and has the advantages of simple circuit structure and low cost.

[0005] The technical solution adopted in the present invention is as follows:

[0006] In a first aspect, the present invention provides a weak signal detection method based on runs test, comprising the following steps:

[0007] The analog signal to be measured is isolated from DC by capacitor C1 and superimposed with Gaussian white noise after being isolated from DC by capacitor C2. The superimposed signal is mixed with the DC voltage signal to obtain a mixed signal;

[0008] The mixed signal is output to the comparator for comparison and quantization with the threshold voltage. The quantized signal is sent to the digital circuit for sampling to obtain single-bit data.

[0009] In digital circuits, a run-length test-based method is used to analyze single-bit data to detect the presence of weak signals.

[0010] Preferably, the method based on runs test comprises the following steps:

[0011] First, count the number of runs (segments with the same value appearing continuously in the data string) R in the sampled single-bit data string, as well as the number of 1s n1 and the number of 0s n2 in the run;

[0012] Calculate the expected value of the run E(R):

[0013]

[0014] Calculate the standard deviation σ(R):

[0015]

[0016] Calculate the test statistic Z:

[0017]

[0018] When the test statistic Z is greater than or equal to a preset threshold, it is determined that a weak signal exists in the analog signal to be tested; otherwise, it is determined that no weak signal exists in the analog signal to be tested.

[0019] Preferably, when a weak signal exists, the signal-to-noise ratio of the weak signal is estimated based on the degree of deviation of the test statistic Z from 0; the larger the value of the test statistic Z, the larger the signal-to-noise ratio, and the smaller the value of the test statistic Z, the smaller the signal-to-noise ratio.

[0020] Preferably, in the digital circuit, a trigger or a latch is used to sample the quantized signal.

[0021] Preferably, the digital circuit is implemented using FPGA.

[0022] In a second aspect, the present invention provides a weak signal detection circuit based on run-length test, comprising a capacitor C1, a capacitor C2, a resistor R1, a resistor R2, a power supply VCC, a comparator, and a digital circuit module;

[0023] The analog signal to be measured is isolated from DC by capacitor C1 and superimposed with the Gaussian white noise after being isolated from DC by capacitor C2 to obtain a superimposed signal;

[0024] The DC voltage output by the power supply VCC is divided by resistors R1 and R2 and mixed with the superimposed signal to obtain a mixed signal which is input to the comparator.

[0025] The mixed signal and the threshold voltage are compared and quantized in the comparator, and the quantized signal is sent to the digital circuit module;

[0026] The digital circuit module samples the input signal to obtain single-bit data, and then uses a run-length test-based method to perform data analysis on the single-bit data to detect whether there is a weak signal and output the detection result.

[0027] The working principle and advantages of the present invention are as follows:

[0028] Gaussian white noise is a type of noise with a normal (Gaussian) amplitude distribution and a flat power spectrum—in other words, completely random noise. Weak signals, on the other hand, are typically periodic sinusoidal or square waves and are non-random. Therefore, the randomness of the resulting mixed signal will vary somewhat compared to Gaussian white noise. By superimposing the signal to be measured with Gaussian white noise, the degree of randomness change can be used to detect the presence of a weak signal and further estimate the signal-to-noise ratio.

[0029] The runs test is a nonparametric statistical test used to determine whether a data sequence is random. It analyzes the continuous changes in positive and negative values ​​or categorical symbols in the data to determine randomness or trends. It is particularly suitable for processing binary or symbolic data. The quantized signal output by the comparator, after being sampled by a digital circuit, produces a binary single-bit data string consisting of 0s and 1s. Therefore, this single-bit data string can be used for runs tests, effectively reducing the computational complexity of the detection technology.

[0030] In summary, the present invention mixes the signal to be measured with high-quality white Gaussian noise. After mixing, the signal is sampled using a comparator and a flip-flop or latch in a digital circuit to obtain single-bit data. The single-bit data string is then used to detect weak signals and estimate the signal-to-noise ratio based on a run-length test. This method can detect weak signals in real time and has the advantages of a simple circuit structure and low cost. BRIEF DESCRIPTION OF THE DRAWINGS

[0031] Figure 1 This is a block diagram of a weak signal detection circuit based on run-length test of the present invention;

[0032] Figure 2 Detailed circuit diagram of a weak signal detection circuit based on run-length test according to an embodiment of the present invention. DETAILED DESCRIPTION

[0033] In order to better illustrate the purpose, advantages and technical ideas of the present invention, the present invention is further described below in conjunction with specific embodiments. It should be noted that the specific examples given below are only used to explain the present invention in detail and do not limit the present invention.

[0034] Figure 1 This is a block diagram of a weak signal detection circuit based on run-length test of the present invention. Figure 2This is a specific circuit diagram of a weak signal detection circuit based on run-length test of an embodiment. As shown in the figure, the circuit of this embodiment includes capacitor C1, capacitor C2, resistor R1, resistor R2, resistor R3, resistor R4, power supply VCC, comparator, DAC module, and FPGA module.

[0035] The analog signal to be measured is isolated from DC by capacitor C1 and superimposed with Gaussian white noise which is isolated from DC by capacitor C2 to obtain a superimposed signal.

[0036] The DC voltage output by the power supply VCC is divided by resistors R1 and R2 and mixed with the superimposed signal to obtain a mixed signal which is input to the comparator.

[0037] The DAC module converts the digital signal into an analog signal and divides it through resistors R3 and R4 to input it into the comparator as a threshold voltage.

[0038] The mixed signal is compared and quantized with the threshold voltage in the comparator, and the quantized signal is sent to the FPGA module.

[0039] The FPGA module samples the input signal to obtain single-bit data, and then uses a run-length test-based method to perform data analysis on the single-bit data to detect whether there is a weak signal and output the detection result.

[0040] Specifically, the method based on the runs test includes the following steps:

[0041] First, count the number of runs R in the sampled single-bit data string, as well as the number of 1s n1 and the number of 0s n2 in the run;

[0042] Calculate the expected value of the run E(R):

[0043]

[0044] Calculate the standard deviation σ(R):

[0045]

[0046] Calculate the test statistic Z:

[0047]

[0048] When the test statistic Z is greater than or equal to a preset threshold, it is determined that a weak signal exists in the analog signal to be tested; otherwise, it is determined that no weak signal exists in the analog signal to be tested.

[0049] When a weak signal is present, the signal-to-noise ratio (SNR) of the weak signal is estimated based on the degree to which the test statistic Z deviates from 0. A larger value for the test statistic Z indicates a higher SNR, while a smaller value indicates a lower SNR. When the data is completely random, the actual number of runs R and the expected number of runs E(R) are very close, so Z tends to 0. However, if the data exhibits a pattern or trend, the difference between R and E(R) increases, causing Z to deviate from 0, indicating that the data may contain non-random components. For example, Gaussian white noise is completely random, so its test statistic Z tends to 0. However, when mixed with a weak, non-random periodic signal, the randomness of the mixed signal is weakened compared to Gaussian white noise. Therefore, the degree to which Z deviates from 0, that is, the degree to which randomness is disrupted, can be used to detect the presence of a weak signal and further estimate its presence.

Claims

1. A weak signal detection method based on runs test, characterized in that: The following steps are involved: The analog signal to be measured is isolated from DC by capacitor C1 and superimposed with Gaussian white noise after being isolated from DC by capacitor C2. The superimposed signal is mixed with the DC voltage signal to obtain a mixed signal; The mixed signal is output to the comparator for comparison and quantization with the threshold voltage. The quantized signal is sent to the digital circuit for sampling to obtain single-bit data. In digital circuits, a run-length test-based method is used to analyze single-bit data to detect the presence of weak signals.

2. A weak signal detection method based on runs test as claimed in claim 1, characterized in that: The method based on runs test comprises the following steps: First, count the number of runs R in the sampled single-bit data string, as well as the number of 1s n1 and the number of 0s n2 in the run; Calculate the expected value of the run E(R): Calculate the standard deviation σ(R): Calculate the test statistic Z: When the test statistic Z is greater than or equal to a preset threshold, it is determined that a weak signal exists in the analog signal to be tested; otherwise, it is determined that no weak signal exists in the analog signal to be tested.

3. A weak signal detection method based on runs test as claimed in claim 2, characterized in that: When there is a weak signal, the signal-to-noise ratio of the weak signal is estimated based on the degree of deviation of the test statistic Z from 0; the larger the value of the test statistic Z, the larger the signal-to-noise ratio, and the smaller the value of the test statistic Z, the smaller the signal-to-noise ratio.

4. A weak signal detection method based on runs test as claimed in claim 2 or 3, characterized in that: In digital circuits, triggers or latches are used to sample the quantized signals.

5. A weak signal detection method based on runs test as claimed in claim 4, characterized in that: The digital circuit is implemented using FPGA.

6. A weak signal detection method based on runs test as claimed in claim 2 or 3, characterized in that: A circuit for implementing the weak signal detection method includes: a capacitor C1, a capacitor C2, a resistor R1, a resistor R2, a power supply VCC, a comparator, and a digital circuit module; The analog signal to be measured is isolated from DC by capacitor C1 and superimposed with the Gaussian white noise after being isolated from DC by capacitor C2 to obtain a superimposed signal; The DC voltage output by the power supply VCC is divided by resistors R1 and R2 and mixed with the superimposed signal to obtain a mixed signal which is input to the comparator. The mixed signal and the threshold voltage are compared and quantized in the comparator, and the quantized signal is sent to the digital circuit module; The digital circuit module samples the input signal to obtain single-bit data, and then uses a run-length test-based method to perform data analysis on the single-bit data to detect whether there is a weak signal and output the detection result.

Images