CN112787964A

CN112787964A - BPSK and QPSK signal modulation identification method based on range median domain features

Info

Publication number: CN112787964A
Application number: CN202110186405.9A
Authority: CN
Inventors: 杨莉; 胡国兵; 赵敦博
Original assignee: Jinling Institute of Technology
Current assignee: Jinling Institute of Technology
Priority date: 2021-02-18
Filing date: 2021-02-18
Publication date: 2021-05-11
Anticipated expiration: 2041-02-18
Also published as: CN112787964B

Abstract

The invention provides a BPSK and QPSK signal modulation identification method based on range median domain features, aiming at the identification problem of two modulation signals of BPSK and QPSK. The method comprises the steps of firstly obtaining a square correction spectrum of a signal to be identified, then averagely grouping the square correction spectrum, respectively extracting a maximum value and a minimum value of each group, and calculating the average value of the two values to obtain a range median sequence. Then, the BPSK signal is converted into a map domain, the standard deviation of the map-degree vector is used as identification statistic, a proper threshold is set, and the identification statistic is compared with the threshold to realize the modulation identification of the BPSK signal and the QPSK signal. Simulation results show that the method can effectively identify BPSK and QPSK modulation signals under the condition of no signal prior information, and has better performance and low algorithm complexity under the condition of low signal-to-noise ratio.

Description

BPSK and QPSK signal modulation identification method based on range median domain features

Technical Field

The invention belongs to the field of signal identification and processing, and particularly relates to a BPSK and QPSK signal modulation identification method based on range median domain features.

Background

In a non-cooperative signal processing scenario such as cognitive radio and electronic reconnaissance, after signal detection is completed at the front end of the digital receiver, the modulation mode of the digital receiver needs to be effectively identified so as to further demodulate the signal, and necessary preconditions are provided for further implementing links such as individual identification and smart interference of the signal.

The previous correlation research is mostly based on the cyclostationarity, the extreme value distribution characteristic, the phase characteristic and the like of the signal. This type of algorithm does not require a priori information of the signal, but it suffers from poor performance at low signal-to-noise ratios. In recent years, with the rapid development of artificial intelligence technology, machine learning (deep learning) based correlation algorithms are widely used. The algorithm has small dependence on manpower in links such as feature definition and extraction, signal classifier design and the like, and is convenient for completing recognition under non-ideal conditions; however, a large number of training samples are required, and the features and the classifiers are learned and trained in advance, so that it is obviously difficult to obtain a large number of training samples under a non-cooperative condition. Therefore, it is important to design a method for effectively identifying the intercepted signal in real time under the condition of low signal-to-noise ratio.

The identification method based on the range median domain features performs graph conversion on the range median spectrum, selects a specific characteristic quantity and a threshold, and completes identification of BPSK and QPSK modulation signals.

Disclosure of Invention

Aiming at the defects in the prior art, the invention provides a BPSK and QPSK signal modulation identification method based on range median domain characteristics.

In order to achieve the purpose, the invention adopts the following technical scheme:

BPSK and QPSK signal modulation identification method based on range median domain features is characterized by comprising the following steps:

step 1: calculating a square correction spectrum of a signal to be identified;

step 2: uniformly grouping the square correction spectrums, extracting a maximum value and a minimum value of each group, and calculating an average value of each group to obtain a range median spectrum;

and step 3: carrying out image domain conversion on the range median spectrum, and extracting the standard deviation of the vector of the image degree as a modulation identification characteristic quantity;

and 4, step 4: setting a corresponding threshold;

and 5: and comparing the identification characteristic quantity with a set threshold, if the identification characteristic quantity is smaller than the threshold, determining the signal is a BPSK signal, otherwise, determining the signal is a QPSK signal.

In order to optimize the technical scheme, the specific measures adopted further comprise:

further, in step 1, let the signal to be identified be x (n), and calculate the frequency spectrum after the square operation as y (k) ═ DFT [ x ×)²(n)]K is 0, 1,.., M-1, where M is the number of sample points of the signal; sequentially arranging the amplitudes of Y (k) in descending order, setting the first c larger spectral lines to be 0, and recording as Y₁(k) And c is the number of correction points, the square correction spectrum of the signal is represented by z (k) ═ Y₁(k)|²。

Further, in step 2, the squared correction spectrum z (k) is divided into groups according to 3-5 spectral lines in each group, and the maximum value B of each group is extracted_max(m) and minimum B_min(m) calculating the mean value thereof to obtain a range median spectrum, i.e.

C(m)＝[B_max(m)+B_min(m)]/2，m＝0，1，...，N-1

Wherein N is the number of packets.

Further, step 3 is specifically as follows:

step 3.1: normalizing the range median spectrum C (m) to obtain a normalized spectrum C' (m):

then, setting a quantization level q, uniformly quantizing C' (m), and obtaining a quantized spectrum U (m) ═ i +1, wherein i/q < U (m) < i +1/q, and 0 < i ≦ q-1;

finally, converting u (m) into a graph domain to form a graph G (V, E), wherein V and E respectively represent a vertex set V ═ V of the graph₁，v₂，...v_qE and set of edges E ═ E_α，β|ν_α∈V，ν_β∈V}，e_α，βRepresents an edge between two vertices of the graph;

step 3.2: extracting and identifying characteristic quantity: calculating degree matrix D of graph G, and extracting degree vector D (D) whose diagonal elements form graph₁，d₂，...，d_j，...，d_q) Wherein d is_jThe degree vector standard deviation std (d) is calculated as the identification feature quantity for the sum of the number of edges connected to the jth vertex.

Further, converting u (m) to the graph domain, and constructing the graph G (V, E) specifically includes: from U (m) to U (m +1),

m

0, 1, N-2 is traversed one by one, when v is present_αTo v_βWhen the level of (2) jumps, the two vertices are connected, e _α，β1 is ═ 1; otherwise, there is no connection between the two vertices, e_α，β＝0。

Further, in step 4, a recognition threshold λ of BPSK and QPSK signals is set_evt，λ_evtTake a real number between 0.5 and 1.

Further, in step 5, the identification characteristic quantity is compared with a threshold to identify BPSK and QPSK signals: when std (d) < lambda_evtThen, BPSK modulation signal is obtained; otherwise, the signal is QPSK modulated.

The invention has the beneficial effects that: the invention extracts the range median of the square correction spectrum, performs graph conversion, determines the identification characteristic quantity according to the degree matrix of the graph, and completes the identification of the BPSK signal and the QPSK signal. Compared with the traditional time domain or transform domain algorithm, the method has better performance under the condition of low signal to noise ratio, low calculation complexity, no need of signal prior information and certain robustness.

Drawings

FIG. 1 is a flow chart of the identification method of the present invention.

Figure 2 is a graph of the performance of the method of the present invention compared to existing algorithms.

Fig. 3a and 3b are a full view and a partial view of the conversion of a BPSK signal and a QPSK signal, respectively.

Detailed Description

The present invention will now be described in further detail with reference to the accompanying drawings.

Fig. 1 is a flowchart of a BPSK and QPSK signal modulation identification method based on the range median domain feature. The method comprises the steps of firstly obtaining a square correction spectrum of a signal to be identified, then averagely grouping the square correction spectrum, respectively extracting a maximum value and a minimum value of each group, and calculating the average value of the two values to obtain a range median sequence. Then, the BPSK signal is converted into a map domain, the standard deviation of the map-degree vector is used as identification statistic, a proper threshold is set, and the identification statistic is compared with the threshold to realize the modulation identification of the BPSK signal and the QPSK signal. Simulation results show that the BPSK and QPSK modulation signals can be identified under the condition of no signal prior information.

The invention specifically comprises the following steps:

step 1: let the signal to be identified be x (n), and calculate the frequency spectrum after square operation as Y (k) ═ DFT [ x [ () ]²(n)]K is 0, 1,.., M-1, where M is the number of sample points of the signal; sequentially arranging the amplitudes of Y (k) in descending order, setting the first c larger spectral lines to be 0(c is the number of correction points, generally 1-5) and recording as Y₁(k) The squared corrected spectrum of the signal can then be expressed as

z(k)＝|Y₁(k)|²

Step 2: grouping the squared correction spectrum z (k) according to about 3-5 spectral lines of each group, and extracting the maximum value B of each group_max(m) and minimum B_min(m) calculating the mean value thereof to obtain the median range of range, i.e.

c(m)＝[B_max(m)+B_min(m)]/2，m＝0，1，...，N-1

Wherein N is the number of packets.

And step 3: and (3) carrying out graph conversion on the range median spectrum, and extracting the standard deviation of the vector of the graph degree as an identification characteristic quantity, wherein the method specifically comprises the following steps:

step 3.1: and (3) graph conversion: normalizing the range median spectrum C (m) to obtain normalized spectrum

Then, the quantization step number q is set and uniformly quantized to obtain U (m) ═ i +1, where i/q < U (m) < i +1/q, and 0 < i ≦ q-1. Finally, converting U (m) to graph domain to form graph G (V, E), wherein V and E respectively represent vertex set V of graph＝{v₁，v₂，...v_qE and set of edges E ═ E_α，β|v_α∈V，v_βIs belonged to V }. The specific method comprises the following steps: from U (m) to U (m +1),

m

0, 1, N-2 is traversed one by one, when v is present_αTo v_βWhen the level of (2) is changed, two vertices are connected, i.e. e _α，β1 is ═ 1; otherwise, there is no connection between the two vertices, i.e. e_α，β＝0。

Step 3.2: extracting and identifying characteristic quantity: calculating the degree matrix D of G, and extracting the degree vector D (D) of the diagram formed by the diagonal elements₁，d₂，...，d_j，...，d_q) Wherein d is_jThe degree vector standard deviation, std (d), is calculated as the identification feature quantity for the sum of the number of edges connected to the jth vertex.

And 4, step 4: setting identification threshold lambda of BPSK and QPSK signals_evtGenerally, a real number is 0.5 to 1.

And 5: the identifying characteristic is compared with a threshold to identify BPSK and QPSK signals. When std (d) < lambda_evtThen, BPSK modulation signal is obtained; otherwise, the signal is QPSK modulated.

Table 1 shows the identification performance of BPSK/QPSK signals under different signal-to-noise ratios, and the simulation conditions are as follows: the signal-to-noise ratio is [ -6, -4, -2, 0, 2, 4, 6, 8], the sampling frequency is 100MHz, the carrier frequency is 20.76MHz, the code element width is 640ns, the number of sample points is 1024, the initial phase is pi/3, each group is 5 spectral lines, the number of correction points is 5, the number of vertexes of the graph is 10, and the simulation is carried out 1000 times under each condition. As can be seen from Table 1, when the SNR is greater than 2dB, the average recognition accuracy can reach more than 95%.

TABLE 1 BPSK/QPSK signal identification performance under different SNR conditions

SNR	-6	-4	-2	0	2	4	6	8
									BPSK	0.962	0.965	0.968	0.956	0.938	0.974	0.959	0.939
QPSK	0.033	0.068	0.209	0.637	0.976	0.981	1	1
									Average	0.4975	0.5165	0.5885	0.7965	0.957	0.9775	0.9795	0.9695

FIG. 2 is a comparison of the performance of the present invention compared to a prior art algorithm, wherein BM is the method of the present invention; FAR is a constant false alarm algorithm, which is derived from a document [1] (Yangli, Hu soldier. an improved BPSK/QPSK signal modulation and identification algorithm [ J ]. telecommunication technology, 2017, 57 (08): 896-; GP is a super-threshold algorithm from document [2] (Yanlie, Hu nationality, Hu Chi dragon. BPSK/QPSK signal modulation identification based on GP distribution fitting test [ J ]. telecommunication technology, 2019, 59 (06): 691-. As can be seen from FIG. 2, the recognition performance of the method of the present invention is better when the signal-to-noise ratio is less than 2dB

In fig. 3a and 3b, the graphs obtained by converting the square correction spectrum extracted from the BPSK signal are complete graphs, and the graphs obtained by converting the square correction spectrum extracted from the QPSK signal are incomplete graphs, so that the modulation identification of the BPSK/QPSK signal can be realized.

The above is only a preferred embodiment of the present invention, and the protection scope of the present invention is not limited to the above-mentioned embodiments, and all technical solutions belonging to the idea of the present invention belong to the protection scope of the present invention. It should be noted that modifications and embellishments within the scope of the invention may be made by those skilled in the art without departing from the principle of the invention.

Claims

1. BPSK and QPSK signal modulation identification method based on range median domain features is characterized by comprising the following steps:

step 1: calculating a square correction spectrum of a signal to be identified;

and 4, step 4: setting a corresponding threshold;

2. The BPSK and QPSK signal modulation identification method as claimed in claim 1, wherein: in step 1, let the signal to be identified be x (n), and calculate the frequency spectrum after square operation as y (k) ═ DFT [ x ×)²(n)]K is 0, 1,.., M-1, where M is the number of sample points of the signal; sequentially arranging the amplitudes of Y (k) in descending order, setting the first c larger spectral lines to be 0, and recording as Y₁(k) And c is the number of correction points, the square correction spectrum of the signal is represented by z (k) ═ Y₁(k)|²。

3. The BPSK and QPSK signal modulation identification method as claimed in claim 1, wherein: in step 2, the square correction spectrum z (k) is divided into groups according to 3-5 spectral lines in each group, and the maximum value B of each group is extracted_max(m) and minimum B_min(m) calculating the mean value thereof to obtain a range median spectrum, i.e.

C(m)＝[B_max(m)+B_min(m)]/2，m＝0，1，...，N-1

Wherein N is the number of packets.

4. The BPSK and QPSK signal modulation identification method as claimed in claim 1, wherein: the step 3 is as follows:

then, setting a quantization level q, uniformly quantizing C' (m), and obtaining a quantized spectrum U (m) ═ i +1, wherein i/q < U (m) < i +1/q, and i is greater than or equal to 0 and less than or equal to q-1;

finally, converting u (m) into a graph domain to form a graph G (V, E), wherein V and E respectively represent a vertex set V ═ V of the graph₁，v₂，...v_qE and set of edges E ═ E_α，β|v_α∈V，v_β∈V}，e_α，βRepresents an edge between two vertices of the graph;

step 3.2: extracting and identifying characteristic quantity: calculating degree matrix D of graph G, and extracting degree vector D (D) whose diagonal elements form graph₁，d₂，...，d_j，..，d_q) Wherein d is_jThe degree vector standard deviation std (d) is calculated as the identification feature quantity for the sum of the number of edges connected to the jth vertex.

5. The BPSK and QPSK signal modulation identification method based on the range median domain feature of claim 4, wherein: converting U (m) into a graph domain to form a graph G (V, E) by the specific steps of: from U (m) to U (m +1), m 0, 1, N-2 is traversed one by one, when v is present_aTo v_βWhen the level of (2) jumps, the two vertices are connected, e_α，β1 is ═ 1; otherwise, there is no connection between the two vertices, e_a，β＝0。

6. The BPSK and QPSK signal modulation identification method based on the range median domain feature of claim 4, wherein: in step 4, set the identification threshold λ of BPSK and QPSK signals_evt，λ_evtTake a real number between 0.5 and 1.

7. The BPSK and QPSK signal modulation identification method as claimed in claim 6, wherein: in step 5, the identification characteristic quantity is compared with a threshold to identify BPSK and QPSK signals: when std (d) < lambda_evtThen, BPSK modulation signal is obtained; otherwise, isThe signal is QPSK modulated.