Networking radar target detection-oriented collaborative suppression interference genetic method

Abstract

The invention discloses a cooperative suppression interference genetic method for networking radar target detection, and the method comprises the steps: firstly, building a networking radar target detection model, and designing an intelligent cooperative interference target function under the condition that interference resources are limited; then designing a smart noise interference signal based on non-uniform sampling frequency modulation forwarding; and finally, improving a traditional genetic algorithm, redesigning a fitness function, and on the basis of constraining a crossover probability function and a mutation probability function of the fitness function, providing a collaborative suppression interference genetic algorithm. According to the invention, effective interference on networking radar target detection is realized; in an aircraft formation or target penetration process, compared with a traditional interference method and other intelligent interference methods, effective suppression interference in a networking radar target detection stage can be better realized; and even in the final stage of target penetration, the comprehensive detection probability of the radar network to the penetration target can still be not more than 81.20%.

Description

Technical field:

[0001] The invention belongs to the technical field of networked radar systems, and mainly relates to a collaborative suppression interference genetic method for networked radar target detection. Background technology:

[0002] In recent years, networked radar systems have attracted the attention of many scholars at home and abroad due to their superior combat performance and broad application prospects. Networked radar systems can make full use of the comprehensive characteristics of spectrum differences, signal style differences, spatial position differences, working mode differences, etc. among multiple radars, and can perform search and tracking tasks in a wide range of airspace and sea areas. Therefore, they play an increasingly important role in the modern defense system from long-range detection to mid-course tracking to terminal interception. At the same time, these advantages of networked radars also make it an important means to counter electronic interference. Its complex and diverse networking characteristics make it difficult for traditional single interference platforms and one-to-one interference methods to produce ideal interference effects on it, which brings severe challenges to traditional interference methods.

[0003] Therefore, the collaborative jamming technology of networking multiple jammers has gradually been widely studied in the industry. The multi-machine accompanying collaborative jamming method forms a jamming system with stronger jamming capability by uniformly allocating the jamming time, jamming power, jamming style and other resources of multiple jammers, and implements suppressive collaborative jamming on the opponent's networked radar system to protect the target and achieve the penetration mission. The traditional collaborative jamming method is more of a stacked superposition jamming of multiple jammers. Usually, the jamming strategy is pre-set according to the mission requirements, and the allocation method of jamming resources in the jamming system is relatively fixed. When the confrontation scenario changes, it is impossible to adjust the jamming strategy adaptively, and the optimization design capability of core parameters such as jamming power, jamming beam, and jamming signal is limited. Summary of the invention:

[0004] In order to overcome the above-mentioned deficiencies, the present invention provides a collaborative suppression interference genetic method for networked radar target detection.

[0005] The technical solution adopted by the present invention to solve its technical problem is:

[0006] A collaborative suppression interference genetic method for networked radar target detection includes the following steps:

[0007] 1) Establish a networked radar target detection model and design an intelligent collaborative jamming objective function under limited jamming resources;

[0008] 2) Design a smart noise jammer signal pattern based on non-uniform sampling frequency modulation forwarding;

[0009] 3) Based on the signal pattern designed in step 2), the crossover, mutation, and fitness function links of the traditional genetic algorithm are improved, the fitness value is redesigned, and its selection and mutation are constrained, so as to optimize the design and configuration of the interference beam, interference power, and interference pattern resources in the collaborative interference, so that it is suitable for the collaborative interference scenario of networked radar detection and achieves effective interference to the networked radar target detection.

[0010] The specific steps of establishing the networked radar target detection model in step 1) are as follows:

[0011] Assume that the network radar adopts the decision fusion structure model for the fusion of target detection probability, the fusion center adopts the KN criterion, and the network radar's decision result υ for target q q It can be expressed as:

[0012]

[0013] in, Indicates whether radar i detects a target. Indicates that the target was not found. Indicates that the target has been found;

[0014] Under the KN fusion criterion, the detection probability of the networked radar for the target q is:

[0015]

[0016] in, represents all permutations and combinations of the sum of the decisions of N radar nodes on target q, is the detection probability of radar i for target q.

[0017] The method for designing the intelligent coordinated interference objective function under the condition of limited interference resources in step 1) is as follows:

[0018] Assuming that there is no interference, the networked radar system simultaneously detects the Q targets in our penetration formation at time t, and summarizes the detection results of each target to obtain the networked radar system detection probability vector Pd(t):

[0019] Pd(t)=[Pd 1 (t) Pd 2 (t) … Pd Q (t)] (3)

[0021] After the coordinated interference is applied, the detection probability vector Pd(t) in this scenario becomes Pd j (t):

[0022]

[0023] The global objective function of cooperative interference for networked radar target detection is defined as:

[0024]

[0025] Among them, the larger the Pd(t), the more likely it is that after the interference is applied, the smaller the Pd j The lower the suppression of (t), the greater the contribution to the objective function and the better the interference effect; the goal of intelligent coordinated interference is to make formula (5) F[Pd(t),Pd j (t)] is maximized, thereby achieving effective interference to the networked radar;

[0026] In the case of limited jamming resources, the maximum jamming effect on networked radar detection can be achieved through intelligent joint optimization configuration of the beams, power and jamming patterns of multiple jammers. The optimization model can be equivalent to a mathematical optimization problem with jamming resource constraints:

[0027]

[0028] in, It is the pulse width constraint of the interference signal.

[0029] The specific method of step 2) designing a smart noise interference signal pattern based on non-uniform sampling frequency modulation forwarding is as follows:

[0030] Firstly, after each sampling pulse in the sampling pulse train obtains its corresponding radar sampling signal, the frequency modulation function is used to perform frequency modulation;

[0031] Then, convolve it with the noise signal;

[0032] Finally, by repeating the forwarding a corresponding number of times, the total forwarding interference signal in a single radar pulse cycle can be obtained as:

[0033]

[0034] Radars generally use matched filters to improve their detection performance. The matched filters can be set to set up To sample the kth sampling pulse, perform noise convolution and frequency modulation, and then perform M (k) The interference signal formed by repeated forwarding, that is, only A section of the signal:

[0035]

[0036] set up for The output signal after pulse compression can be obtained:

[0037]

[0038] The specific steps of improving the traditional genetic algorithm in step 3) are as follows:

[0039] First, the population is initialized; then, the fitness value of the population is calculated based on the interference objective function of the above formula (6), and the population with large fitness is selected as the high-quality population for crossover mutation. In this process, it will be restricted by constraints such as the interference signal pulse width of the NSFF-INJ interference signal;

[0040] Then, a new population is formed, completing a genetic iteration;

[0041] Finally, calculate the fitness value of the new population and determine whether to continue with the crossover and mutation steps;

[0042] The crossover probability and mutation probability of GA under unconstrained conditions are:

[0043]

[0044] Among them, P c1 , P c2 , P v1 , P v2 are the crossover and mutation probability coefficients, F m is the fitness value of the individual to be mutated, F avg is the average fitness value of individuals in the population, F max is the maximum fitness value of individuals in the population. When the NSFF-INJ interference signal is used, the corresponding crossover probability and mutation probability become

[0045]

[0046] Due to the adoption of the above-mentioned technical solution, the present invention has the following advantages:

[0047] The invention provides a genetic method for cooperative suppression interference for networked radar target detection, which optimizes the interference beam, interference power and interference signal pattern in cooperative interference. The invention can effectively interfere with networked radar target detection; in the process of aircraft formation or target penetration, compared with traditional interference methods and other intelligent interference methods, it can better achieve effective suppression interference in the target detection stage of networked radar; even in the final stage of target penetration, the comprehensive detection probability of the radar network to the penetration target can still be no more than 81.20%. Description of the drawings:

[0048] Figure 1 It is the flow chart of cooperative interference genetic algorithm;

[0049] Figure 2 It is a schematic diagram of the coordinated jamming scenario for target penetration of a networked radar system;

[0050] Figure 3 is the detection probability of target 1 by the networked radar under the three jamming methods;

[0051] Figure 4 is the detection probability of target 2 by the networked radar under the three interference methods. Specific implementation method:

[0052] In order to make the purpose, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Generally, the components of the embodiments of the present invention described and shown in the drawings here can be arranged and designed in various different configurations.

[0053] A collaborative suppression interference genetic method for networked radar target detection includes the following steps:

[0054] 1) Establish a networked radar target detection model and design an intelligent collaborative jamming objective function under limited jamming resources;

[0055] 2) Design a smart noise jammer signal pattern based on non-uniform sampling frequency modulation forwarding;

[0056] 3) Based on the signal pattern designed in step 2), the crossover, mutation, and fitness function links of the traditional genetic algorithm are improved, the fitness value is redesigned, and its selection and mutation are constrained, so as to optimize the design and configuration of the interference beam, interference power, and interference pattern resources in the collaborative interference, so that it is suitable for the collaborative interference scenario of networked radar detection and achieves effective interference to the networked radar target detection.

[0057] The specific steps for establishing the networked radar target detection model in step 1) are as follows:

[0058] Assume that the network radar adopts the decision fusion structure model for the fusion of target detection probability, the fusion center adopts the KN criterion, and the network radar's decision result υ for target q q It can be expressed as:

[0059]

[0060] in, Indicates whether radar i detects a target. Indicates that the target was not found. Indicates that the target has been found;

[0061] Under the KN fusion criterion, the detection probability of the networked radar for the target q is:

[0062]

[0063] in, represents all permutations and combinations of the sum of the decisions of N radar nodes on target q, is the detection probability of radar i for target q.

[0064] The method for designing the intelligent coordinated interference objective function under the condition of limited interference resources in step 1) is as follows:

[0065] Assuming that there is no interference, the networked radar system simultaneously detects the Q targets in our penetration formation at time t, and summarizes the detection results of each target to obtain the networked radar system detection probability vector Pd(t):

[0066] Pd(t)=[Pd 1 (t) Pd 2 (t) … Pd Q (t)] (15)

[0067] After the coordinated interference is applied, the detection probability vector Pd(t) in this scenario becomes Pd j (t):

[0068]

[0069] The more important the target is, the greater the probability of detecting it can be made by adjusting the system resources. However, the jammer needs to reduce the probability of detecting it by the networked radar. Therefore, Pd(t) / Pd j (t) will be larger. In this paper, the global objective function for cooperative interference of networked radar target detection is defined as:

[0070]

[0071] Among them, the larger the Pd(t), the more likely it is that after the interference is applied, the smaller the Pd j The lower the suppression of (t), the greater the contribution to the objective function and the better the interference effect; the goal of intelligent coordinated interference is to make formula (5) F[Pd(t),Pd j (t)] is maximized, thereby achieving effective interference to the networked radar;

[0072] In the case of limited jamming resources, the maximum jamming effect on networked radar detection can be achieved through intelligent joint optimization configuration of the beams, power and jamming patterns of multiple jammers. The optimization model can be equivalent to a mathematical optimization problem with jamming resource constraints:

[0073]

[0074] in, It is the pulse width constraint of the interference signal.

[0075] Since beam pointing and transmit power are coupled to each other, and beam pointing is a binary variable, the above optimization problem is non-convex and it is difficult to find the optimal solution. Therefore, this paper proposes a cooperative suppression jamming method based on an improved genetic algorithm to solve the optimization problem in order to obtain the optimal cooperative jamming effect for networked radars.

[0076] Noise jamming is the most primitive jamming style in the field of electronic countermeasures. It is low-cost and easy to implement. However, there is no coherence between noise jamming and radar signals, and its jamming effect on pulse compression radar is limited. The smart noise jamming method uses noise to modulate the intercepted signal, and after matched filtering, it can obtain radar processing gain. Compared with conventional noise suppression jamming, it brings an improvement in the interference-to-signal ratio, and the jamming result takes into account the dual effects of suppression and deception. The emergence of smart noise jamming poses a great threat to pulse compression radar. Since intermittent sampling forwarding jamming has the advantages of strong correlation, suppression and deception jamming effects, wide range of applicable radar signal styles, flexible jamming parameter modulation, small required jamming power, and applicability to accompanying jamming, the intermittent sampling noise modulation forwarding type of smart noise jamming has always been a difficulty in radar jamming countermeasures.

[0077] In this paper, intermittent sampling forwarding interference is selected as the radar coherent interference signal in smart noise jamming, and convolved with the noise signal to generate the basic interference pattern in cooperative interference, the Intelligent Noise Jamming Signal Based on Non-Uniform Sampling Frequency-Modulation Forwarding (NSFF-INJ), which is used to study the cooperative jamming technology for networked radar detection.

[0078] Step 2) The specific method for designing a smart noise interference signal pattern based on non-uniform sampling frequency modulation forwarding is as follows:

[0079] Firstly, after each sampling pulse in the sampling pulse train obtains its corresponding radar sampling signal, the frequency modulation function is used to perform frequency modulation;

[0080] Then, convolve it with the noise signal;

[0081] Finally, by repeating the forwarding a corresponding number of times, the total forwarding interference signal in a single radar pulse cycle can be obtained as:

[0082]

[0083] Radars generally use matched filters to improve their detection performance. The matched filters can be set to set up To sample the kth sampling pulse, perform noise convolution and frequency modulation, and then perform M (k) The interference signal formed by repeated forwarding, that is, only A section of the signal:

[0084]

[0085] set up for The output signal after pulse compression can be obtained:

[0086]

[0087]

[0088] It can be seen that the effect of repeated forwarding is that the interference pulse signal after noise convolution and frequency shift modulation will be matched filtered and superimposed with multiple radar signals of the same time width. The closer to the center, the higher the interference signal amplitude, and it decreases to both sides. Since the interference signal repeatedly forwarded in the same forwarding cycle uses the same frequency shift function, the frequency shift characteristics after the superposition of multiple interference pulses remain unchanged. At the same time, due to the existence of convolution noise, the interference effect not only has the ISRJ multi-false target interference characteristics, but also has the suppression interference effect. In addition, due to the influence of factors such as non-uniform sampling time, repeated forwarding times, and Doppler modulation frequency, the NSFF-INJ interference waveform has the advantages of non-uniformity, weak regularity, strong adjustability, can cover real targets, and has both deception and suppression dual interference characteristics.

[0089] The specific steps for improving the traditional genetic algorithm in step 3) are as follows: Figure 1 As shown,

[0090] First, the population is initialized; then, the fitness value of the population is calculated based on the interference objective function of the above formula (6), and the population with large fitness is selected as the high-quality population for crossover mutation. In this process, it will be restricted by constraints such as the interference signal pulse width of the NSFF-INJ interference signal;

[0091] Then, a new population is formed, completing a genetic iteration;

[0092] Finally, calculate the fitness value of the new population and determine whether to continue with the crossover and mutation steps;

[0093] The crossover probability and mutation probability of GA under unconstrained conditions are:

[0094]

[0095] Among them, P c1 , P c2 , P v1 , P v2 are the crossover and mutation probability coefficients, F m is the fitness value of the individual to be mutated, F avg is the average fitness value of individuals in the population, F max is the maximum fitness value of individuals in the population. When the NSFF-INJ interference signal is used, the corresponding crossover probability and mutation probability become:

[0096]

[0097] Comparative Experiment and Analysis

[0098] The proposed method is compared with the traditional interference method and the particle swarm-based intelligent optimization method through simulation experiments. Figure 2 As shown in the figure, in the target penetration application scenario, the enemy networked radar system consists of 6 radars, 2 targets intend to penetrate the networked radar system, 3 jammers implement team jamming, and the fusion center adopts the rank 4 criterion. The traditional method is that all jammers use a single-beam jamming system and the resource configuration method is fixed; the method in this paper adopts that jammer 1 is responsible for jamming radar 1 in the networked radar, jammer 2 is responsible for jamming radar 3 in the networked radar, and jammer 3 is responsible for jamming radar 5 in the networked radar.

[0099] The parameters of a single jammer and radar are shown in Table 1 and Table 2 respectively:

[0100] Table 1 Working parameters of a single jammer in cooperative jamming

[0101]

[0102] Table 2 Operating parameters of a single radar in the networked radar

[0103]

[0104] The entire penetration process is divided into 30 frames according to the time process in the experiment. At each frame position, the detection probability of the networked radar for two penetration targets under three interference methods is calculated. The experimental results are as follows: Figure 3 and Figure 4 shown.

[0105] It can be seen that the method proposed in this paper is better than the other two methods. This is because the traditional jamming method is limited by the number of jammers. A single jammer can only interfere with a single radar in a fixed manner. Other radars in the networked radar system can detect the target without interference, and the target will be exposed to the detection range of the radar network at a long distance. The particle swarm optimization method mainly adopts a two-step solution method, namely, beam pointing solution based on uniform distribution of beam transmission power and transmission power distribution. It adopts noise suppression interference and does not fully utilize the signal information of the interfered radar. At the specific node of target penetration, the method proposed in this paper fully considers the relative spatial position relationship between our penetration formation and the enemy radar, and intelligently and reasonably allocates the beam and interference power of the jammer. It also adopts the NSFF-INJ jamming signal, which can fully utilize the coherent information of the radar transmission signal while taking advantage of the noise interference, so that the radar anti-interference ability is greatly reduced.

[0106] In addition, until the two targets penetrated to a position very close to the radar network, the coordinated jamming system could still effectively suppress the networked radar system. For example, after 25 frames, the target detection probability increased, but the coordinated jamming could still effectively interfere with the networked radar system, making it impossible to obtain a high detection probability for both targets at the same time. For example, at the 29th frame, the detection probability of target 1 reached 90.10% ( Figure 3 ), but the detection probability of target 2 is only 70.67% (as shown Figure 4 As shown in Figure 2, the comprehensive detection probability of the penetration target is 80.39%; at the 30-frame position, the detection probability of target 2 reaches 94.21% (as shown in Figure 2). Figure 4 However, the detection probability of target 1 is reduced to 68.19% (as shown in Figure 3 As shown in the figure, the comprehensive detection probability of the penetration target is 81.20%. Therefore, the method proposed in this paper can effectively interfere with the radar network's stable detection of the target and then interfere with the establishment of the target's track, achieving effective interference with the enemy's radar network and effective protection against the penetration of our target.

[0107] The parts not described in detail in the above content are prior art, so they are not described in detail.

Claims

1. A collaborative suppression interference genetic method for networked radar target detection, characterized by: The steps include: 1) Establish a networked radar target detection model and design an intelligent collaborative jamming objective function under limited jamming resources; 2) Design a smart noise jammer signal pattern based on non-uniform sampling frequency modulation forwarding; 3) Based on the signal pattern designed in step 2), the crossover, mutation, and fitness function links of the traditional genetic algorithm are improved, the fitness value is redesigned, and its selection and mutation are constrained, so as to optimize the design and configuration of the interference beam, interference power, and interference pattern resources in the collaborative interference, so that it is suitable for the collaborative interference scenario of networked radar detection and achieves effective interference to the networked radar target detection.

2. According to claim 1, a collaborative suppression interference genetic method for networked radar target detection is characterized by: The specific steps of establishing the networked radar target detection model in step 1) are as follows: Assume that the network radar adopts the decision fusion structure model for the fusion of target detection probability, the fusion center adopts the KN criterion, and the network radar's decision result υ for target q q It can be expressed as: in, Indicates whether radar i detects a target. Indicates that the target was not found. Indicates that the target has been found; Under the KN fusion criterion, the detection probability of the networked radar for the target q is: in, represents all permutations and combinations of the sum of the decisions of N radar nodes on target q, is the detection probability of radar i for target q.

3. The collaborative suppression interference genetic method for networked radar target detection according to claim 1 is characterized in that: The method for designing the intelligent coordinated interference objective function under the condition of limited interference resources in step 1) is as follows: Assuming that there is no interference, the networked radar system simultaneously detects Q targets in our penetration formation at time t, and summarizes the detection results of each target to obtain the networked radar system detection probability vector Pd(t): Pd(t)=[Pd 1 (t) Pd 2 (t) … Pd Q (t)] (3) After the coordinated interference is applied, the detection probability vector Pd(t) in this scenario becomes Pd j (t): The global objective function of cooperative interference for networked radar target detection is defined as: Among them, the larger the Pd(t), the more likely it is that after the interference is applied, the smaller the Pd j The lower the suppression of (t), the greater the contribution to the objective function and the better the interference effect; the goal of intelligent coordinated interference is to make formula (5) F[Pd(t),Pd j (t)] is maximized, thereby achieving effective interference to the networked radar; In the case of limited jamming resources, the maximum jamming effect on networked radar detection can be achieved through intelligent joint optimization configuration of the beams, power and jamming patterns of multiple jammers. The optimization model can be equivalent to a mathematical optimization problem with jamming resource constraints: in, It is the pulse width constraint of the interference signal.

4. The collaborative suppression interference genetic method for networked radar target detection according to claim 3 is characterized by: The specific method of step 2) designing a smart noise interference signal pattern based on non-uniform sampling frequency modulation forwarding is as follows: Firstly, after each sampling pulse in the sampling pulse train obtains its corresponding radar sampling signal, the frequency modulation function is used to perform frequency modulation; Then, convolve it with the noise signal; Finally, by repeating the forwarding a corresponding number of times, the total forwarding interference signal in a single radar pulse cycle can be obtained as: Radars generally use matched filters to improve their detection performance. The matched filters can be set to set up The interference signal is formed by sampling the kth sampling pulse, performing noise convolution and frequency modulation, and then repeating M(k) again, that is, only A section of the signal: set up for The output signal after pulse compression can be obtained:

5. The collaborative suppression interference genetic method for networked radar target detection according to claim 4 is characterized in that: The specific steps of improving the traditional genetic algorithm in step 3) are as follows: First, the population is initialized; then, the fitness value of the population is calculated based on the interference objective function of the above formula (6), and the population with large fitness is selected as the high-quality population for crossover mutation. In this process, it will be restricted by constraints such as the interference signal pulse width of the NSFF-INJ interference signal; Then, a new population is formed, completing a genetic iteration; Finally, calculate the fitness value of the new population and determine whether to continue with the crossover and mutation steps; The crossover probability and mutation probability of GA under unconstrained conditions are: Among them, P c1 , P c2 , P v1 , P v2 are the crossover and mutation probability coefficients, F m is the fitness value of the individual to be mutated, F avg is the average fitness value of individuals in the population, F max is the maximum fitness value of individuals in the population. When NSFF-INJ interferes with the signal, the corresponding crossover probability and mutation probability become:

Images