Time-varying multi-interference source direct tracking method

Abstract

The invention discloses a time-varying multi-interference source direct tracking method. The method comprises the following steps: respectively modeling a new target hypothesis and a survival target hypothesis in a prediction stage; in the updating stage, a direct tracking spatial spectrum function is used for replacing a likelihood function in the filter updating step on each prediction target particle set, target-measurement data association is not needed, and information loss from interference source DOA tracking to position tracking is avoided. According to the invention, direct position tracking in a time-varying scene of the number of interference sources can be realized, and the adaptability of interference source tracking is improved.

Description

Technical Field

[0001] This invention belongs to the field of array signal technology, specifically relating to a time-varying method for direct tracking of multiple interference sources. Background Technology

[0002] Interference sources refer to devices that interfere with the normal operation of enemy electronic equipment, communication links, and sensor systems by emitting or relaying electromagnetic signals. The core objective is to weaken or paralyze the enemy's information acquisition, transmission, and processing capabilities through electromagnetic spectrum countermeasures. On the modern battlefield, enemy-launched electromagnetic interference can easily cause significant positioning errors or even render our navigation terminals unable to locate themselves, thus affecting our combat effectiveness. Tracking the location of interference sources is a prerequisite for implementing countermeasures such as targeted suppression.

[0003] Direct tracking of interference sources is a motion parameter estimation method for received signals. It uses sensors such as antenna arrays to passively receive the arrival wave of the signal radiated by the moving interference source and directly calculate the motion parameters of the interference source, including its geographical location and velocity. This avoids the information loss caused by the separation of parameter estimation and position and velocity calculation, thus achieving higher tracking accuracy.

[0004] Existing direct tracking methods for interference sources typically assume that the number of interference sources remains constant. However, in complex battlefield environments, this assumption is difficult to uphold. The movement, emergence, or disappearance of interference sources can easily cause both their number and state to change over time. Existing interference source tracking methods struggle to maintain stable tracking performance in time-varying scenarios. There is an urgent need to design a time-varying direct tracking method for multiple interference sources to improve the scenario applicability of existing methods, enabling robust interference suppression and allowing us to gain the initiative in electromagnetic spectrum warfare.

[0005] Tracking methods based on random finite sets are an effective approach for handling tracking problems with a time-varying number of targets. They can model the states and measurements of different multiple targets as sets and incorporate track initiation and termination mechanisms, enabling parallel estimation of the number and state of targets. A representative algorithm is... Filtering algorithms. Current tracking methods based on random finite sets are only applicable to point measurement models in active tracking. These models satisfy the following characteristics: each measurement is generated by a single target or clutter, and changes in any target do not affect measurements generated by other targets. However, active tracking easily exposes our facilities, thus inviting enemy attacks. In combat scenarios involving interference source tracking, passive tracking is a more commonly used method. The measurement model in passive tracking is usually a superposition measurement model. The measurements received by sensors such as antenna arrays are the superposition of signals and noise from all interference sources. Changes in any interference source will cause changes in the received measurements. Existing time-varying tracking methods are difficult to apply to time-varying multi-interference source tracking scenarios. Summary of the Invention

[0006] To overcome the shortcomings of existing technologies, this invention provides a time-varying multi-interference source direct tracking method. In the prediction phase, it models the hypotheses of newly emerging targets and surviving targets separately; in the update phase, it replaces the prediction target particle set with a direct tracking spatial spectrum function. The likelihood function in the filter update step does not require target-measurement data correlation and avoids information loss from DOA tracking of interference sources to position tracking. This invention enables direct position tracking in scenarios where the number of interference sources varies over time, improving the adaptability of interference source tracking.

[0007] The technical solution adopted by this invention to solve its technical problem is as follows: Step 1: Establish an interference source model; Step 2: Establish a measurement model; Step 3: Model the multi-objective state of the previous time step; Step 4: Predict the state of multiple objectives at the current moment; Step 5: Calculate the likelihood function based on the array received measurements at the current time; Step 6: Update the multi-objective state density at the current time step.

[0008] Preferably, step 1 specifically comprises: For a multi-source interference system, the number of interference sources is: The state of a single interference source is represented as ,in and These represent the x-direction and y-direction positions of the interference source, respectively. and These represent the velocity in the x-direction and the velocity in the y-direction, respectively; to distinguish each interference source, each interference source is assigned a label. ,in Indicates a label, This indicates the moment when the interference source is generated. This represents the sequence number of different targets appearing at the same time; the state of a labeled single target is represented as... Labeled multi-objective states are represented as .

[0009] Preferably, step 2 specifically comprises: For a distributed array observation system, the number of observation stations is: Each observation station is equipped with A uniform linear array of n elements, the nth The location of the interference source is determined by It means that the first The location of each observation station is determined by Indicates that the time The array receiver measurement model is represented as follows:

[0010] in The meaning has been explained in step 1. Indicates the first The observed signals from each observation station Indicates the first The observation station for the first The steering vector of each interference source Indicates the first The signal envelope of an interference source Indicates from the first The first interference source to the first Signal propagation delay at each observation station Indicates the first Additive white Gaussian noise between channels of each observation station.

[0011] Preferably, step 3 specifically comprises: right Multi-objective posterior probability density modeling at time step 1 is expressed as follows:

[0012] in Represents the posterior probability density of multiple objectives. This represents the labeled multi-objective state density. Indicates a label distinct indicator, used to describe The distinctness of the tags of each element in the text. Indicates time A set of labels, i.e., a hypothesis; This indicates the weight of the hypothesis. This represents the probability density of multiple targets. and Involved The function is the Kronecker function, and its specific expression is as follows: .

[0013] Preferably, step 4 specifically comprises: The multi-objective state density predicted at the current moment is calculated by the following formula:

[0014] in

[0015]

[0016]

[0017]

[0018]

[0019] in This indicates the prediction of multi-objective state density. Indicates the prediction of multi-objective states. This represents a combination of label hypotheses in the prediction hypothesis. This indicates the weight of the prediction hypothesis. This represents the probability density of predicted multiple targets. Indicates the survival hypothesis weight. Represents the live tag space, Indicates the weight of the newborn hypothesis. Represents the new tag space. This represents the probability density of surviving targets. This represents the probability density of newly generated targets. Indicates the probability of the target's survival. Indicates the probability of a target newborn. Represents the state transition function. Represents the inner product of integrals over the state space. and The meaning and expression of have been explained in step 3. and This indicates the inclusion of a function, and the specific expression is as follows:

[0020] Each predicted single-target density is calculated using a weighted particle set. It means that among them Indicates the number of particles. Indicates the first The position of each particle. This represents the weight of the particle; the particle follows a Gaussian distribution. ,in This represents the mean of the particle distribution. This represents the variance of the particle distribution.

[0021] Preferably, step 5 specifically comprises: Calculate the covariance matrix of the array received signal at the current time for each receiving station. ,in Indicates the first The covariance matrix of the signals received by each observation station. This indicates the number of snapshots of the received signal; The direct tracking spatial spectrum function is calculated for each predicted target particle set according to the following formula, and is used as the likelihood function of the tracking algorithm:

[0022] in Indicates the first The direct tracking spatial spectrum function at the location of each particle. Indicates the first The steering vector of each observation station relative to the location of a potential interference source.

[0023] Preferably, step 6 specifically comprises: The expression for the multi-objective state density at the current moment is as follows:

[0024] in:

[0025]

[0026] in This represents the updated multi-objective state density. This represents the updated label combination hypothesis. This represents the measurement likelihood function. This indicates the weight corresponding to the hypothesis. This indicates the updated multi-target density. This represents the probability density of the predicted target.

[0027] An electronic device includes: a processor and a memory; the memory is used to store a computer program, and the processor is used to execute the computer program stored in the memory to cause the electronic device to perform the above-described direct tracking method for time-varying multiple interference sources.

[0028] A computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the above-described direct tracking method for time-varying multiple interference sources.

[0029] A chip includes a processor for calling and running a computer program from a memory, causing a device equipped with the chip to perform the above-described direct tracking method for time-varying multiple interference sources.

[0030] The beneficial effects of this invention are as follows: 1. Compared with existing direct tracking methods for interference sources, this invention explicitly models the emergence and disappearance of interference sources. It uses the idea of ​​random finite sets to model the interference sources at each moment as a set of newly emerging interference sources and a set of surviving interference sources. It also incorporates the trajectory start and end mechanism, which can realize effective multi-interference source tracking in scenarios where the number of interference sources varies over time.

[0031] 2. The present invention is in Based on the filter, the likelihood function is replaced by the direct tracking spatial spectrum function of each predicted target particle set in the update stage. This solves the problem of target-measurement correlation difficulties caused by the superposition of array received signal measurements, and avoids information loss from DOA tracking of interference source to position tracking. Attached Figure Description

[0032] Figure 1 shows the tracking trajectory results: (a) X-direction trajectory tracking, (b) Y-direction trajectory tracking. Figure 2 shows a comparison of the tracking accuracy of the tracking algorithm of the present invention with that of existing tracking algorithms when time-varying information is known. (a) Comparison of tracking accuracy with different numbers of array elements, (b) Comparison of tracking accuracy with different numbers of snapshots. Detailed Implementation

[0033] The present invention will be further described below with reference to the accompanying drawings and embodiments.

[0034] This invention provides a method based on A time-varying multi-interference source direct tracking technique for filters is proposed. A time-varying multi-interference source direct tracking method applicable to array measurement frameworks is suggested. In the prediction stage, the hypotheses of newly generated targets and surviving targets are modeled separately; in the update stage, the direct tracking spatial spectrum function is used to replace the modeling on each predicted target particle set. The likelihood function in the filter update step does not require target-measurement data correlation and avoids information loss from DOA tracking of interference sources to position tracking. This invention enables direct position tracking in scenarios where the number of interference sources varies over time, improving the adaptability of interference source tracking.

[0035] The technical solution adopted by this invention to solve its technical problem includes the following steps: Step 1: Establish an interference source model; For a multi-source interference system, the number of interference sources is: The state of a single interference source can be represented as ,in and These represent the x-direction and y-direction positions of the interference source, respectively. and These represent the velocity in the x-direction and the velocity in the y-direction, respectively. To distinguish between different interference sources, each interference source is assigned a unique label. ,in Indicates a label, This indicates the moment when the interference source is generated. This represents the sequence number of different targets appearing at the same time. A labeled single-target state can be represented as... A labeled multi-objective state can be represented as .

[0036] Step 2: Establish a measurement model; For a distributed array observation system, the number of observation stations is: Each observation station is equipped with A uniform linear array of n elements, the nth The location of the interference source is determined by It means that the first The location of each observation station is determined by Indicates that the time The array receiver measurement model can be represented as follows:

[0037] in The meaning has been explained in step 1. Indicates the first The observed signals from each observation station Indicates the first The observation station for the first The steering vector of each interference source Indicates the first The signal envelope of an interference source Indicates from the first The first interference source to the first Signal propagation delay at each observation station Indicates the first Additive white Gaussian noise between channels of each observation station.

[0038] Step 3: Model the multi-objective state of the previous time step; right Multi-objective posterior probability density modeling at time step 1, the expression is as follows:

[0039] in Represents the posterior probability density of multiple objectives. This represents the labeled multi-objective state density. Indicates a label distinct indicator, used to describe The distinctness of the tags of each element in the text. Indicates time A set of labels, that is, a hypothesis. This indicates the weight of the hypothesis. This represents the probability density of multiple targets. and Involved The function is the Kronecker function, and its specific expression is as follows:

[0040] Step 4: Predict the state of multiple objectives at the current moment; The multi-objective state density predicted at the current moment can be calculated by the following formula.

[0041] in

[0042]

[0043]

[0044]

[0045]

[0046] in This indicates the prediction of multi-objective state density. Indicates the prediction of multi-objective states. This represents a combination of label hypotheses in the prediction hypothesis. This indicates the weight of the prediction hypothesis. This represents the probability density of predicted multiple targets. Indicates the survival hypothesis weight. Represents the live tag space, Indicates the weight of the newborn hypothesis. Represents the new tag space. This represents the probability density of surviving targets. This represents the probability density of newly generated targets. Indicates the probability of the target's survival. Indicates the probability of a target newborn. Represents the state transition function. Represents the inner product of integrals over the state space. and The meaning and expression of have been explained in step 3. and This indicates a function inclusion function, and the specific expression is as follows:

[0047] To facilitate algorithm implementation, a weighted particle set is used for each predicted single-target density. It means that among them Indicates the number of particles. Indicates the first The position of each particle. This represents the weight of the particle. The particles follow a Gaussian distribution. ,in This represents the mean of the particle distribution. This represents the variance of the particle distribution.

[0048] Step 5: Calculate the likelihood function based on the array received measurements at the current time; Calculate the covariance matrix of the array received signal at the current time for each receiving station. ,in Indicates the first The covariance matrix of the signals received by each observation station. This indicates the number of snapshots of the received signal.

[0049] The direct tracking spatial spectrum function is calculated for each predicted target particle set according to the following formula, and used as the likelihood function of the tracking algorithm.

[0050] in Indicates the first The direct tracking spatial spectrum function at the location of each particle. Indicates the first The steering vector of each observation station relative to the location of a potential interference source.

[0051] Step 6: Update the multi-objective state density at the current time step; The expression for the multi-objective state density at the current moment is as follows:

[0052] in:

[0053]

[0054] in This represents the updated multi-objective state density. This represents the updated label combination hypothesis. This represents the measurement likelihood function. This indicates the weight corresponding to the hypothesis. This indicates the updated multi-target density. This represents the probability density of the predicted target.

[0055] Example: Step 1: Establish an interference source model; For a multi-source interference system, with 3 interference sources, the state of a single interference source can be represented as follows: ,in and These represent the x-direction and y-direction positions of the interference source, respectively. and These represent the velocity in the x-direction and the velocity in the y-direction, respectively. To distinguish between different interference sources, each interference source is assigned a unique label. ,in Indicates a label, This indicates the moment when the interference source is generated. This represents the sequence number of different targets appearing at the same time. A labeled single-target state can be represented as... A labeled multi-objective state can be represented as .

[0056] Step 2: Establish a measurement model; For a distributed array observation system, there are 4 observation stations, each equipped with a uniform linear array of 7 elements. The location of the interference source is determined by It means that the first The location of each observation station is determined by Indicates that the time The array receiver measurement model can be represented as follows:

[0057] in Indicates the first The observed signals from each observation station Indicates the first The observation station for the first The steering vector of each interference source Indicates the first The signal envelope of an interference source Indicates from the first The first interference source to the first Signal propagation delay at each observation station Indicates the first Additive white Gaussian noise between channels of each observation station.

[0058] Step 3: Model the multi-objective state of the previous time step; right Multi-objective posterior probability density modeling at time step 1 is expressed as follows:

[0059] in Represents the posterior probability density of multiple objectives. This represents the labeled multi-objective state density. Indicates a label distinct indicator, used to describe The distinctness of the tags of each element in the text. Indicates time A set of labels, that is, a hypothesis. This indicates the weight of the hypothesis. This represents the probability density of multiple targets. and Involved The function is the Kronecker function, and its specific expression is as follows:

[0060] Step 4: Predict the state of multiple objectives at the current moment; The multi-objective state density predicted at the current moment can be calculated by the following formula:

[0061] in

[0062]

[0063]

[0064]

[0065]

[0066] in This indicates the prediction of multi-objective state density. Indicates the prediction of multi-objective states. This represents a combination of label hypotheses in the prediction hypothesis. This indicates the weight of the prediction hypothesis. This represents the probability density of predicted multiple targets. Indicates the survival hypothesis weight. Represents the live tag space, Indicates the weight of the newborn hypothesis. Represents the new tag space. This represents the probability density of surviving targets. This represents the probability density of newly generated targets. Indicates the probability of the target's survival. Indicates the probability of a target newborn. Represents the state transition function. Represents the inner product of integrals over the state space. and The meaning and expression of have been explained in step 3. and This indicates the inclusion of a function, and the specific expression is as follows:

[0067] To facilitate algorithm implementation, each predicted single-target density is represented by a weighted particle set of 200 particles. It means that among them Indicates the first The position of each particle. This represents the weight of the particle. The particles follow a Gaussian distribution. ,in This represents the mean of the particle distribution. This represents the variance of the particle distribution.

[0068] Step 5: Calculate the likelihood function based on the array received measurements at the current time; Calculate the covariance matrix of the array received signal at the current time for each receiving station. ,in Indicates the first The covariance matrix of the signals received by each observation station.

[0069] The direct tracking spatial spectrum function is calculated for each predicted target particle set according to the following formula, and used as the likelihood function of the tracking algorithm.

[0070] in Indicates the first The direct tracking spatial spectrum function at the location of each particle. Indicates the first The steering vector of each observation station relative to the location of a potential interference source.

[0071] Step 6: Update the multi-objective state density at the current time step; The expression for the multi-objective state density at the current moment is as follows:

[0072] in

[0073]

[0074] in This represents the updated multi-objective state density. This represents the updated label combination hypothesis. This represents the measurement likelihood function. This indicates the weight corresponding to the hypothesis. This indicates the updated multi-target density. This represents the probability density of the predicted target.

[0075] Figure 1 shows the invention based on... The tracking effect of the time-varying multi-interference source direct tracking method of the filter is shown in Figure 1. The entry and exit times of each interference source in the observation scene are set as follows: interference source 1 enters at time 1 and leaves at time 35; interference source 2 enters at time 5 and leaves at time 47; interference source 3 appears at time 43 and persists until the end of the observation, i.e., time 60. As can be seen from Figure 1, the time-varying multi-interference source tracking method proposed in this invention can accurately track the emergence and disappearance of interference sources even when the number of interference sources changes over time, and can achieve high-precision tracking that distinguishes the trajectories of different interference sources.

[0076] Figure 2 compares the tracking accuracy of the proposed tracking algorithm with existing tracking algorithms after 500 Monte Carlo experiments. Since existing interference source tracking algorithms are only applicable when the number of interference sources is constant or known, they cannot be applied in scenarios where the number of interference sources varies over time. For ease of comparison, it is assumed that the time-varying information of the interference sources is known in existing tracking algorithms, while the proposed tracking algorithm is still applied under the condition that the time-varying information of the interference sources is unknown. As can be seen from Figure 2, the proposed algorithm based on… The time-varying multi-interference source direct tracking method of the filter can achieve accurate tracking of multiple interference sources with a time-varying number. The tracking accuracy increases with the increase of the signal-to-noise ratio. When the signal-to-noise ratio is high, the tracking performance tends to be consistent with the existing methods with known time-varying information. Increasing the number of array elements and particles can improve the tracking performance.

Claims

1. A time-varying multi-interference source direct tracking method, characterized in that, Includes the following steps: Step 1: Establish an interference source model; Step 2: Establish a measurement model; Step 3: Model the multi-objective state of the previous time step; Step 4: Predict the state of multiple objectives at the current moment; Step 5: Calculate the likelihood function based on the array received measurements at the current time; Step 6: Update the multi-objective state density at the current time step.

2. The time-varying multi-interference source direct tracking method according to claim 1, characterized in that, Step 1 specifically involves: For a multi-source interference system, the number of interference sources is: The state of a single interference source is represented as ,in and These represent the x-direction and y-direction positions of the interference source, respectively. and These represent the velocity in the x-direction and the velocity in the y-direction, respectively; to distinguish each interference source, each interference source is assigned a label. ,in Indicates a label, This indicates the moment when the interference source is generated. This represents the sequence number of different targets appearing at the same time; the state of a labeled single target is represented as... Labeled multi-objective states are represented as .

3. The time-varying multi-interference source direct tracking method according to claim 2, characterized in that, Step 2 specifically involves: For a distributed array observation system, the number of observation stations is: Each observation station is equipped with A uniform linear array of n elements, the nth The location of the interference source is determined by It means that the first The location of each observation station is determined by Indicates that the time The array receiver measurement model is represented as follows: in The meaning has been explained in step 1. Indicates the first The observed signals from each observation station Indicates the first The observation station for the first The steering vector of each interference source Indicates the first The signal envelope of an interference source Indicates from the first The first interference source to the first Signal propagation delay at each observation station Indicates the first Additive white Gaussian noise between channels of each observation station.

4. The time-varying multi-interference source direct tracking method according to claim 3, characterized in that, Step 3 specifically involves: right Multi-objective posterior probability density modeling at time step 1 is expressed as follows: in Represents the posterior probability density of multiple objectives. This represents the labeled multi-objective state density. Indicates a label distinct indicator, used to describe The distinctness of the tags of each element in the text. Indicates time A set of labels, i.e., a hypothesis; This indicates the weight of the hypothesis. Denotes the probability density of multiple targets; where and Involved The function is the Kronecker function, and its specific expression is as follows: 。 5. The time-varying multi-interference source direct tracking method according to claim 4, characterized in that, Step 4 specifically involves: The multi-objective state density predicted at the current moment is calculated by the following formula: in in This indicates the prediction of multi-objective state density. Indicates the prediction of multi-objective states. This represents a combination of label hypotheses in the prediction hypothesis. This indicates the weight of the prediction hypothesis. This represents the probability density of predicted multiple targets. Indicates the survival hypothesis weight. Represents the live tag space, Indicates the weight of the newborn hypothesis. Represents the new tag space. This represents the probability density of surviving targets. This represents the probability density of newly generated targets. Indicates the probability of the target's survival. Indicates the probability of a target newborn. Represents the state transition function. Represents the inner product of integrals over the state space. and The meaning and expression of have been explained in step 3. and This indicates the inclusion of a function, and the specific expression is as follows: Each predicted single-target density is calculated using a weighted particle set. It means that, among them Indicates the number of particles. Indicates the first The position of each particle. This indicates the weight of the particle; The particles follow a Gaussian distribution ,in This represents the mean of the particle distribution. This represents the variance of the particle distribution.

6. The time-varying multi-interference source direct tracking method according to claim 5, characterized in that, Step 5 specifically involves: Calculate the covariance matrix of the array received signal at the current time for each receiving station. ,in Indicates the first The covariance matrix of the signals received by each observation station. This indicates the number of snapshots of the received signal; The direct tracking spatial spectrum function is calculated for each predicted target particle set according to the following formula, and is used as the likelihood function of the tracking algorithm: in Indicates the first The direct tracking spatial spectrum function at the location of each particle. Indicates the first The steering vector of each observation station relative to the location of a potential interference source.

7. The time-varying multi-interference source direct tracking method according to claim 6, characterized in that, Step 6 specifically involves: The expression for the multi-objective state density at the current moment is as follows: in: in This represents the updated multi-objective state density. This represents the updated label combination hypothesis. This represents the measurement likelihood function. This indicates the weight corresponding to the hypothesis. This indicates the updated multi-target density. This represents the probability density of the predicted target.

8. An electronic device, characterized in that, include: Processor and memory; The memory is used to store a computer program, and the processor is used to execute the computer program stored in the memory to cause the electronic device to perform the method as described in any one of claims 1 to 7.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, it implements the method as described in any one of claims 1 to 7.

10. A chip, characterized in that, include: A processor for retrieving and running a computer program from memory, causing a device on which the chip is mounted to perform the method as described in any one of claims 1 to 7.

Images