Darkroom probe selection method, device, equipment, medium and product

Abstract

The invention relates to the technical field of wireless communication, in particular to a darkroom probe selection method and device, equipment, a medium and a product, and the method comprises the steps: selecting a first probe subset, a second probe subset and a third probe subset from a current population; identifying an attack subset and a victim subset according to an electrostatic discharge algorithm, and updating the victim subset based on the attack subset; performing damage inspection on the current population to generate a next-generation population; and updating the current population to a next generation population, carrying out iteration until a preset condition for exiting iteration is met, and selecting an optimal probe subset from the final population. By introducing an electrostatic discharge algorithm and setting a fitness function, a subset (damaged subset) with low fitness learns from a subset (attack subset) with high fitness, an optimal probe subset, namely an effective probe, can be accurately and rapidly screened from a large number of candidate probes, and the probe selection accuracy and speed are remarkably improved.

Description

Technical Field

[0001] This disclosure relates to the field of wireless communication technology, and in particular to a method, apparatus, device, medium, and product for selecting probes in an anechoic chamber. Background Technology

[0002] With the application of cutting-edge technologies, the antenna systems (transmitting and receiving signals) of devices such as terminals and base stations are increasingly exhibiting high complexity and integration. However, the highly integrated nature of antenna systems limits the deployment space of conducted test connectors, thus restricting conducted testing. Therefore, the multi-probe anechoic chamber method has become a feasible solution for performance testing of complex device antenna systems.

[0003] The multi-probe anechoic chamber method, a technique based on an anechoic chamber and probe antenna arrays, often requires deploying a large number of probes within the anechoic chamber to expand the coverage of the test area. However, in the wireless signal testing of 5G communication equipment, how to more accurately and quickly simulate the real communication environment using fewer probes within the anechoic chamber has become a pressing problem to be solved. Summary of the Invention

[0004] In view of the above problems, this disclosure is made to provide a method, apparatus, equipment, medium and product for selecting anechoic chamber probes.

[0005] According to one aspect of this disclosure, a method for selecting a probe in an anechoic chamber is provided, comprising: Select a first probe subset, a second probe subset, and a third probe subset from the current population; wherein, the current population includes multiple probe subsets, the probe subsets are generated based on the probes deployed in the darkroom and the preset selection number, and the first probe subset, the second probe subset, and the third probe subset are sorted according to their fitness; Based on the electrostatic discharge algorithm, the attack subset and victim subset in the first probe subset, the second probe subset and the third probe subset are identified. Based on the attack subset, the victim subset is updated; where the victim subset is the party that suffers the discharge attack. Perform damage checks on the current population, including the updated victim subset; Based on the current population after damage inspection, generate the next generation population; Update the current population to the next generation population, and repeat the steps of selecting the first probe subset, the second probe subset and the third probe subset from the current population until the preset condition for exiting the iteration is met, and then use the current population as the final population. Based on fitness, select the optimal subset of probes from the final population.

[0006] By introducing an electrostatic discharge algorithm and setting a fitness function, subsets with low fitness (victim subsets) learn from subsets with high fitness (attack subsets). This allows for precise and rapid selection of the optimal subset of probes (effective probes) from a large number of candidate probes, significantly improving the accuracy and speed of probe selection. This disclosure exhibits fast convergence speed, achieving convergence within 100 iterations under different channel models and probe weight optimization methods. It maintains good performance under various channel models and test scenarios, demonstrating strong robustness.

[0007] Furthermore, according to one aspect of the anechoic chamber probe selection method of this disclosure, based on an electrostatic discharge algorithm, attack subsets and victim subsets are identified within a first probe subset, a second probe subset, and a third probe subset, including: Generate a random number and determine whether the random number is greater than a preset threshold; If the random number is greater than the preset threshold, the first probe subset is the attack subset and the second probe subset is the victim subset. If the random number is less than or equal to the preset threshold, then the first probe subset and the second probe subset are the attack subset, and the third probe subset is the victim subset.

[0008] By introducing random numbers, the victim subset of this iteration is determined, which is the subset of probes that need to be updated. The worse victim subset "discharges" to the better attack subset, learns from the attack subset, and absorbs some of the advantages of the attack subset to update its own probe combination, thereby promoting the quality of the next generation of the population to evolve upward.

[0009] Furthermore, according to the darkroom probe selection method of one aspect of this disclosure, damage inspection is performed on the current population, including the updated victim subset, comprising: Calculate the number of times each probe subset in the current population is subjected to discharge attacks; Based on the number of times, identify whether there is a damaged subset in the current population; where the damaged subset has been subjected to discharge attacks more than the number of times it has been subjected to the attack. When a damaged subset exists in the current population, a new probe subset is generated based on the probes deployed in the darkroom and the preset selection number, and the damaged subset is replaced by the new probe subset. If no damaged subset exists in the current population, identify whether a partial damaged subset exists in the current population. If so, select a probe from the darkroom to replace any probe in the partial damaged subset.

[0010] By performing a damage check at the end of each iteration, a subset of damage that has been subjected to too many discharge attacks is identified and replaced with a randomly generated new subset of probes, introducing randomness and preventing the population from becoming stagnant.

[0011] Furthermore, according to one aspect of the anechoic chamber probe selection method of this disclosure, a next-generation population is generated based on the current population after damage examination, including: The subset of probes included in the current population after damage inspection is used as the candidate subset; Calculate the Euclidean distance between any two candidate subsets; The candidate subset is divided into multiple microhabitats based on Euclidean distance. The microhabitat represents the behavior and position of organisms in the ecosystem. The radius of the microhabitat decreases linearly to a minimum with the number of iterations. Obtain representative subsets from each niche; the representative subsets are the candidate subsets with the highest fitness. Multiple next-generation probe subsets are generated based on the probe and a preset selection quantity; The next generation population is generated based on the representative subset and the next generation probe subset.

[0012] Dividing the candidate subset into multiple niches using Euclidean distance, and focusing on each potential "region" (niche) in the solution space, can prevent the loss of population diversity. Each niche retains only the candidate subset with the highest fitness, eliminating other similar solutions. This avoids the "monopoly" of the entire population by a dominant probe subset, forcing the algorithm to explore different probe combinations and avoiding getting trapped in local optima. This is particularly suitable for scenarios in multi-cluster channel models that need to cover multiple signal arrival angles.

[0013] Furthermore, the anechoic chamber probe selection method according to one aspect of this disclosure also includes: Each representative subset is then used as the target subset in turn; Calculate the normalized Hamming distance between the target subset and each of the remaining representative subsets; whereby the normalized Hamming distance is used to measure the difference between the target subset and each of the remaining representative subsets. Based on the shared radius and normalized Hamming distance, the shared function values ​​between the target subset and each remaining representative subset are determined; among them, the shared radius decreases linearly to a minimum value with the number of iterations. The shared fitness of the target subset is calculated based on the fitness of the target subset and the values ​​of each shared function. Use shared fitness as the fitness of the target subset.

[0014] By calculating the shared function value between the target subset and each remaining representative subset, we can determine whether the target subset is located in a high-density region. If the target subset is in a high-density region, its fitness needs to be penalized, resulting in a shared fitness that is much lower than its original fitness. If the target subset is not in a high-density region, its fitness needs to be rewarded, meaning its shared fitness is almost equal to its original fitness, encouraging the algorithm to explore that region. As the shared radius gradually decreases with the number of iterations, this dynamic calculation of shared fitness prevents the algorithm from local convergence. By penalizing the fitness of probe subsets in high-density regions, it simulates resource competition in nature, forcing the algorithm to expand into sparser regions of the solution space.

[0015] Furthermore, according to one aspect of the anechoic chamber probe selection method of this disclosure, a next-generation population is generated based on a representative subset and a next-generation probe subset, including: Based on the fitness of the representative subset and the next-generation probe subset, obtain the m candidate subsets with the smallest fitness; When the current iteration meets the preset conditions, m historical probe subsets are retrieved from the historical archive; the historical archive is used to store diverse subsets from the historical iterations. Replace the candidate subset with a subset of historical probes and generate the next generation population.

[0016] By establishing a historical archive to preserve high-quality and diverse subsets that emerge during the iteration process and periodically re-injecting them into the solution space, high-quality solutions are prevented from being eliminated during iteration, while new search directions are introduced. For example, it can be set up such that every 10 iterations, the worst-performing candidate subset is randomly sampled from the historical archive to replace it, injecting diversity into the next iteration. This simulates "gene recombination" in biological evolution, using the diversity of historical solutions to break the current search deadlock. This is especially suitable for quickly adjusting the probe selection strategy when the channel model changes abruptly.

[0017] According to another aspect of this disclosure, an anechoic chamber probe selection device is provided, comprising: The subset selection module is used to select a first probe subset, a second probe subset, and a third probe subset from the current population. The current population includes multiple probe subsets, which are generated based on the probes deployed in the darkroom and a preset selection number. The first probe subset, the second probe subset, and the third probe subset are sorted according to their fitness. The electrostatic discharge simulation module is used to identify the attack subset and the victim subset in the first probe subset, the second probe subset and the third probe subset according to the electrostatic discharge algorithm, and update the victim subset based on the attack subset; wherein, the victim subset is the party that suffers the discharge attack. The damage inspection module is used to perform damage inspection on the current population, including the updated victim subset; The population update module is used to generate the next generation population based on the current population after damage inspection; The iteration module is used to update the current population to the next generation population, repeatedly executing the steps of selecting the first probe subset, the second probe subset, and the third probe subset from the current population until the preset conditions for exiting the iteration are met, and the current population is taken as the final population. The probe selection module is used to select the optimal subset of probes from the final population based on fitness.

[0018] According to another aspect of this disclosure, a computer device is provided, including a memory, a processor, and a computer program stored in the memory, the processor executing the computer program to implement the method of one aspect above.

[0019] According to another aspect of this disclosure, a computer-readable storage medium is provided having a computer program stored thereon that, when executed by a processor, implements the method of one aspect above.

[0020] According to another aspect of this disclosure, a computer program product is provided, including a computer program that, when executed by a processor, implements the method of the above-described aspect.

[0021] It should be understood that both the foregoing general description and the following detailed description are exemplary and intended to provide further illustration of the claimed technology. Attached Figure Description

[0022] The above and other objects, features, and advantages of this disclosure will become more apparent from the more detailed description of the embodiments thereof in conjunction with the accompanying drawings. The drawings are provided to further illustrate the embodiments of this disclosure and form part of the specification. They are used together with the embodiments of this disclosure to explain the disclosure and do not constitute a limitation thereof. In the drawings, the same reference numerals generally represent the same components or steps.

[0023] Figure 1 This is a system architecture diagram illustrating the application of an anechoic chamber probe selection method according to an embodiment of this disclosure.

[0024] Figure 2 This is a flowchart illustrating a method for selecting a probe in an anechoic chamber according to an embodiment of this disclosure.

[0025] Figure 3 This is a flowchart illustrating another method for selecting a probe in an anechoic chamber according to an embodiment of the present disclosure.

[0026] Figure 4 This is a schematic diagram of the structure of a darkroom probe selection device according to an embodiment of the present disclosure.

[0027] Figure 5 This is a schematic diagram illustrating the structure of a computer device according to an embodiment of the present disclosure.

[0028] Figure 6 This is a schematic diagram illustrating a computer program product according to an embodiment of the present disclosure. Detailed Implementation

[0029] To make the objectives, technical solutions, and advantages of this disclosure more apparent, exemplary embodiments according to this disclosure will now be described in detail with reference to the accompanying drawings. Obviously, the described embodiments are merely some embodiments of this disclosure, and not all embodiments of this disclosure. It should be understood that this disclosure is not limited to the exemplary embodiments described herein.

[0030] With the application of cutting-edge technologies, the antenna systems (transmitting and receiving signals) of devices such as terminals and base stations are increasingly exhibiting high complexity and integration. However, the highly integrated nature of antenna systems limits the deployment space of conducted test connectors, thus restricting conducted testing. Therefore, the multi-probe anechoic chamber method has become a feasible solution for performance testing of complex device antenna systems.

[0031] The multi-probe anechoic chamber method, a technique based on an anechoic chamber and probe antenna arrays, often requires deploying a large number of probes within the anechoic chamber to expand the coverage of the test area. However, in the wireless signal testing of 5G communication equipment, how to more accurately and quickly simulate the real communication environment using fewer probes within the anechoic chamber has become a pressing problem to be solved.

[0032] In over-the-air (OTA) testing of 5G Multiple-Input Multiple-Output (MIMO) terminals, when constructing the target channel in a multi-probe anechoic chamber, fixed algorithms are often used for probe selection and weight optimization, making it difficult to balance accuracy, efficiency, and cost. Specifically, traditional iterative algorithms (such as multi-shot) achieve selection by successively eliminating low-weight probes. However, due to the unidirectional iterative nature, key probe combinations are easily mistakenly deleted, leading to local optima and insufficient global simulation accuracy. Its core flaw lies in the irreversibility of the algorithm's logic, making it impossible to backtrack and verify the necessity of eliminated probes. Especially when dealing with 3GPP-defined three-dimensional channel models that include azimuth and elevation angles, it struggles to cover complex spatial features, requiring a large number of probes to maintain accuracy, resulting in a surge in hardware costs.

[0033] Heuristic algorithms in related technologies, such as Differential Evolution (DE) and Particle Swarm Optimization (PSO), can explore a wider solution space, but they are sensitive to initialization parameters. Different initial values ​​can easily lead to fluctuations in results, and their convergence speed is relatively slow. For example, when dealing with a large number of probes (e.g., 481), traditional DE algorithms lack a dynamic parameter adjustment mechanism, resulting in low search efficiency in the early stages and a tendency to get stuck in local convergence later. Repeated iterations are required to approach the optimal solution, making it difficult to meet the needs of real-time testing. The root cause lies in the failure to fully utilize the spatial correlation and dynamic characteristics of the channel and the inability to adaptively adjust the search strategy. This leads to the need for redundant probes to compensate for accuracy in highly directional 5G millimeter-wave channels, resulting in a simultaneous increase in computational complexity and hardware costs.

[0034] Furthermore, related technologies generally treat probe location and weight optimization separately, failing to achieve joint optimization. For example, weight calculation based on spatial correlation is not closely integrated with probe location distribution, leading to deviations between the reconstructed channel's angular power spectrum and the target. This is especially problematic in multi-cluster channel models, where the dynamic correlation between clusters is ignored, requiring additional computational resources to compensate for the error. The key obstacle to solving these problems lies in designing an algorithmic framework that balances global search and local fine-tuning, and dynamically integrates the spatiotemporal characteristics of the channel model. Traditional algorithms, due to architectural limitations, struggle to overcome these bottlenecks.

[0035] The above description, with reference to the accompanying drawings, illustrates a method, apparatus, device, medium, and product for selecting probes in an anechoic chamber according to embodiments of the present disclosure. By introducing an electrostatic discharge algorithm and setting a fitness function, subsets with low fitness (victim subsets) learn from subsets with high fitness (attack subsets), enabling precise and rapid selection of the optimal probe subset, i.e., effective probes, from a large number of candidate probes, significantly improving the accuracy and speed of probe selection. This disclosure exhibits fast convergence speed; under different channel models and probe weight optimization methods, the fitness converges within 100 iterations, maintaining good performance under different channel models and test scenarios, demonstrating strong robustness.

[0036] To facilitate understanding of this embodiment, a detailed description of the anechoic chamber probe selection method disclosed in this disclosure is provided first. The execution entity of the anechoic chamber probe selection method provided in this disclosure is generally a computer device with certain computing capabilities. This computer device may include, for example, a terminal device, a server, or other processing devices. The terminal device may be a user equipment (UE), mobile device, user terminal, terminal, cellular phone, cordless phone, personal digital assistant (PDA), handheld device, computing device, vehicle-mounted device, wearable device, etc. In some possible implementations, this anechoic chamber probe selection method can be implemented by a processor calling computer-readable instructions stored in memory.

[0037] like Figure 1 The diagram shown illustrates the system architecture of the anechoic chamber probe selection method provided in this embodiment. A Simple Sectored Multiprobe Anechoic Chamber (SS-MPAC) system is used to simulate a 5G three-dimensional channel, including an anechoic chamber 1, probe antennas 2, a millimeter-wave switching circuit 3, a millimeter-wave channel simulator 4, and a base station simulator 5. The anechoic chamber 1 is an anechoic chamber. The probe antennas 2 are mounted on a probe wall with a fan-shaped structure, covering the direction of the incident signal. The probe antennas 2 maintain the same test distance from the device under test and are spaced at a certain angle. The millimeter-wave channel simulator 4 and the base station simulator 5 are used to create a millimeter-wave multipath fading environment and transmit signals, respectively. The millimeter-wave switching circuit 3 connects the RF port of the millimeter-wave channel simulator 4 to the corresponding probes to achieve dynamic beamforming testing.

[0038] After setting up the SS-MPAC system, determine the optimization goals, which include the following steps: In channel simulation, an optimization method based on beamforming power spectrum is used to determine probe weights. The beamforming power spectrum is derived using the Bartlett beamforming algorithm. Under ideal channel conditions, for a given target power angle spectrum... The spatial correlation matrix of the target channel is expressed as:

[0039] in, Represents a spatial angle, derived from the angle of elevation. and azimuth composition, Indicates the angle from the measured device to the space angle. The normalized direction vector, Indicates the given target power angular spectrum. This indicates the conjugate transpose.

[0040] The nth element can be represented as: .in, , It is a wave vector. It is the position vector of the nth probe in the DUT.

[0041] Assuming weight vector The ideal Bartlett beamforming power spectrum is:

[0042] In the SS-MPAC system, with K probes, the simulated Bartlett beamforming power spectrum is:

[0043] in, The spatial correlation matrix representing the received signal is determined by the probe weights and the antenna structure of the DUT:

[0044] in, It is the probe weight vector. This indicates the position of the k-th probe. From array antenna to space angle directional vector, The nth element is:

[0045] in, It is the distance from the k-th probe to the n-th antenna element in the device under test. This represents the free space path loss.

[0046] To obtain the probe power weights, the optimization objective is set as minimizing the difference between the target power spectrum and the reconstructed power spectrum:

[0047]

[0048] in, M represents the number of sampling points within the directional sector covered by the fan-shaped probe wall. With the probe position fixed, the probe power weights can be obtained by solving this optimization problem. PSP (Fitness Performance Profile) is used as the system evaluation metric for millimeter-wave terminal testing.

[0049] Power spectrum distortion factor Defined as:

[0050] Under ideal conditions, the power spectrum of the target channel beamforming power spectrum and the simulated beamforming power spectrum of the SS-MPAC system are defined as follows:

[0051] The value of PSP ranges from [0,1], and the larger the value of PSP, the higher the similarity.

[0052] Based on the above embodiments, this embodiment provides a method for selecting a probe in an anechoic chamber, such as... Figure 2 The diagram shows a flowchart of the anechoic chamber probe selection method, including steps S201-S207: S201: Initialization, generating a subset of probes.

[0053] Randomly generate N probe subsets The initial population is formed, with each probe subset... , which means that K probes are randomly selected from M probes in the SS-MPAC system.

[0054] The PSP is used as the fitness function for each probe subset, and a counter is set for each probe subset to record the number of times that probe subset becomes a victim subset. The archive counter is initialized. Empty.

[0055] S202: Electrostatic discharge event simulation.

[0056] Based on the improved electrostatic discharge (ESD) algorithm, the process of discharge between objects is simulated. The probe subset is updated by "discharge" to explore a better solution.

[0057] Specifically, in each iteration, from N probe subsets Three probe subsets were randomly selected from the data. and ,calculate and After determining the fitness, categorize the groups according to their fitness from highest to lowest. and Sort. Then generate random numbers. Through random numbers To determine the discharge type, this embodiment sets the preset threshold to 0.5, as follows: (1) When When the value is greater than 0.5, an ESD event is simulated between two probe subsets, and the attack subset is... (The subset with the highest fitness), the victim subset is (The subset with moderate fitness), then the updated for:

[0058] in, , It is a random factor.

[0059] (2) When At that time, an ESD event was simulated between two subsets of probes, and the attacked subset was... and (The subset with the highest fitness and the middle value), the victim subset is (The subset with the smallest fitness), after update for:

[0060] in, , It is a random factor.

[0061] Each time an ESD event occurs in a victim subset, the counter for that probe subset is incremented by 1.

[0062] The essence of simulating an ESD event in this embodiment is to enable the weaker approach (victim) to learn from the stronger approach (attacker), absorbing some of the attacker's advantages to update its own probe setup. For example, The probes included have been compatible with Unlike other methods, updating a probe subset means updating the combination of probes included in the subset, i.e., the probe position, which specifically refers to the spatial angle (including azimuth) of the probe.

[0063] S203: Boundary handling.

[0064] The system identifies whether a probe position exceeds the search space, i.e., whether it is an invalid probe index. If so, it resets the probe to a random valid position within the probe wall. Specifically, it ensures that the probe positions in the updated probe subset are valid and do not exceed the probe wall range.

[0065] S204: Damage inspection.

[0066] The counters for each probe subset are checked one by one to identify whether the number of ESD events experienced by each probe subset exceeds a threshold. In this embodiment, the threshold is set to 3. (1) If a certain probe subset is subjected to more than 3 ESDs, it is considered that the probe subset is damaged. It is necessary to randomly select K probes from the M probes in the SS-MPAC system to regenerate a new probe subset and replace the damaged probe subset.

[0067] (2) If a subset of probes is subjected to ESD less than or equal to 3 times, generate random numbers. ,when If the probe subset is considered to be partially damaged, it means that one probe in the probe subset is damaged, and one probe is randomly selected from the M probes in the SS-MPAC system for replacement.

[0068] S205: Add a diversity preservation mechanism.

[0069] To address the issues of traditional ESD algorithms easily getting trapped in local optima and insufficient population diversity during iteration, this embodiment introduces three diversity preservation mechanisms into the basic ESD algorithm framework: niche-based elite selection, dynamic calculation of shared fitness, and injection of diversity archives.

[0070] These diversity preservation mechanisms, by simulating population differentiation, resource competition, and historical memory strategies in biological evolution, ensure a dynamic balance between global search and local optimization in the ESD algorithm, avoiding premature convergence. The following are the specific implementation details: (1) Niche-based elite selection: During the elite selection process, any two probe subsets are calculated. and Euclidean distance between them:

[0071] in, Represents a subset of probes and The Euclidean distance between them Represents a subset of probes The Middle The position coordinates of each probe.

[0072] Based on the Euclidean distance, the probe subset is divided into multiple "niches" in the probe selection space. Each niche retains only the optimal probe subset, forcing the ESD algorithm to explore different regions rather than focusing on a single local optimum.

[0073] In the field of biological evolution, a microhabitat is defined as the behavior and position of an organism within an ecosystem, encompassing its taxonomic status, morphological characteristics, physiological responses, and behaviors. An initial microhabitat radius is then set. As the number of iterations t increases, the niche radius is dynamically adjusted:

[0074] in, Denotes the niche radius during the t-th iteration. To maximize the number of iterations, the ESD algorithm shifts from coarse-grained search to fine-grained optimization.

[0075] For each subset of probes within a small population, only the subset with the highest PSP value is retained, while other similar solutions (probe subsets) are removed. This avoids the dominant probe subset "monopolizing" the entire population and forces the ESD algorithm to explore different probe combinations. It is especially suitable for scenarios in multi-cluster channel models that need to cover multiple signal arrival angles.

[0076] For ease of understanding, assume that N=100 probe subsets are randomly generated initially, and the number of niches can be set to 40. The 100 probe subsets are grouped according to Euclidean distance, and only the probe subset with the highest PSP value is retained for each niche. This yields 40 high-quality probe subsets, which are then used in the next iteration. It should be noted that the number of probe subsets and niches can be chosen according to actual needs. This embodiment is merely for better explanation of the technical solution and should not be considered a limitation of this embodiment.

[0077] (2) Dynamically calculate shared fitness: To prevent local convergence of the ESD algorithm, this embodiment penalizes the PSP (Programmable Surface Spatial Components) of a subset of probes in high-density regions, simulating resource competition in nature and forcing the ESD algorithm to expand into sparser regions of the solution space. Specifically: First, by normalizing the Hamming distance Represents any two probe subsets and Differences:

[0078] in, Represents a subset of probes and The normalized Hamming distance, This indicates the number of probes shared by the two probe subsets. Used to characterize a subset of probes and The degree of similarity (or difference) of the probes included, for example: This indicates that the two probe subsets selected exactly the same probes. This indicates that the two probe subsets have selected completely different probes.

[0079] For each probe subset Its shared fitness The calculation formula is:

[0080] in, Represents a subset of probes Original fitness (PSP); Indicates the shared radius. As the number of iterations decreases linearly to the minimum value This is used to prevent the denominator from being zero.

[0081] Will Defined as a shared function, it represents how fitness should be "shared" between two solutions. The value of this shared function is... Maximum at time (meaning the subset of probes) and (If they are too close to each other, severe penalties are needed), the value of the shared function changes accordingly. Increases and decreases. A "peaked" function (such as a trigonometric function or Gaussian function) is typically used when... Greater than the shared radius At that time, the value of the shared function is 0 (meaning a subset of probes). and (They are too far apart to be punished). By punishing clustered subsets of probes, the clustering of solutions into a single dense region can be suppressed.

[0082] Summing the shared function in the denominator to calculate the probe subset. The shared fitness is calculated by summing the shared fitness values ​​of the subsets of probes selected from all other probe subsets in the population, where n is the number of niches. This step calculates the shared fitness for the high-quality probe subsets selected from the niches. A large sum indicates that many probe subsets share the shared fitness with the population. Very similar (small distance) If a region is very crowded, the denominator becomes larger, and its shared fitness will be much smaller than the original value. This is the penalty; if the sum is very small (close to 1), it means that there are almost no probe subsets with... resemblance, If the value is in a sparse region, then the denominator is approximately 1, and its shared fitness is almost equal to the original value. This is a reward.

[0083] (3) Injecting diversity archives: This embodiment saves historical diverse solutions that appear during the historical iteration process by establishing a "historical archive". For example, some high-quality diverse subsets that are deleted in the elite selection of niches are periodically reinjected into the solution space to avoid high-quality solutions being eliminated due to iteration, and at the same time, new search directions can be introduced.

[0084] First, create an archive of a certain size. To preserve historical diverse solutions, the archive update rules are as follows for each iteration:

[0085] The iteration replacement cycle can be preset. For example, every 10 iterations, random samples are taken from the diverse subsets of the historical archive, and the m candidate subsets with the smallest PSP in the entire candidate subset are replaced. The number of m is greater than or equal to 1. In this embodiment, m=1.

[0086] By injecting diverse archives, it simulates "gene recombination" in biological evolution, and uses the diversity of historical solutions to break the current search deadlock. It is especially suitable for quickly adjusting the probe selection strategy when the channel model changes suddenly.

[0087] In summary, niche elite selection ensures the population covers multiple solution space regions, shared fitness guides the algorithm to search sparse regions, and historical archiving replenishes the next generation of the population with diverse historical solutions. Through the synergistic effect of these three mechanisms, the improved ESD algorithm can maintain solution space diversity in complex channel models, significantly enhancing the global optimality and convergence efficiency of probe selection.

[0088] S206: Population Update.

[0089] Repeat S201. Based on the M probes in the SS-MPAC system, regenerate multiple probe subsets. The difference is that the regenerated probe subsets and the high-quality probe subsets retained in the niche are a total of N.

[0090] Taking N=100 and the number of niches=40 as an example, the number of probe subsets in the population remains unchanged at N in each iteration. Therefore, 60 probe subsets need to be regenerated, and 100 candidate subsets are needed to form the next generation population. The PSP of each candidate subset is calculated to form a new solution set, ensuring that the population always evolves towards better solutions. Additionally, every 10 iterations, a random sample is taken from the historical archive of diverse subsets, replacing the candidate subset with the smallest PSP.

[0091] S207: Output the optimal probe subset and weights.

[0092] After satisfying the maximum number of iterations, the optimal solution, i.e., the subset of probes with the largest PSP, is extracted from the final population. The probe set and its corresponding power weights are determined based on the optimal solution for deployment in the actual OTA testing system. The maximum number of iterations can be set according to actual needs, such as 300 or 400.

[0093] This embodiment fully considers the complex characteristics of 5G millimeter-wave channels and the cost issues brought about by the increased number of probes, effectively improving test performance and efficiency. By simulating the electrostatic discharge process and introducing a diversity retention mechanism, effective probes can be accurately screened from a large number of candidate probes, significantly improving the accuracy of probe selection and thus enhancing the accuracy of channel reconstruction. When simulating the CDL-A and CDL-B channel models defined in the 3GPP TR 38.901 protocol, the PSP of the improved ESD algorithm is 6.66% and 4.93% higher than that of the traditional multi-shot algorithm, respectively, and is closer to the power spectrum of the target channel, proving that this embodiment can more effectively simulate the spatial characteristics of the target channel.

[0094] This embodiment demonstrates fast convergence. Under different channel models and probe weight optimization methods, the improved ESD algorithm enables PSP to converge within 100 iterations, while the basic ESD algorithm requires many more iterations. This means that in practical testing, using the improved ESD algorithm of this embodiment can find suitable probe combinations more quickly, significantly shortening testing time and improving testing efficiency. Moreover, this embodiment maintains good performance under different channel models and testing scenarios, exhibiting strong robustness. Whether it's a simple single-cluster channel model or a complex multi-cluster model, this embodiment effectively reduces the number of probes required while ensuring channel reproduction accuracy. Even with only 5 probes selected, the PSP value can still exceed 90%, fully demonstrating its adaptability in complex scenarios. By comparing the simulation results of the improved ESD algorithm and the traditional ESD algorithm under different channel models, the effectiveness of the method proposed in this embodiment is clearly demonstrated. Compared with traditional methods, the method in this embodiment overcomes problems such as easy getting trapped in local optima and sensitivity to initialization, meets the requirements of high precision, high efficiency and low cost for 5G MIMO OTA testing, and provides reliable technical support for equipment testing, network deployment and performance evaluation of 5G and future communication systems.

[0095] Based on the above embodiments, this embodiment also provides a method for selecting a probe for an anechoic chamber, such as... Figure 3 The diagram shown is another flowchart for anechoic chamber probe selection, including S301-S306: S301: Select the first probe subset, the second probe subset, and the third probe subset from the current population.

[0096] The current population includes multiple probe subsets, which are generated based on the probes deployed in the darkroom and a preset selection number. The first probe subset, the second probe subset, and the third probe subset are sorted according to their fitness.

[0097] S302: Based on the electrostatic discharge algorithm, identify the attack subset and victim subset in the first probe subset, the second probe subset and the third probe subset, and update the victim subset based on the attack subset.

[0098] The victim subset is the party that suffers the electrical attack.

[0099] S303: Perform damage checks on the current population, including the updated victim subset.

[0100] S304: Generate the next generation population based on the current population after damage inspection.

[0101] S305: Update the current population to the next generation population, and repeat the steps of selecting the first probe subset, the second probe subset and the third probe subset from the current population until the preset condition for exiting the iteration is met, and use the current population as the final population.

[0102] S306: Select the optimal subset of probes from the final population based on fitness.

[0103] In one or more embodiments, S302 further includes: Generate a random number and determine whether the random number is greater than a preset threshold. If the random number is greater than the preset threshold, the first probe subset is the attack subset and the second probe subset is the victim subset. If the random number is less than or equal to the preset threshold, the first probe subset and the second probe subset are the attack subset and the third probe subset is the victim subset.

[0104] This embodiment uses random numbers. The preset threshold is set to 0.5, as detailed below: (1) When When the value is greater than 0.5, an ESD event is simulated between two probe subsets, and the attack subset is... (First probe subset), victim subset is (Second probe subset), then the updated for:

[0105] in, , It is a random factor.

[0106] (2) When At that time, an ESD event was simulated between two subsets of probes, and the attacked subset was... and (First probe subset and second probe subset), victim subset is (Third probe subset), updated for:

[0107] in, , It is a random factor.

[0108] In one or more embodiments, S303 further includes: Calculate the number of times each probe subset in the current population is subjected to discharge attacks; based on the number of attacks, identify whether there is a damaged subset in the current population, wherein the number of times the damaged subset is subjected to discharge attacks is greater than a threshold; if there is a damaged subset in the current population, generate a new probe subset based on the probes deployed in the darkroom and a preset selection quantity, and replace the damaged subset with the new probe subset; if there is no damaged subset in the current population, identify whether there is a partial damaged subset in the current population, and if so, select a probe from the darkroom to replace any probe in the partial damaged subset.

[0109] Specifically, assuming the threshold number of times is 3, if the number of times the probe is subjected to discharge attack is greater than 3, then the probe subset is identified as the damaged subset, and a new probe subset needs to be generated by randomly selecting K probes from the darkroom to replace this damaged subset.

[0110] If the number of times the system suffers a discharge attack is less than or equal to 3, a random number must also be generated. ,judge Is it less than a preset threshold (e.g., 0.2)? If the probe subset is considered to be partially damaged, it means that one probe in the probe subset is damaged and needs to be replaced by any probe selected from the darkroom.

[0111] In one or more embodiments, S304 further includes: The current population, after damage inspection, includes a subset of probes as candidate subsets. The Euclidean distance between any two candidate subsets is calculated. Based on the Euclidean distance, the candidate subsets are divided into multiple microhabitats. A representative subset is obtained from each microhabitat. Multiple next-generation probe subsets are generated based on the number of probes and a preset selection quantity. The next-generation population is generated based on the representative subsets and the next-generation probe subsets. Here, a microhabitat represents the behavior and position of an organism in the ecosystem, and the radius of the microhabitat decreases linearly to a minimum with the number of iterations. The representative subset is the candidate subset with the highest fitness.

[0112] Specifically, the formula for calculating Euclidean distance is:

[0113] in, Represents any two candidate subsets and The Euclidean distance between them Indicates the subset to be selected The Middle The position coordinates of each probe.

[0114] Based on the Euclidean distance, the candidate subset is divided into multiple "niches" in the probe selection space. Each niche retains only the optimal representative subset, forcing the ESD algorithm to explore different regions rather than focusing on a single local optimum.

[0115] In one or more embodiments, it further includes: Each representative subset is sequentially selected as the target subset; the normalized Hamming distance between the target subset and each remaining representative subset is calculated; based on the shared radius and the normalized Hamming distance, the shared function value between the target subset and each remaining representative subset is determined; based on the fitness of the target subset and each shared function value, the shared fitness of the target subset is calculated; and the shared fitness is used as the fitness of the target subset. The normalized Hamming distance measures the difference between the target subset and each remaining representative subset; the shared radius decreases linearly to a minimum with the number of iterations.

[0116] Specifically, the formula for calculating the normalized Hamming distance is:

[0117] in, Represents any two target subsets and representative subset The normalized Hamming distance, This indicates the number of probes shared by the two subsets. Used to characterize subsets and The degree of similarity (or difference) of the probes included, for example: This indicates that the two subsets selected exactly the same probes. This indicates that the probes selected by the two subsets are completely different.

[0118] Shared function value , indicating two subsets and How should we "share" fitness among them? The maximum value of the time-sharing function means the target subset and The elements are too close together, so a target subset is needed. Severe penalties, shared function values ​​as Increases and decreases. When Greater than the shared radius When the shared function value is 0, it means the target subset and They are too far apart to need a target subset. punish.

[0119] The formula for calculating shared fitness is:

[0120] in, Represents the target subset Original fitness (PSP); Indicates the shared radius. As the number of iterations decreases linearly to the minimum value This is used to prevent the denominator from being zero.

[0121] Summing the shared function values ​​in the denominator to calculate the target subset. The shared function values ​​of the other representative subsets are summed. If the sum is large, it indicates that many representative subsets share the same function values. Very similar (small distance) If a region is very crowded, the denominator becomes larger, and its shared fitness will be much smaller than the original value. This is the penalty; if the sum is very small (close to 1), it means that there are almost no representative subsets. resemblance, If the value is in a sparse region, then the denominator is approximately 1, and its shared fitness is almost equal to the original value. This is a reward.

[0122] In one or more embodiments, a next-generation population is generated based on a representative subset and a next-generation probe subset, including: Based on the fitness of the representative subset and the next-generation probe subset, obtain the m candidate subsets with the smallest fitness; when the current iteration meets the preset conditions, obtain m historical probe subsets from the historical archive; replace the candidate subsets with the historical probe subsets and generate the next-generation population. The historical archive is used to store diverse subsets from previous iterations.

[0123] Specifically, a "historical archive" is established to preserve high-quality historical probe subsets that emerge during the iterative process. These subsets might be high-quality, diverse subsets that were removed during niche elite selection. These subsets are periodically reinjected into the solution space to prevent high-quality solutions from being eliminated during iteration, and also to introduce new search directions. An iterative replacement cycle (preset conditions) can be pre-set; for example, every 10 iterations, random samples are taken from the historical probe subsets in the historical archive to replace the m smallest candidate subsets of the PSP, where m is greater than or equal to 1. In this embodiment, m=1.

[0124] By injecting diverse archives, it simulates "gene recombination" in biological evolution, and uses the diversity of historical solutions to break the current search deadlock. It is especially suitable for quickly adjusting the probe selection strategy when the channel model changes suddenly.

[0125] According to another aspect of the embodiments of this disclosure, an anechoic chamber probe selection device is provided, such as... Figure 4 As shown, the device includes: The subset selection module 401 is used to select a first probe subset, a second probe subset, and a third probe subset from the current population; wherein, the current population includes multiple probe subsets, the probe subsets are generated based on the probes deployed in the darkroom and a preset selection number, and the first probe subset, the second probe subset, and the third probe subset are sorted according to their fitness. The electrostatic discharge simulation module 402 is used to identify the attack subset and the victim subset in the first probe subset, the second probe subset and the third probe subset according to the electrostatic discharge algorithm, and update the victim subset based on the attack subset; wherein, the victim subset is the party that suffers the discharge attack. Damage inspection module 403 is used to perform damage inspection on the current population, including the updated victim subset; Population update module 404 is used to generate the next generation population based on the current population after damage inspection; The iteration module 405 is used to update the current population to the next generation population, and repeatedly execute the steps of selecting the first probe subset, the second probe subset and the third probe subset from the current population until the preset condition for exiting the iteration is met, and the current population is taken as the final population. The probe selection module 406 is used to select the optimal subset of probes from the final population according to fitness.

[0126] The anechoic chamber probe selection device and the anechoic chamber probe selection method provided in this disclosure are based on the same inventive concept and have the same beneficial effects as the methods they employ, operate, or implement.

[0127] This disclosure also provides a computer device for performing the above-described anechoic chamber probe selection method. Please refer to... Figure 5 It illustrates a schematic diagram of a computer device provided by some embodiments of this disclosure. For example... Figure 5 As shown, the computer device 5 includes: a processor 500, a memory 501, a bus 502, and a communication interface 503. The processor 500, the communication interface 503, and the memory 501 are connected via the bus 502. The memory 501 stores a computer program that can run on the processor 500. When the processor 500 runs the computer program, it executes the anechoic chamber probe selection method provided in any of the foregoing embodiments of this disclosure.

[0128] The memory 501 may include high-speed random access memory (RAM) or non-volatile memory, such as at least one disk storage device. Communication between this device network element and at least one other network element is achieved through at least one communication interface 503 (which can be wired or wireless), such as the Internet, wide area network, local area network, metropolitan area network, etc.

[0129] Bus 502 can be an ISA bus, PCI bus, or EISA bus, etc. The bus can be divided into an address bus, a data bus, a control bus, etc. The memory 501 is used to store programs. After receiving an execution instruction, the processor 500 executes the program. The anechoic chamber probe selection method disclosed in any of the foregoing embodiments of this disclosure can be applied to the processor 500, or implemented by the processor 500.

[0130] The processor 500 may be an integrated circuit chip with signal processing capabilities. In implementation, each step of the above method can be completed by the integrated logic circuitry in the hardware of the processor 500 or by instructions in software form. The processor 500 may be a general-purpose processor, including a central processing unit (CPU), a network processor (NP), etc.; it may also be a digital signal processor (DSP), an application-specific integrated circuit (ASIC), an off-the-shelf programmable gate array (FPTA), or other programmable logic devices, discrete gate or transistor logic devices, or discrete hardware components. It can implement or execute the methods, steps, and logic block diagrams disclosed in the embodiments of this disclosure. The general-purpose processor may be a microprocessor or any conventional processor. The steps of the methods disclosed in the embodiments of this disclosure can be directly embodied in the execution of a hardware decoding processor, or executed by a combination of hardware and software modules in the decoding processor. The software modules may reside in random access memory, flash memory, read-only memory, programmable read-only memory, electrically erasable programmable memory, registers, or other mature storage media in the art. The storage medium is located in memory 501. The processor 500 reads the information in memory 501 and, in conjunction with its hardware, completes the steps of the above method.

[0131] The computer equipment provided in this disclosure and the anechoic chamber probe selection method provided in this disclosure are based on the same inventive concept and have the same beneficial effects as the methods they employ, operate, or implement.

[0132] This disclosure also provides a computer-readable storage medium corresponding to the anechoic chamber probe selection method provided in the foregoing embodiments. The computer-readable storage medium is an optical disc, on which a computer program (i.e., a computer program product) is stored. When the computer program is run by a processor, it executes the anechoic chamber probe selection method provided in any of the foregoing embodiments.

[0133] It should be noted that examples of the computer-readable storage medium may also include, but are not limited to, phase-change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other optical and magnetic storage media, which will not be elaborated here.

[0134] The computer-readable storage medium provided in the above embodiments of this disclosure and the anechoic chamber probe selection method provided in the embodiments of this disclosure are based on the same inventive concept and have the same beneficial effects as the methods adopted, run or implemented by the applications stored therein.

[0135] This disclosure also provides a computer program product; please refer to [reference needed]. Figure 6 The computer program product 600 carries program code, namely computer program 601. The instructions included in the computer program 601 can be used to execute the steps of the anechoic chamber probe selection method described in the above method embodiments. For details, please refer to the above method embodiments, which will not be repeated here.

[0136] The aforementioned computer program product can be implemented through hardware, software, or a combination thereof. In one optional embodiment, the computer program product is specifically embodied in a computer storage medium; in another optional embodiment, the computer program product is specifically embodied in a software product, such as a software development kit (SDK), etc.

[0137] The basic principles of this disclosure have been described above with reference to specific embodiments. However, it should be noted that the advantages, benefits, and effects mentioned in this disclosure are merely examples and not limitations, and should not be considered as essential features of each embodiment of this disclosure. Furthermore, the specific details disclosed above are for illustrative and facilitative purposes only, and are not limitations. These details do not limit the scope of this disclosure to the necessity of employing the aforementioned specific details for implementation.

[0138] The block diagrams of devices, apparatuses, devices, and systems disclosed herein are merely illustrative examples and are not intended to require or imply that they must be connected, arranged, or configured in the manner shown in the block diagrams. As those skilled in the art will recognize, these devices, apparatuses, devices, and systems can be connected, arranged, and configured in any manner. Words such as “comprising,” “including,” “having,” etc., are open-ended terms meaning “including but not limited to,” and are used interchangeably with them. The terms “or” and “and” as used herein refer to the terms “and / or,” and are used interchangeably with them unless the context clearly indicates otherwise. The term “such as” as used herein refers to the phrase “such as but not limited to,” and is used interchangeably with it.

[0139] Additionally, as used herein, the "or" used in a list of items beginning with "at least one" indicates a separate list, such that a list of, for example, "at least one of A, B, or C" means A or B or C, or AB or AC or BC, or ABC (i.e., A and B and C). Furthermore, the word "exemplary" does not imply that the described example is preferred or better than other examples.

[0140] It should also be noted that in the systems and methods of this disclosure, the components or steps can be decomposed and / or recombined. These decompositions and / or recombinations should be considered as equivalent solutions to this disclosure.

[0141] Various changes, substitutions, and modifications can be made to the technology described herein without departing from the teachings defined by the appended claims. Furthermore, the scope of the claims of this disclosure is not limited to the specific aspects of the processes, machines, manufactures, events, means, methods, and actions described above. Currently existing or later-developed processes, machines, manufactures, events, means, methods, or actions that perform substantially the same function or achieve substantially the same result as the corresponding aspects described herein can be utilized. Therefore, the appended claims include such processes, machines, manufactures, events, means, methods, or actions within their scope.

[0142] The above description of the disclosed aspects is provided to enable any person skilled in the art to make or use this disclosure. Various modifications to these aspects will be readily apparent to those skilled in the art, and the general principles defined herein may be applied to other aspects without departing from the scope of this disclosure. Therefore, this disclosure is not intended to be limited to the aspects shown herein, but rather to be carried out within the widest scope consistent with the principles and novel features disclosed herein.

[0143] The above description has been given for purposes of illustration and description. Furthermore, this description is not intended to limit the embodiments of this disclosure to the forms disclosed herein. Although numerous exemplary aspects and embodiments have been discussed above, those skilled in the art will recognize certain variations, modifications, alterations, additions, and sub-combinations thereof.

Claims

1. A method for selecting a probe in an anechoic chamber, characterized in that, include: Select a first probe subset, a second probe subset, and a third probe subset from the current population; wherein, the current population includes multiple probe subsets, the probe subsets are generated based on the probes deployed in the darkroom and a preset selection number, and the first probe subset, the second probe subset, and the third probe subset are sorted according to their fitness; According to the electrostatic discharge algorithm, the attack subset and the victim subset in the first probe subset, the second probe subset and the third probe subset are identified, and the victim subset is updated based on the attack subset; wherein, the victim subset is the party that suffers the discharge attack. Perform damage checks on the current population, including the updated victim subset; Based on the current population after damage inspection, generate the next generation population; The current population is updated to the next generation population. The steps of selecting the first probe subset, the second probe subset and the third probe subset from the current population are repeated until the preset condition for exiting the iteration is met, and the current population is taken as the final population. Based on the fitness, the optimal subset of probes is selected from the final population.

2. The anechoic chamber probe selection method as described in claim 1, characterized in that, Based on the electrostatic discharge algorithm, the attack subsets and victim subsets in the first probe subset, the second probe subset, and the third probe subset are identified, including: Generate a random number and determine whether the random number is greater than a preset threshold; If the random number is greater than the preset threshold, then the first probe subset is the attack subset and the second probe subset is the victim subset; If the random number is less than or equal to the preset threshold, then the first probe subset and the second probe subset are attack subsets, and the third probe subset is victim subset.

3. The anechoic chamber probe selection method as described in claim 1, characterized in that, Damage checks are performed on the current population, including the updated victim subset, including: Calculate the number of times each of the probe subsets in the current population suffers a discharge attack; Based on the number of times, identify whether there is a damaged subset in the current population; wherein, the damaged subset has been subjected to discharge attacks more than a number threshold. When a damaged subset exists in the current population, a new probe subset is generated based on the probes deployed in the darkroom and the preset selection number, and the damaged subset is replaced by the new probe subset; If no damaged subset exists in the current population, identify whether a partial damaged subset exists in the current population. If so, select a probe from the darkroom to replace any probe in the partial damaged subset.

4. The anechoic chamber probe selection method as described in claim 1, characterized in that, Based on the current population after damage inspection, a next-generation population is generated, including: The subset of probes included in the current population after damage inspection is used as the candidate subset; Calculate the Euclidean distance between any two of the candidate subsets; The candidate subset is divided into multiple microhabitats based on the Euclidean distance, wherein the microhabitat represents the behavior and position of organisms in the ecosystem, and the radius of the microhabitat decreases linearly to a minimum with the number of iterations; Obtain a representative subset from each of the aforementioned niches; wherein, the representative subset is the candidate subset with the highest fitness. Multiple next-generation probe subsets are generated based on the probe and the preset selection quantity; Based on the representative subset and the next-generation probe subset, a next-generation population is generated.

5. The anechoic chamber probe selection method as described in claim 4, characterized in that, Also includes: Each of the aforementioned representative subsets is then used as the target subset. Calculate the normalized Hamming distance between the target subset and each of the remaining representative subsets; wherein the normalized Hamming distance is used to measure the difference between the target subset and each of the remaining representative subsets; Based on the shared radius and the normalized Hamming distance, the shared function value between the target subset and each of the remaining representative subsets is determined; wherein, the shared radius decreases linearly to a minimum value with the number of iterations; The shared fitness of the target subset is calculated based on the fitness of the target subset and the shared function values ​​of each subset. The shared fitness is used as the fitness of the target subset.

6. The anechoic chamber probe selection method as described in claim 4, characterized in that, Based on the representative subset and the next-generation probe subset, a next-generation population is generated, including: Based on the fitness of the representative subset and the next-generation probe subset, obtain the m candidate subsets with the smallest fitness; When the current iteration meets the preset conditions, m historical probe subsets are obtained from the historical archive; wherein, the historical archive is used to store diverse subsets in the historical iterations; The candidate subset is replaced with the historical probe subset, and the next generation population is generated.

7. A anechoic chamber probe selection device, characterized in that, include: The subset selection module is used to select a first probe subset, a second probe subset, and a third probe subset from the current population; wherein, the current population includes multiple probe subsets, the probe subsets are generated based on the probes deployed in the darkroom and a preset selection number, and the first probe subset, the second probe subset, and the third probe subset are sorted according to their fitness. The electrostatic discharge simulation module is used to identify the attack subset and the victim subset in the first probe subset, the second probe subset and the third probe subset according to the electrostatic discharge algorithm, and update the victim subset based on the attack subset; wherein, the victim subset is the party that suffers the discharge attack. Damage inspection module, used to perform damage inspection on the current population including the updated victim subset; The population update module is used to generate the next generation population based on the current population after damage inspection; The iteration module is used to update the current population to the next generation population, and repeatedly execute the steps of selecting the first probe subset, the second probe subset and the third probe subset from the current population until the preset condition for exiting the iteration is met, and the current population is taken as the final population. The probe selection module is used to select the optimal subset of probes from the final population according to the fitness.

8. A computer embedded device, comprising a memory, a processor, and a computer program stored in the memory, characterized in that, The processor executes the computer program to implement the method according to any one of claims 1 to 6.

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 described in any one of claims 1 to 6.

10. A computer program product, comprising a computer program, characterized in that, When the computer program is executed by a processor, it implements the method described in any one of claims 1 to 6.

Images