Multi-source signal interference parameter adaptive optimization configuration method and system

Abstract

The invention provides a multi-source signal interference parameter adaptive optimization configuration method and system, and relates to the technical field of signal interference, and the method comprises the steps: building an interference efficiency evaluation model through obtaining historical interference efficiency data and real-time parameter state data of a to-be-optimized signal interference source; and optimizing a resource allocation scheme according to the fitness function by adopting a self-adaptive genetic algorithm, and executing constraint verification and fitness correction based on the dynamic resource constraint matrix and the self-adaptive penalty function. According to the invention, the distribution efficiency and interference efficiency of signal interference resources can be improved, and system resource consumption is reduced.

Description

Technical Field

[0001] The present invention relates to signal interference technology, and particularly to a method and system for adaptively optimizing the configuration of multi-source signal interference parameters. Background Art

[0002] In modern electronic warfare and information warfare, signal interference technology, as an important countermeasure, has been widely applied in military and civilian fields. With the increasingly complex electromagnetic environment, multi-source signal interference systems need to achieve the optimal interference effect on target signals under limited resource conditions. How to reasonably allocate interference resources and optimize interference efficiency has become a key issue in current research.

[0003] Traditional signal interference parameter configuration methods mainly rely on empirical models or fixed algorithms, and are unable to flexibly allocate resources when facing complex and changeable electromagnetic environments. With the development of electronic countermeasure technology, especially in multi-source signal interference scenarios, multiple factors such as the cooperative effect between signal sources, resource constraint conditions, and real-time interference efficiency evaluation need to be considered.

[0004] The existing technologies have the following deficiencies in the optimization configuration of signal interference parameters: First, most of the existing interference parameter configuration methods adopt static optimization strategies and cannot be adaptively adjusted according to real-time interference efficiency and environmental changes, resulting in unstable interference effects and low resource utilization rates in complex and changeable electromagnetic environments.

[0005] Second, when traditional optimization algorithms deal with the resource allocation problem of multi-source signal interference, it is often difficult to effectively balance the relationship between global search and local optimization, and is prone to falling into local optimal solutions, unable to obtain the best interference configuration scheme. Especially when the resource constraint conditions are relatively strict, the practicality of the optimization results is limited.

[0006] Finally, most of the existing interference efficiency evaluation models are based on theoretical analysis or simplified assumptions, lack of full utilization of historical interference efficiency data, and are difficult to accurately evaluate interference effects, resulting in deviations between resource allocation schemes and actual requirements, and unable to achieve precise interference and efficient countermeasures. Summary of the Invention

[0007] Embodiments of the present invention provide a method and system for adaptively optimizing the configuration of multi-source signal interference parameters, which can solve the problems in the existing technologies.

[0008] In the first aspect of the embodiments of the present invention, a method for adaptively optimizing the configuration of multi-source signal interference parameters is provided, including: Obtaining historical interference efficiency data and real-time interference parameter status data of a signal interference source to be optimized, and constructing an interference efficiency evaluation model based on the historical interference efficiency data; The resource allocation scheme of the signal interference source to be optimized is optimized and iterated by using an adaptive genetic algorithm. The fitness value of the iterative population is calculated according to a preset fitness function, and the crossover probability and mutation probability of the adaptive genetic algorithm are dynamically adjusted based on the fitness value; A dynamic resource constraint matrix and an adaptive penalty function are constructed based on the interference effectiveness evaluation model. Constraint verification and fitness correction are performed on the resource allocation scheme based on the dynamic resource constraint matrix and the adaptive penalty function; The constraint violation degree of the population individuals is calculated, a penalty factor is determined according to the constraint violation degree, the individual fitness value is adaptively adjusted through the penalty factor, and the constraint parameters of the dynamic resource constraint matrix are updated based on the constraint violation degree; Based on the constraint parameters, it is judged whether the number of iterations reaches a preset iteration threshold. When the condition is met, the optimal resource allocation scheme is output.

[0009] Constructing an interference effectiveness evaluation model based on the historical interference effectiveness data includes: Interference effectiveness evaluation features are extracted from the historical interference effectiveness data, the influence relationship between the historical power allocation data and historical frequency band coverage data of the interference source on the historical interference gain data is established to form an interference effectiveness evaluation matrix, and an effectiveness evaluation function for each interference source is established based on the interference effectiveness evaluation matrix. The effectiveness evaluation function is used to predict the interference gain under different power allocations and frequency band coverages; An interference effectiveness evaluation structure is constructed according to the effectiveness evaluation function, and the evaluation parameters of the interference effectiveness evaluation structure are optimized and adjusted by using the historical interference effectiveness data to generate an interference effectiveness evaluation model that meets the preset evaluation accuracy.

[0010] Optimizing and iterating the resource allocation scheme of the signal interference source to be optimized by using an adaptive genetic algorithm, calculating the fitness value of the iterative population according to a preset fitness function, and dynamically adjusting the crossover probability and mutation probability of the genetic algorithm based on the fitness value includes: A fitness function for the resource allocation scheme of the signal interference source to be optimized is established. The resource allocation scheme of the signal interference source to be optimized is encoded according to the fitness function to obtain an initial population, the fitness value of each individual in the initial population is calculated, and the maximum population fitness value and the average population fitness value are calculated based on the fitness value of each individual; The crossover probability is dynamically adjusted according to the difference between the fitness value of the individuals to be crossed and the maximum population fitness value and the average population fitness value. When the fitness value of the individuals to be crossed is greater than the average population fitness value, a dynamically compressed crossover probability is adopted. When the fitness value of the individuals to be crossed is less than or equal to the average population fitness value, a fixed-maintained crossover probability is adopted; Dynamically adjust the mutation probability according to the difference between the fitness value of the individual to be mutated and the maximum fitness value and the average fitness value of the population. When the fitness value of the individual to be mutated is less than the maximum fitness value of the population, use a dynamically compressed mutation probability. When the fitness value of the individual to be mutated is greater than or equal to the average fitness value of the population, use a fixed-maintained mutation probability.

[0011] Establish a fitness function for the resource allocation scheme of the signal interference source to be optimized. Encode the resource allocation scheme of the signal interference source to be optimized according to the fitness function to obtain an initial population, including: According to the interference effectiveness, resource constraints and performance objectives of the signal interference source to be optimized, establish a fitness function for the resource allocation scheme of the signal interference source to be optimized. The fitness function includes an interference effectiveness term and a resource constraint penalty term; Perform real-number encoding on the resource allocation scheme of the signal interference source to be optimized according to interference power, frequency bandwidth and time allocation, and generate an initial population with a preset scale.

[0012] Construct a dynamic resource constraint matrix and an adaptive penalty function based on the interference effectiveness evaluation model. Perform constraint verification and fitness correction on the resource allocation scheme based on the dynamic resource constraint matrix and the adaptive penalty function, including: Obtain the transmission power, antenna gain, frequency band range and beam pointing angle of multiple interference sources. Construct power constraint matrix elements based on the ratio of the transmission power to the antenna gain, construct frequency constraint matrix elements based on the overlapping integral of the frequency band ranges, and construct space constraint matrix elements based on the deviation of the beam pointing angles; Combine the power constraint matrix elements, the frequency constraint matrix elements and the space constraint matrix elements to form a dynamic resource constraint matrix. The dynamic resource constraint matrix is used to represent the power constraint relationship, frequency constraint relationship and space constraint relationship between interference sources; Obtain the expected interference effectiveness value and the current interference effectiveness value. Construct an effectiveness penalty term based on the square difference between the expected interference effectiveness value and the current interference effectiveness value. Multiply the effectiveness penalty term by a dynamic weight to obtain a comprehensive effectiveness penalty value; Calculate the constraint penalty term based on the dynamic resource constraint matrix. The constraint penalty term is obtained by applying constraint type weights to the power constraint matrix elements, the frequency constraint matrix elements and the space constraint matrix elements respectively; Calculate the power violation degree, frequency violation degree and space violation degree according to the dynamic resource constraint matrix; correct the original fitness value according to the comprehensive effectiveness penalty value to obtain an effectiveness-corrected fitness value, and correct the effectiveness-corrected fitness value based on the power violation degree, the frequency violation degree and the space violation degree to obtain a final fitness value.

[0013] Calculate the power violation degree, frequency violation degree, and space violation degree according to the dynamic resource constraint matrix; correct the original fitness value according to the comprehensive effectiveness penalty value to obtain the effectiveness-corrected fitness value, and correct the effectiveness-corrected fitness value based on the power violation degree, the frequency violation degree, and the space violation degree to obtain the final fitness value, including: The dynamic resource constraint matrix includes a power constraint matrix, a frequency constraint matrix, and a space constraint matrix; obtain the node power, frequency set, and space distance parameters, and obtain the original power violation degree by summing the product of the deviation square of the node power ratio and the power constraint matrix element and the dynamic weight based on the power constraint matrix; Calculate the original frequency violation degree by summing the product of the deviation square of the ratio of the frequency set intersection to the union and the frequency constraint matrix element and the frequency weight based on the frequency constraint matrix; calculate the original space violation degree by summing the product of the deviation square of the ratio of the space distance to the maximum distance and the space constraint matrix element and the space weight based on the space constraint matrix; perform noise filtering processing on the original power violation degree, the original frequency violation degree, and the original space violation degree to obtain the filtered violation degree; Use the comprehensive effectiveness penalty value to correct the original fitness value to obtain the effectiveness-corrected fitness value, perform weighted fusion on the filtered violation degree based on the effectiveness-corrected fitness value to obtain the overall violation degree, and correct the effectiveness-corrected fitness value using the overall violation degree to obtain the final fitness value, where the final fitness value is calculated by the difference of the product term of the effectiveness-corrected fitness value and the overall violation degree.

[0014] Calculate the constraint violation degree of the population individuals, determine the penalty factor according to the constraint violation degree, perform adaptive adjustment on the individual fitness value through the penalty factor, and update the constraint parameters of the dynamic resource constraint matrix based on the constraint violation degree, including: Calculate the constraint violation degree for each interference source, where the constraint violation degree is obtained by summing the product of the deviation square of the product ratio of the transmission power and the antenna gain and the power constraint threshold and the dynamic weight; Construct an individual penalty factor based on the constraint violation degree, where the individual penalty factor is calculated by the exponential function of the sum of the basic penalty coefficient and the weighted violation degree, and the weighted violation degree is obtained by weighting each violation degree by the constraint type weight; Calculate the group cooperation penalty factor according to the individual penalty factor, where the group cooperation penalty factor is obtained by the weighted sum of the mean of the individual penalty factors and the variance of the violation degree, and the variance of the violation degree is multiplied by the smoothing coefficient; Calculate the gradient of the constraint violation degree with respect to the constraint parameters based on the group collaborative penalty factor, and dynamically update the constraint parameters based on the gradient.

[0015] In a second aspect of the embodiments of the present invention, there is provided a multi-source signal interference parameter adaptive optimization and configuration system, including: A first unit, configured to obtain historical interference effectiveness data and real-time interference parameter status data of a signal interference source to be optimized, and construct an interference effectiveness evaluation model based on the historical interference effectiveness data; A second unit, configured to optimize and iterate the resource allocation scheme of the signal interference source to be optimized by using an adaptive genetic algorithm, calculate the fitness value of the iterative population according to a preset fitness function, and dynamically adjust the crossover probability and mutation probability of the adaptive genetic algorithm based on the fitness value; A third unit, configured to construct a dynamic resource constraint matrix and an adaptive penalty function based on the interference effectiveness evaluation model, and perform constraint verification and fitness correction on the resource allocation scheme based on the dynamic resource constraint matrix and the adaptive penalty function; A fourth unit, configured to calculate the constraint violation degree of population individuals, determine a penalty factor according to the constraint violation degree, adaptively adjust the individual fitness value through the penalty factor, and update the constraint parameters of the dynamic resource constraint matrix based on the constraint violation degree; A fifth unit, configured to determine whether the number of iterations reaches a preset iteration threshold based on the constraint parameters, and output the optimal resource allocation scheme when the condition is satisfied.

[0016] In a third aspect of the embodiments of the present invention, there is provided an electronic device, including: A processor; A memory for storing instructions executable by the processor; Wherein, the processor is configured to call the instructions stored in the memory to execute the method described above.

[0017] In a fourth aspect of the embodiments of the present invention, there is provided a computer-readable storage medium, on which computer program instructions are stored, and when the computer program instructions are executed by a processor, the method described above is implemented.

[0018] The beneficial effects of this application are as follows: The multi-source signal interference parameter adaptive optimization and configuration method provided by the present invention realizes the intelligent optimization and configuration of interference parameters, effectively improves the overall combat effectiveness of the electronic countermeasure system, and solves the problem that the traditional manual configuration method cannot meet the requirements of complex battlefield environments.

[0019] This method optimizes parameters by constructing an interference effectiveness evaluation model and an adaptive genetic algorithm, can respond to changes in the battlefield environment in real time, dynamically adjust the interference parameter configuration, achieve the optimal allocation of resources in a multi-source signal interference scenario, and improve the flexibility and effectiveness of the countermeasure system.

[0020] By adopting the innovative design of a dynamic resource constraint matrix and an adaptive penalty function, the problem of resource conflict in multi-source signal interference is solved, ensuring the practical executability of the scheme. At the same time, through the calculation of the constraint violation degree and the dynamic adjustment of the penalty factor, the optimization result better meets the actual battlefield requirements, achieving the maximization of interference effectiveness. Brief Description of the Drawings

[0021] Figure 1 It is a schematic flowchart of the method for adaptively optimizing the configuration of multi-source signal interference parameters in an embodiment of the present invention; Figure 2 It is a logic block diagram for constructing an interference effectiveness evaluation model in an embodiment of the present invention; Figure 3 It is a schematic diagram showing the optimization effect of the resource allocation scheme under different interference scenarios in an embodiment of the present invention; Figure 4 It is a flowchart of the method for dynamically optimizing the resources of interference sources based on multi-dimensional constraints in an embodiment of the present invention; Figure 5 It is a flowchart of the method for calculating the violation degree and optimizing the fitness based on multi-dimensional constraints in an embodiment of the present invention. Detailed Embodiment

[0022] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only some of the embodiments of the present invention, rather than all of them. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0023] The technical solutions of the present invention will be described in detail below with specific embodiments. These specific embodiments can be combined with each other, and the same or similar concepts or processes may not be repeated in some embodiments.

[0024] Figure 1 It is a schematic flowchart of the method for adaptively optimizing the configuration of multi-source signal interference parameters in an embodiment of the present invention, as Figure 1 shown, the method includes: Obtain the historical interference effectiveness data and real-time interference parameter status data of the signal interference source to be optimized, and construct an interference effectiveness evaluation model based on the historical interference effectiveness data; An adaptive genetic algorithm is used to optimize and iterate the resource allocation scheme of the signal interference source to be optimized. The fitness value of the iterative population is calculated according to a preset fitness function, and the crossover probability and mutation probability of the adaptive genetic algorithm are dynamically adjusted based on the fitness value; A dynamic resource constraint matrix and an adaptive penalty function are constructed based on the interference effectiveness evaluation model, and constraint verification and fitness correction are performed on the resource allocation scheme based on the dynamic resource constraint matrix and the adaptive penalty function; The constraint violation degree of the population individuals is calculated, a penalty factor is determined according to the constraint violation degree, the individual fitness value is adaptively adjusted through the penalty factor, and the constraint parameters of the dynamic resource constraint matrix are updated based on the constraint violation degree; Based on the constraint parameters, it is judged whether the number of iterations reaches a preset iteration threshold, and when the condition is satisfied, the optimal resource allocation scheme is output.

[0025] In an alternative embodiment, constructing an interference effectiveness evaluation model based on the historical interference effectiveness data includes: Interference effectiveness evaluation features are extracted from the historical interference effectiveness data, the influence relationship between the historical power allocation data, historical frequency band coverage data of the interference source and the historical interference gain data is established to form an interference effectiveness evaluation matrix, and an effectiveness evaluation function for each interference source is established based on the interference effectiveness evaluation matrix. The effectiveness evaluation function is used to predict the interference gain under different power allocations and frequency band coverages; An interference effectiveness evaluation structure is constructed according to the effectiveness evaluation function, and the evaluation parameters of the interference effectiveness evaluation structure are optimized and adjusted by using the historical interference effectiveness data to generate an interference effectiveness evaluation model that meets the preset evaluation accuracy.

[0026] Interference effectiveness evaluation features are extracted from historical interference effectiveness data. In practical applications, historical interference effectiveness data includes historical power allocation data, historical frequency band coverage data and corresponding historical interference gain data of multiple interference sources.

[0027] For each interference source, historical data within the past six months is collected, including power allocation values, frequency band coverage ranges and corresponding interference gain values at different time points. For example, for interference source A, 500 groups of historical data samples are collected, and each group of samples includes power allocation values (such as 23dBm, 27dBm, 30dBm, etc.), frequency band coverage ranges (such as 2.4 - 2.5GHz, 5.1 - 5.8GHz, etc.) and actually measured interference gain values (such as 12dB, 15dB, 18dB, etc.).

[0028] For the collected historical data, perform feature engineering to extract effective evaluation features. Through data analysis, it is found that there is a positive correlation between power allocation and interference gain, and there is also a certain correlation between the width of frequency band coverage and interference gain. Therefore, parameters such as power allocation value, frequency band coverage width, and center frequency offset are used as the main feature parameters. At the same time, considering the influence of environmental factors, auxiliary features such as environmental noise level and relative distance between the interference source and the target are also introduced.

[0029] Establish the influence relationship of the historical power allocation data and historical frequency band coverage data of the interference source on the historical interference gain data to form an interference effectiveness evaluation matrix. In this embodiment, the following method is adopted: Perform normalization processing on the historical data of each interference source, map the power allocation data to the range of 0 - 1, and convert the frequency band coverage data into two feature parameters of center frequency and bandwidth. For example, the power allocation value of 30 dBm is normalized to 0.9, and the frequency band coverage of 2.4 - 2.5 GHz is converted into a center frequency of 2.45 GHz and a bandwidth of 100 MHz.

[0030] Construct an interference effectiveness evaluation matrix. The rows of this matrix represent different historical data samples, and the columns represent the extracted feature parameters, including the normalized power allocation value, center frequency, bandwidth, etc. The last column is the corresponding interference gain value. For example, for the first set of historical data of interference source A, a row of the evaluation matrix may be [0.8, 2.45, 0.1, 15], representing the normalized power value, center frequency (GHz), bandwidth (GHz), and interference gain (dB), respectively.

[0031] By analyzing the relationship between each feature in the evaluation matrix and the interference gain, establish an effectiveness evaluation function for each interference source based on the evaluation matrix. In this embodiment, for each interference source, a random forest algorithm is used to construct the effectiveness evaluation function: For interference source A, through processing 500 sets of historical data, a random forest model containing 100 decision trees is constructed. Each decision tree is trained based on different feature subsets and sample subsets, and finally the prediction result of the interference gain is obtained through the way of ensemble learning. This model can effectively capture the non - linear influence relationship between power allocation and frequency band coverage on the interference gain.

[0032] When the power of interference source A is set to 28 dBm (normalized to 0.85) and the frequency band coverage is 2.4 - 2.48 GHz (center frequency 2.44 GHz, bandwidth 80 MHz), the interference gain value predicted by the effectiveness evaluation function is 14.6 dB, which is not much different from the actual measured value of 14.8 dB, indicating that the model has good prediction ability.

[0033] Construct an interference effectiveness evaluation structure according to the effectiveness evaluation function, and optimize and adjust the evaluation parameters using historical interference effectiveness data.

[0034] Based on the previously constructed random forest model, further design an interference effectiveness evaluation structure. This structure includes four parts: a data preprocessing module, a feature extraction module, a model prediction module, and a result output module. The data preprocessing module is responsible for normalizing the input power allocation and frequency band coverage data; the feature extraction module is responsible for extracting effective features from the processed data; the model prediction module calls the trained random forest model to predict the interference gain; the result output module converts the prediction result into a standard format and outputs it.

[0035] To ensure that the evaluation accuracy reaches the preset standard, it is necessary to optimize and adjust the model parameters. In this embodiment, the cross-validation method is used to optimize the model. The historical data is divided into a training set and a validation set in a ratio of 8:2, and the key parameters of the random forest, including the number of decision trees, the maximum depth of the tree, and the feature selection method, are optimized through the grid search method.

[0036] Taking interference source A as an example, through parameter optimization, the optimal number of decision trees is determined to be 120, the maximum depth is 15, and the minimum number of samples in the leaf nodes is 5. After retraining the model with this set of optimized parameters, the average prediction error on the validation set is reduced from the initial 1.2 dB to 0.8 dB, meeting the preset evaluation accuracy requirement (error less than 1.0 dB).

[0037] To verify the generalization ability of the model, actual application tests are also carried out. Under different environmental conditions, multiple groups of new power and frequency band parameters are set for interference source A, and the interference gain values are obtained through model prediction and actual measurement respectively. The test results show that in 10 groups of new parameter settings, the average error between the model prediction value and the actual measurement value is 0.85 dB, and the maximum error is 1.3 dB, which verifies the effectiveness and applicability of the constructed model.

[0038] Through the above methods, an interference effectiveness evaluation model based on historical interference effectiveness data is successfully constructed. This model can accurately predict the interference gain under different power allocation and frequency band coverage conditions, providing an effective decision support tool for the optimal configuration of the interference system.

[0039] Figure 2 The logic block diagram for constructing the interference effectiveness evaluation model in the embodiment of the present invention is as follows: The figure shows the construction process of an interference effectiveness evaluation model, which is a complete system from historical data analysis to model construction. Starting from historical interference effectiveness data, through the interference effectiveness evaluation feature extraction link, the data is divided into three parallel analysis directions: historical power allocation data analysis, historical frequency band coverage data analysis, and historical interference gain data analysis. The analysis results of these three dimensions of data will be summarized in the influence relationship modeling link, and then an interference effectiveness evaluation matrix will be constructed. On this basis, an effectiveness evaluation function is established, and a closed-loop optimization is formed through the optimization and adjustment of evaluation parameters, and finally a complete interference effectiveness evaluation model is constructed. The entire process reflects the systematic thinking from data collection, feature extraction, multi-dimensional analysis, relationship modeling to final model construction. In particular, the comprehensiveness and accuracy of the evaluation are ensured through the three-dimensional analysis of historical data, and the dynamic optimization ability of the model is ensured through the feedback loop of parameter optimization and adjustment.

[0040] In an optional implementation manner, an adaptive genetic algorithm is used to optimize and iterate the resource allocation scheme of the signal interference source to be optimized. Calculating the fitness value of the iterative population according to a preset fitness function, and dynamically adjusting the crossover probability and mutation probability of the genetic algorithm based on the fitness value, including: Establish a fitness function for the resource allocation scheme of the signal interference source to be optimized, encode the resource allocation scheme of the signal interference source to be optimized according to the fitness function to obtain an initial population, calculate the fitness value of each individual in the initial population, and calculate the maximum population fitness value and the average population fitness value based on the fitness value of each individual; Dynamically adjust the crossover probability according to the difference between the fitness value of the individual to be crossed and the maximum population fitness value and the average population fitness value. When the fitness value of the individual to be crossed is greater than the average population fitness value, use a dynamically compressed crossover probability. When the fitness value of the individual to be crossed is less than or equal to the average population fitness value, use a fixed maintenance crossover probability; Dynamically adjust the mutation probability according to the difference between the fitness value of the individual to be mutated and the maximum population fitness value and the average population fitness value. When the fitness value of the individual to be mutated is less than the maximum population fitness value, use a dynamically compressed mutation probability. When the fitness value of the individual to be mutated is greater than or equal to the average population fitness value, use a fixed maintenance mutation probability.

[0041] Establishing the fitness function for the resource allocation scheme of signal interference sources is the first step in implementing this method. The design of the fitness function is directly related to the effect and direction of optimization. In a certain military electronic countermeasure system, the fitness function can comprehensively consider interference effectiveness, resource consumption, and tactical requirements. For example, the fitness value can be represented by the weighted sum of the interference success rate multiplied by 0.6, the spectrum utilization rate multiplied by 0.3, and the power efficiency multiplied by 0.1. If the interference success rate of a certain scheme is 85%, the spectrum utilization rate is 70%, and the power efficiency is 60%, then its fitness value is 85×0.6 + 70×0.3 + 60×0.1 = 78.5.

[0042] Encode the resource allocation scheme of signal interference sources according to the fitness function to obtain the initial population. The encoding method can adopt real number encoding, directly representing the parameter values of resource allocation. For example, for a resource allocation scheme containing three parameters: transmit power, operating frequency, and antenna gain, it can be represented by a chromosome composed of three real numbers. If an individual is encoded as [25, 2450, 12], it means the transmit power is 25 watts, the operating frequency is 2450 MHz, and the antenna gain is 12 dB.

[0043] The initial population can be composed of randomly generated multiple resource allocation schemes that meet the constraint conditions. Assume the population size is 50, then randomly generate 50 resource allocation schemes as the initial population. Calculate the fitness values of each individual in the initial population, and record the maximum fitness value and the average fitness value of the population. For example, if the fitness value of the optimal individual in the initial population is 85.2, and the average fitness value of all individuals is 72.6, then record the maximum fitness value of the population as 85.2 and the average fitness value of the population as 72.6.

[0044] During the optimization iteration process, the core innovation lies in dynamically adjusting the crossover probability and mutation probability according to the individual fitness. For the crossover operation, an adaptive adjustment strategy is adopted. When the fitness value of the individuals to be crossed is greater than the average fitness value of the population, a dynamically compressed crossover probability is used. For example, if the fitness value of an individual to be crossed is 78.5, which is greater than the average fitness value of the population 72.6, then its crossover probability can be dynamically calculated by subtracting the fitness difference ratio from the basic crossover probability (such as 0.9). The fitness difference ratio is (78.5 - 72.6) / (85.2 - 72.6) = 0.467, multiplied by the compression coefficient 0.5, resulting in 0.234. Subtracting this value from the basic crossover probability gives the actual crossover probability 0.9 - 0.234 = 0.666.

[0045] When the fitness value of the individual to be crossed is less than or equal to the average fitness value of the population, a fixed-maintained crossover probability is adopted. For example, if the fitness value of the individual to be crossed is 70.2, which is less than the average fitness value of the population 72.6, then the basic crossover probability 0.9 is directly used as its crossover probability. This strategy ensures that individuals with lower fitness have a higher probability of performing crossover operations, increasing population diversity.

[0046] An adaptive adjustment strategy is also adopted for the mutation operation. When the fitness value of the individual to be mutated is less than the maximum fitness value of the population, a dynamically compressed mutation probability is adopted. For example, the fitness value of a certain individual to be mutated is 80.3, which is less than the maximum fitness value of the population 85.2. Then its mutation probability can be dynamically calculated by subtracting the proportion of the fitness difference from the basic mutation probability (such as 0.1). The proportion of the fitness difference is (85.2 - 80.3) / (85.2 - 72.6) = 0.389. Multiplying it by the compression coefficient 0.7 gives 0.272. Subtracting this value from the basic mutation probability gives the actual mutation probability 0.1 - 0.272×0.1 = 0.073.

[0047] When the fitness value of the individual to be mutated is greater than or equal to the average fitness value of the population, a fixed-maintained mutation probability is adopted. For example, if the fitness value of the individual to be mutated is 85.2, which is equal to the maximum fitness value of the population, then the basic mutation probability 0.1 is directly used as its mutation probability. This strategy ensures that individuals with higher fitness have a lower probability of mutating, maintaining excellent genes.

[0048] In the crossover operation, the arithmetic crossover method can be adopted. For example, the two parent individuals are [25, 2450, 12] and [30, 2400, 10] respectively. The crossover point is selected as the second position and the crossover coefficient is 0.7. Then the two offspring individuals are [25, 2415, 12] and [30, 2435, 10] respectively, where 2415 = 2450×0.7 + 2400×0.3 and 2435 = 2450×0.3 + 2400×0.7.

[0049] In the mutation operation, the uniform mutation method can be adopted. For example, the third position of an individual [25, 2450, 12] mutates, and the mutation range is [8, 15]. Then it may mutate to [25, 2450, 9.5], where 9.5 is a randomly generated value within the range of [8, 15].

[0050] After the crossover and mutation operations, calculate the fitness values of the newly generated individuals and merge them with the original population. Select excellent individuals according to the fitness values to form a new generation of population. The selection method can adopt the elite retention strategy to ensure that the optimal individual directly enters the next generation, and the remaining individuals are selected by the roulette wheel method.

[0051] The iterative process continuously repeats the above steps, updates the maximum fitness value and average fitness value of the population, and dynamically adjusts the crossover probability and mutation probability until the termination condition is met. The termination condition can be reaching a preset maximum number of iterations (such as 500 times), or the change in the optimal fitness value for multiple consecutive generations (such as 50 generations) being less than a threshold (such as 0.001).

[0052] Practical application cases show that this adaptive genetic algorithm has remarkable effects in the optimization of signal interference source resource allocation. In the optimization of a certain radar jamming system, after adopting this method, the jamming success rate has increased from the original 78% to 92%, the spectrum utilization rate has increased by 15%, the power efficiency has increased by 20%, and the algorithm convergence speed is about 35% faster than that of the traditional genetic algorithm.

[0053] The strategy of adaptively adjusting the crossover probability and mutation probability effectively balances the global search ability and local fine search ability of the algorithm, avoids the premature convergence problem, and at the same time speeds up the convergence speed. By adopting different parameter adjustment strategies for individuals with different fitness values, both the diversity of the population and the optimization efficiency are maintained, providing an efficient and reliable optimization method for signal interference source resource allocation.

[0054] In an optional implementation manner, a fitness function for the resource allocation scheme of the signal interference source to be optimized is established, and the resource allocation scheme of the signal interference source to be optimized is encoded according to the fitness function to obtain an initial population, including: According to the interference effectiveness, resource constraints, and performance objectives of the signal interference source to be optimized, a fitness function for the resource allocation scheme of the signal interference source to be optimized is established. The fitness function includes an interference effectiveness term and a resource constraint penalty term; The resource allocation scheme of the signal interference source to be optimized is encoded in terms of interference power, frequency bandwidth, and time allocation to generate an initial population of a preset scale.

[0055] According to the interference effectiveness, resource constraints, and performance objectives of the signal interference source to be optimized, a fitness function for the resource allocation scheme of the signal interference source to be optimized is established. This fitness function consists of two parts: an interference effectiveness term and a resource constraint penalty term.

[0056] When establishing the interference effectiveness term, the interference degree of the interference source on the target communication system is comprehensively considered. Specifically, the interference effectiveness term can be characterized by the reduction degree of the signal-to-interference-plus-noise ratio of the target system. Suppose the target system has N channels. For the i-th channel, its signal-to-noise ratio before being interfered is SNR_i, and the signal-to-interference-plus-noise ratio after being interfered is SINR_i. Then the interference effectiveness of the i-th channel can be expressed as the difference between SNR_i and SINR_i. The interference effectiveness of the entire system is the weighted average of the interference effectiveness of all channels, and the weights can be set according to the importance of each channel.

[0057] For a communication system with three channels, if the importance weights of each channel are 0.5, 0.3, and 0.2 respectively, the original signal-to-noise ratios are 20 dB, 18 dB, and 15 dB respectively, and the signal-to-interference-plus-noise ratios after being interfered drop to 5 dB, 8 dB, and 10 dB respectively, then the total interference effectiveness is: 0.5×(20 - 5) + 0.3×(18 - 8) + 0.2×(15 - 10) = 7.5 + 3 + 1 = 11.5 dB. The higher the interference effectiveness, the more effective the interference is.

[0058] When establishing the resource constraint penalty term, the power constraint, bandwidth constraint, and time constraint of the interference source are mainly considered. When the resource allocation scheme exceeds the resource constraint, appropriate penalties need to be given to reduce the fitness function value.

[0059] In terms of power constraint, it is assumed that the upper limit of the total power of the interference source is P_max, and the total power of the current allocation scheme is P_total. If P_total is greater than P_max, then the power penalty term is the penalty coefficient α multiplied by (P_total - P_max); otherwise, the power penalty term is 0.

[0060] In terms of bandwidth constraint, it is assumed that the upper limit of the total bandwidth of the interference source is B_max, and the total bandwidth of the current allocation scheme is B_total. If B_total is greater than B_max, then the bandwidth penalty term is the penalty coefficient β multiplied by (B_total - B_max); otherwise, the bandwidth penalty term is 0.

[0061] In terms of time constraint, it is assumed that the upper limit of the total time resource of the interference source is T_max, and the total time resource of the current allocation scheme is T_total. If T_total is greater than T_max, then the time penalty term is the penalty coefficient γ multiplied by (T_total - T_max); otherwise, the time penalty term is 0.

[0062] The fitness function can be expressed as the interference effectiveness term minus the sum of each penalty term. The penalty coefficients α, β, and γ need to be adjusted according to the specific application scenario and resource importance to ensure better interference effects under resource constraint conditions.

[0063] Set α = 10, β = 8, γ = 5. Assume that the interference effectiveness of a certain resource allocation scheme is 11.5 dB, the total power exceeds the upper limit by 0.5 W, the bandwidth does not exceed the upper limit, and the time resource exceeds the upper limit by 2 ms. Then the fitness function value of this scheme is: 11.5 - 10×0.5 - 8×0 - 5×2 = 11.5 - 5 - 0 - 10 = -3.5. The negative fitness function value indicates that this scheme is not feasible and needs to be further optimized.

[0064] The resource allocation scheme of the signal interference source to be optimized is encoded in real numbers according to interference power, frequency bandwidth, and time allocation to generate an initial population of a preset scale.

[0065] Real number encoding directly represents the resource allocation scheme as a real number vector, facilitating subsequent operations of the genetic algorithm. For a communication system with N channels, the encoding vector contains 3N elements, representing the power allocation, bandwidth allocation, and time allocation for each channel respectively.

[0066] The encoding vector X can be expressed as [P_1, P_2, ..., P_N, B_1, B_2, ..., B_N, T_1, T_2,..., T_N], where P_i represents the power allocated to the i-th channel, B_i represents the bandwidth allocated to the i-th channel, and T_i represents the time resource allocated to the i-th channel.

[0067] When generating the initial population, a random initialization method can be adopted. First, set the population size M, and then randomly generate M encoding vectors to form the initial population. When randomly generating, the reasonable range of resources needs to be considered. For example, the power allocation can be randomly generated within the range of [0, P_max], the bandwidth allocation within the range of [0, B_max], and the time allocation within the range of [0, T_max].

[0068] For a communication system with 3 channels, assuming the maximum power of the interference source is 10W, the maximum bandwidth is 100MHz, and the maximum time resource is 50ms, a possible encoding vector is [3.5, 2.8, 3.2, 30, 25, 40, 15, 20, 10], indicating that 3.5W, 2.8W, and 3.2W of power, 30MHz, 25MHz, and 40MHz of bandwidth, and 15ms, 20ms, and 10ms of time resources are allocated to the three channels respectively.

[0069] To ensure the diversity of the initial population, a uniform distribution random number generation method can be adopted, or prior knowledge can be combined to increase the sampling point density in key regions. For example, according to experience, sampling points may be increased in the high-power and narrow-bandwidth regions to improve the quality of the initial population.

[0070] Set the population size M = 100, then the initial population contains 100 such encoding vectors. Each encoding vector represents a possible resource allocation scheme, and these schemes will be evolutionarily optimized through the genetic algorithm later to find the optimal resource allocation scheme.

[0071] Through the above method, the establishment of the fitness function and the generation of the initial population of the resource allocation scheme of the signal interference source to be optimized are completed, laying a foundation for the subsequent optimization process.

[0072] Figure 3 Schematic diagram for optimizing the resource allocation scheme under different interference scenarios in the embodiments of the present invention: This figure is a hexagonal radar chart, showing the performance comparison between the present technical solution and the prior art in six different scenarios. The triangular markers in the figure represent the present technical solution, and the circular markers represent the prior art (fixed weight). From the specific data, in the communication interference scenario, the present technical solution reaches 92.5 points, an increase of 8.2 points compared with 84.3 points of the prior art; in the radar interference scenario, the present technical solution reaches 88.7 points, better than 78.5 points of the prior art; in the electronic reconnaissance interference aspect, the present technical solution scores 90.3 points, significantly higher than 81.2 points of the prior art; in the multi-target cooperative interference field, the present technical solution reaches 86.9 points, with an obvious increase compared with 73.6 points of the prior art; in the complex electromagnetic environment, the present technical solution obtains 84.2 points, better than 69.8 points of the prior art; in the low signal-to-noise ratio scenario, the present technical solution reaches 93.6 points, far exceeding 82.7 points of the prior art. Generally speaking, the present technical solution has achieved significant performance improvement in all six evaluation dimensions, with an average improvement of about 10 points, especially being the most prominent in the low signal-to-noise ratio and communication interference scenarios.

[0073] In an optional implementation manner, a dynamic resource constraint matrix and an adaptive penalty function are constructed based on the interference effectiveness evaluation model, and the constraint verification and fitness correction of the resource allocation scheme based on the dynamic resource constraint matrix and the adaptive penalty function include: Obtain the transmission power, antenna gain, frequency band range, and beam pointing angle of multiple interference sources, construct power constraint matrix elements based on the ratio of the transmission power to the antenna gain, construct frequency constraint matrix elements based on the overlapping integral of the frequency band ranges, and construct spatial constraint matrix elements based on the deviation of the beam pointing angles; Combine the power constraint matrix elements, the frequency constraint matrix elements, and the spatial constraint matrix elements to form a dynamic resource constraint matrix, and the dynamic resource constraint matrix is used to characterize the power constraint relationship, frequency constraint relationship, and spatial constraint relationship between interference sources; Obtain the expected interference effectiveness value and the current interference effectiveness value, construct an effectiveness penalty term based on the square difference between the expected interference effectiveness value and the current interference effectiveness value, and multiply the effectiveness penalty term by a dynamic weight to obtain a comprehensive effectiveness penalty value; Calculate a constraint penalty term based on the dynamic resource constraint matrix, and the constraint penalty term is obtained by applying constraint type weights to the power constraint matrix elements, the frequency constraint matrix elements, and the spatial constraint matrix elements respectively; Calculate the power violation degree, frequency violation degree, and spatial violation degree according to the dynamic resource constraint matrix; correct the original fitness value according to the comprehensive effectiveness penalty value to obtain the effectiveness-corrected fitness value, and correct the effectiveness-corrected fitness value based on the power violation degree, the frequency violation degree, and the spatial violation degree to obtain the final fitness value.

[0074] Obtain the key parameters of multiple interference sources, including transmission power, antenna gain, frequency band range, and beam pointing angle. Taking a certain wireless communication scenario as an example, assume that there are 4 interference sources in the system. Among them, the transmission power of interference source 1 is 15 watts, the antenna gain is 6 dB, and the frequency band range is 2.4 - 2.5 GHz. The beam pointing angle of interference source 2 is 45 degrees, the transmission power is 12 watts, and the frequency band range is 2.35 - 2.45 GHz.

[0075] Calculate the ratio of the transmission power to the antenna gain between each pair of interference sources when constructing the elements of the power constraint matrix. For interference source 1 and interference source 2, calculate the transmission power ratio 15 / 12 = 1.25, the antenna gain ratio 6 / 8 = 0.75, and the product of the two gives the power constraint matrix element 0.9375. For other pairs of interference sources, use a similar method to calculate and form the power constraint matrix.

[0076] Calculate the overlapping degree of the frequency band ranges of each pair of interference sources when constructing the elements of the frequency constraint matrix. For interference source 1 and interference source 2, the frequency band ranges are 2.4 - 2.5 GHz and 2.35 - 2.45 GHz respectively. The overlapping part is 2.4 - 2.45 GHz, the overlapping width is 0.05 GHz, which accounts for 50% of the frequency band width of interference source 1 and 50% of the frequency band width of interference source 2. Take the average value to obtain the frequency constraint matrix element 0.5. Calculate the frequency constraint matrix elements for all pairs of interference sources in this way.

[0077] Calculate the deviation degree of the beam pointing angles of each pair of interference sources when constructing the elements of the spatial constraint matrix. For interference source 1 and interference source 2, the beam pointing angles are 30 degrees and 45 degrees respectively, and the angle deviation is 15 degrees. Assuming that the angle deviation threshold is 90 degrees, the normalized deviation is 15 / 90 = 0.167, and the spatial constraint matrix element is 1 - 0.167 = 0.833. Form the spatial constraint matrix in this way.

[0078] Combine the power constraint matrix, the frequency constraint matrix, and the spatial constraint matrix to form a dynamic resource constraint matrix. This matrix is a three-dimensional matrix, where the third dimension represents the constraint type (power, frequency, space).

[0079] Obtain the expected interference effectiveness value and the current interference effectiveness value for constructing the effectiveness penalty term. Assume that the expected interference effectiveness value of the system is 85 points, and the current interference effectiveness evaluation value is 72 points. Calculate the square of the difference and multiply it by the dynamic weight to obtain the effectiveness penalty term. Based on the ratio of the interference effectiveness value to the expected value 72 / 85 = 0.847, calculate the comprehensive effectiveness penalty value as 0.153.

[0080] Calculate the constraint penalty term based on the dynamic resource constraint matrix. Apply a constraint type weight of 0.4 to the elements of the power constraint matrix, a constraint type weight of 0.35 to the elements of the frequency constraint matrix, and a constraint type weight of 0.25 to the elements of the space constraint matrix. Taking interference source 1 and interference source 2 as examples, the calculated constraint penalty term is 0.7584.

[0081] Calculate the power violation degree, frequency violation degree, and space violation degree according to the dynamic resource constraint matrix. For the power violation degree, calculate the deviation between the actual power value of each interference source and the elements of the power constraint matrix. Assume that the actual power adjustment of interference source 1 is 17 watts, and the square of the deviation from the corresponding constraint element in the power constraint matrix is multiplied by the dynamic weight to obtain the power violation degree. Similar calculations are performed for all pairs of interference sources and summed to obtain the total power violation degree.

[0082] When calculating the frequency violation degree, assume that the actual frequency band of interference source 1 is adjusted to 2.38 - 2.48 GHz, and the square of the deviation from the corresponding constraint element in the frequency constraint matrix is multiplied by the frequency weight to obtain the frequency violation degree. After summing, the total frequency violation degree is obtained.

[0083] When calculating the space violation degree, assume that the beam pointing angle of interference source 1 is adjusted to 35 degrees, and the square of the deviation from the corresponding constraint element in the space constraint matrix is multiplied by the space weight to obtain the space violation degree. After summing, the total space violation degree is obtained.

[0084] Correct the original fitness value according to the comprehensive effectiveness penalty value. Assume that the original fitness value is 680, and apply the comprehensive effectiveness penalty value of 0.153 for correction to obtain the effectiveness-corrected fitness value of 576.

[0085] Further correct the effectiveness-corrected fitness value based on the power violation degree, frequency violation degree, and space violation degree. Set the weights of the three violation degrees for weighted summation to obtain the overall violation degree. Apply the overall violation degree to correct the effectiveness-corrected fitness value to obtain the final fitness value.

[0086] This method can be used for interference suppression in radar systems. By dynamically adjusting the transmission power, frequency band range, and beam pointing angle of each interference source, the interference effectiveness is maximized while satisfying resource constraints. After optimization, the interference effectiveness is significantly improved, the resource utilization rate is improved, and the configurations of all interference sources meet the constraint conditions, achieving effective interference suppression.

[0087] This method is also applicable to the spectrum management of mobile communication networks. In a multi-cell cooperation scenario, the base stations are regarded as interference sources. By constructing a dynamic resource constraint matrix, the interference relationships between base stations are evaluated, and an adaptive penalty function is used to guide the resource allocation scheme to optimize in the direction of meeting the constraint conditions. Practice has proved that this method can effectively reduce the interference of users at the cell edge and improve the overall network throughput and spectrum efficiency.

[0088] This method also performs excellently in the multi-UAV cooperative reconnaissance mission. By establishing the power, frequency, and spatial constraint relationships between UAVs, the communication resource allocation of each UAV is optimized to ensure stable communication quality during the mission execution and avoid mutual interference at the same time. The test results show that the optimized resource allocation scheme can reduce the communication interference by about 30% and increase the information transmission success rate by about 25%.

[0089] Figure 4 The flowchart of the dynamic resource optimization method for interference sources based on multi-dimensional constraints according to the embodiments of the present invention is as follows: This flowchart shows a complete process of dynamic resource constraint and fitness optimization for interference sources. First, the system obtains the key parameters of multiple interference sources, including transmission power, antenna gain, frequency band range, and beam pointing angle, and constructs three types of constraint matrix elements based on these parameters: the power constraint matrix elements come from the ratio of transmission power to antenna gain, the frequency constraint matrix elements are derived from the overlapping integral of the frequency band range, and the spatial constraint matrix elements are based on the deviation of the beam pointing angle. Subsequently, these three types of constraint matrix elements are combined into a dynamic resource constraint matrix to characterize the power constraint, frequency constraint, and spatial constraint relationships between interference sources. The system obtains the expected interference efficiency value and the current efficiency value, constructs an efficiency penalty term through the square difference, and obtains a comprehensive efficiency penalty value through dynamic weights. Different types of weights are applied to the constraint penalty terms based on the dynamic resource constraint matrix. Finally, the system calculates the power violation degree, frequency violation degree, and spatial violation degree, corrects the original fitness value, and further optimizes the efficiency-corrected fitness value according to these violation degree indicators to finally obtain the final fitness value.

[0090] In an alternative embodiment, calculating the power violation degree, frequency violation degree, and spatial violation degree according to the dynamic resource constraint matrix; correcting the original fitness value according to the comprehensive efficiency penalty value to obtain an efficiency-corrected fitness value, and correcting the efficiency-corrected fitness value based on the power violation degree, the frequency violation degree, and the spatial violation degree to obtain the final fitness value includes: The dynamic resource constraint matrix includes a power constraint matrix, a frequency constraint matrix, and a space constraint matrix; obtaining the node power, frequency set, and space distance parameters, and obtaining the original power violation degree by summing the product of the deviation square of the node power ratio and the power constraint matrix element and the dynamic weight based on the power constraint matrix; Based on the frequency constraint matrix, calculating the sum of the product of the deviation square of the ratio of the intersection to the union of the frequency sets and the frequency constraint matrix element and the frequency weight to obtain the original frequency violation degree; based on the space constraint matrix, calculating the sum of the product of the deviation square of the ratio of the space distance to the maximum distance and the space constraint matrix element and the space weight to obtain the original space violation degree; performing noise filtering processing on the original power violation degree, the original frequency violation degree, and the original space violation degree to obtain the filtered violation degree; Using the comprehensive efficiency penalty value to correct the original fitness value to obtain the efficiency-corrected fitness value, based on the efficiency-corrected fitness value, performing weighted fusion on the filtered violation degree to obtain the overall violation degree, and using the overall violation degree to correct the efficiency-corrected fitness value to obtain the final fitness value, and the final fitness value is calculated by the difference of the product term of the efficiency-corrected fitness value and the overall violation degree.

[0091] Obtain the current power values of each node in the system. Suppose there are 5 nodes in a wireless communication system, and their power values are 12 watts, 15 watts, 9 watts, 18 watts, and 14 watts respectively. In the power constraint matrix, the constraint threshold between node 1 and node 2 is 0.7, indicating that the ratio of the power of node 1 to the power of node 2 should not exceed 0.7. Calculate the actual ratio as 12 / 15 = 0.8, and the deviation from the constraint threshold 0.7 is 0.1. Squaring this deviation gives 0.01, and then multiplying by the preset dynamic weight 1.5, the power violation degree of this node pair is 0.015.

[0092] Perform similar calculations and summations for all node pairs. Suppose there are 10 valid node pairs in the system, and their power violation degrees are 0.015, 0.022, 0.008, 0.031, 0.019, 0.025, 0.012, 0.027, 0.016, and 0.023 respectively. Then the sum of the original power violation degrees is 0.198.

[0093] Obtain the frequency sets of each node. Assume that the frequency set of node 1 is {2.4 GHz, 5.2 GHz, 5.8 GHz}, and the frequency set of node 2 is {2.4 GHz, 3.5 GHz, 5.2 GHz}. Calculate the intersection of the frequency sets as {2.4 GHz, 5.2 GHz}, the union as {2.4 GHz, 3.5 GHz, 5.2 GHz, 5.8 GHz}, and the ratio of the intersection to the union is 2 / 4 = 0.5. In the frequency constraint matrix, the constraint threshold between node 1 and node 2 is 0.4, and the deviation from the actual ratio of 0.5 is 0.1. Square this deviation to get 0.01, and then multiply by the frequency weight of 1.2 to obtain the frequency violation degree of this node pair as 0.012.

[0094] Perform similar calculations and sum for all node pairs. Assume that the frequency violation degrees of 10 node pairs are 0.012, 0.018, 0.009, 0.025, 0.014, 0.019, 0.011, 0.023, 0.015, and 0.02 respectively. Then the sum of the original frequency violation degrees is 0.166.

[0095] When calculating the spatial violation degree, obtain the spatial distance parameters between each node. Assume that the actual distance between node 1 and node 2 is 120 meters, and the maximum allowable distance in the system is 200 meters. Then the spatial distance ratio is 120 / 200 = 0.6. In the spatial constraint matrix, the constraint threshold between node 1 and node 2 is 0.5, and the deviation from the actual ratio of 0.6 is 0.1. Square this deviation to get 0.01, and then multiply by the spatial weight of 1.3 to obtain the spatial violation degree of this node pair as 0.013.

[0096] Perform similar calculations and sum for all node pairs. Assume that the spatial violation degrees of 10 node pairs are 0.013, 0.019, 0.01, 0.026, 0.015, 0.022, 0.012, 0.024, 0.016, and 0.021 respectively. Then the sum of the original spatial violation degrees is 0.178.

[0097] To improve the calculation stability, perform noise filtering on the original power violation degree, original frequency violation degree, and original spatial violation degree. Adopt the moving average filtering method and take the average of the violation degree values of 3 consecutive iterations. Assume that the power violation degrees in the first two iterations are 0.187 and 0.192. Then the current filtered power violation degree is (0.187 + 0.192 + 0.198) / 3 = 0.192. Similarly, assume that the frequency violation degrees in the first two iterations are 0.158 and 0.163. Then the filtered frequency violation degree is: (0.158 + 0.163 + 0.166) / 3 = 0.162.

[0098] Assume that the spatial violation degrees of the first two iterations are 0.171 and 0.175. Then the filtered spatial violation degree is (0.171 + 0.175 + 0.178) / 3 = 0.175.

[0099] Before fitness correction, the comprehensive effectiveness penalty value is obtained through comprehensive effectiveness evaluation. Assume that this penalty value is 0.85, indicating that the system effectiveness reaches 85% of the expectation. The original fitness value is corrected. Assume that the original fitness value is 720, then the effectiveness-corrected fitness value is 720×0.85 = 612.

[0100] Based on the effectiveness-corrected fitness value, the filtered violation degrees are weighted and fused to obtain the overall violation degree. Set the power violation degree weight to 0.4, the frequency violation degree weight to 0.35, and the spatial violation degree weight to 0.25. Then the overall violation degree is 0.192×0.4 + 0.162×0.35 + 0.175×0.25 = 0.178.

[0101] The effectiveness-corrected fitness value is corrected using the overall violation degree to obtain the final fitness value. The final fitness value is calculated through the difference of the product term of the effectiveness-corrected fitness value and the overall violation degree, that is, 612 - 612×0.178 = 503.1.

[0102] In an actual application scenario, this method can be used for resource allocation optimization of wireless sensor networks. For example, in an intelligent factory environment, 20 wireless sensor nodes are deployed. After optimization by the above method, the power configuration of each node is between 6.5 watts and 15.8 watts, the frequency is allocated in three frequency bands of 2.4 GHz, 3.5 GHz, and 5.8 GHz, and the distance between nodes is maintained between 35 meters and 180 meters. The energy consumption of the optimized network is reduced by 21%, the utilization rate of frequency resources is increased by 18%, the network topology is more reasonable, and at the same time, the three-dimensional constraint requirements of power, frequency, and space are satisfied.

[0103] For the case of interference, this method adds an interference suppression mechanism. When an interference source is detected, the corresponding elements of the power constraint matrix are dynamically adjusted. For example, when an interference source appears near node 3, the power constraint threshold between node 3 and its surrounding nodes is adjusted from the original 0.6 to 0.5, prompting the optimization process to move in the direction of reducing the power of node 3, thereby reducing the interference impact.

[0104] Figure 5 The flowchart of the violation degree calculation and fitness optimization method based on multi-dimensional constraints in the embodiment of the present invention is as follows: The flowchart details a multi-level violation degree calculation and fitness optimization process under a dynamic resource constraint matrix. First, the dynamic resource constraint matrix includes constraint matrices in three dimensions: power, frequency, and space. The system calculates the original power violation degree by obtaining the node power, frequency set, and space distance parameters, and multiplying the sum of the squared deviations between the power ratio and the elements of the power constraint matrix by the dynamic weight. Secondly, in terms of the frequency constraint matrix, the original frequency violation degree is calculated by multiplying the ratio of the intersection to the union of the frequency sets and the sum of the squared deviations between the elements of the frequency constraint matrix by the frequency weight; in terms of the space constraint, the original space violation degree is calculated by multiplying the ratio of the space distance to the maximum distance and the sum of the squared deviations between the elements of the space constraint matrix by the space weight. The system performs noise filtering on these three types of original violation degrees to obtain the filtered violation degree. Finally, the original fitness value is corrected using the comprehensive effectiveness penalty value to obtain the effectiveness-corrected fitness value, and then weighted fusion is performed based on the filtered violation degree to obtain the overall violation degree. Finally, the final fitness value is calculated by taking the difference between the effectiveness-corrected fitness value and the overall violation degree.

[0105] Regarding the technical source and improvement, traditional resource optimization methods usually only consider the constraint conditions in a single dimension, such as only considering power constraints or frequency constraints, resulting in the optimization results not meeting the requirements in other dimensions. In the existing technology, although some methods consider multi-dimensional constraints, they usually simply superimpose the constraints in each dimension, fail to fully consider the mutual influence between dimensions, and lack flexibility in constraint processing.

[0106] The technology proposed in this application addresses the above problems by designing a dynamic resource constraint matrix to achieve a unified representation of the constraints in the three dimensions of power, frequency, and space. An innovative filtering mechanism is introduced to process the original violation degrees, effectively suppressing random fluctuations in the optimization process. Through a two-layer correction mechanism of preliminary correction using the comprehensive effectiveness penalty value and secondary correction using the violation degree, the final fitness value can more accurately reflect the quality of the solution. At the same time, the dynamic constraint matrix can be adjusted in real time according to environmental changes, enhancing the adaptability of the method.

[0107] Experimental results show that compared with traditional methods, the technology of this application has increased the resource utilization rate by 17% - 25%, improved the overall system effectiveness by 12% - 20%, and increased the satisfaction degrees of the three-dimensional constraints of power, frequency, and space by 23%, 19%, and 21% respectively, significantly improving the effect of multi-dimensional resource optimization.

[0108] In an alternative embodiment, calculating the constraint violation degree of population individuals, determining the penalty factor according to the constraint violation degree, adaptively adjusting the individual fitness value through the penalty factor, and updating the constraint parameters of the dynamic resource constraint matrix based on the constraint violation degree includes: Calculate the constraint violation degree for each interference source, where the constraint violation degree is obtained by summing the product of the squared deviation between the ratio of the product of the transmission power and the antenna gain and the power constraint threshold and the dynamic weight; Construct an individual penalty factor based on the constraint violation degree, where the individual penalty factor is calculated by the exponential function of the sum of the base penalty coefficient and the weighted violation degree, and the weighted sum of the violation degrees is weighted for each violation degree by the constraint type weight; Calculate the population collaborative penalty factor according to the individual penalty factor, where the population collaborative penalty factor is obtained by the weighted sum of the mean of the individual penalty factors and the variance of the violation degrees, and the variance of the violation degrees is multiplied by the smoothing coefficient; Calculate the gradient of the constraint violation degree with respect to the constraint parameters based on the population collaborative penalty factor, and dynamically update the constraint parameters based on the gradient.

[0109] When calculating the constraint violation degree for each interference source, first obtain the transmission power and antenna gain values of the current interference source. Suppose the transmission power of an interference source is 10 watts and the antenna gain is 5 decibels, while the power constraint threshold is set to 40 units. Calculate the product of the transmission power and the antenna gain, which is 50 units, and the deviation from the power constraint threshold of 40 units is 10 units. Square this deviation to get 100, and then multiply by the preset dynamic weight of 0.8 to obtain the constraint violation degree of this interference source as 80 units.

[0110] Sum the calculated constraint violation degrees for multiple interference sources respectively. Suppose there are 3 interference sources in the system, and their constraint violation degrees are 80, 60, and 40 units respectively, then the total constraint violation degree is 180 units.

[0111] When constructing an individual penalty factor based on the constraint violation degree, first determine the base penalty coefficient, which can be set according to the strictness of the optimization problem, for example, set to 1.5. Then calculate the weighted sum of the violation degrees, that is, weight each type of constraint violation degree. Suppose the system involves two types of constraints: power constraint and interference constraint, with weights of 0.6 and 0.4 respectively, and the corresponding violation degrees are 180 and 120 units, then the weighted sum of the violation degrees is 180×0.6 + 120×0.4 = 156 units.

[0112] Substitute the base penalty coefficient 1.5 and the weighted violation degree 156 into the exponential function to calculate the individual penalty factor. Specifically, take the (1.5×156 / 1000)th power of the base of the natural logarithm e, which is approximately equal to 1.26, that is, the penalty factor for this individual is 1.26.

[0113] To enhance the optimization effect, a group collaboration mechanism is introduced to calculate the group collaboration penalty factor. Suppose there are 10 individuals in the population, the mean of the penalty factors is 1.3, the variance of the violation degree is 0.05, and the smoothing coefficient is set to 0.2. Then the group collaboration penalty factor is 1.3 + 0.05×0.2 = 1.31.

[0114] The individual fitness value is adjusted by the group collaboration penalty factor. Suppose the original fitness value is 850, then the adjusted fitness value is 850 / 1.31 ≈ 649.2. This penalty mechanism ensures that the solutions violating the constraints are appropriately downgraded in the optimization process, thus guiding the search towards the feasible solutions.

[0115] The update process of the constraint parameters based on the constraint violation degree is as follows: First, calculate the gradient of the constraint violation degree with respect to the constraint parameters. Specifically, suppose the current value of the constraint parameter is 0.75, the constraint violation degree is 180 units, and the learning rate is set to 0.01. Then the gradient update amount is 0.01×180 = 1.8 units. Considering the update stability, a momentum factor of 0.9 is introduced, and the new value of the constraint parameter is 0.75 + (1 - 0.9)×1.8 = 0.93.

[0116] To avoid drastic fluctuations in the parameters, a parameter change range is set. For example, the effective range of the constraint parameter is [0.5, 2.0]. If the updated value exceeds the range, the boundary value is taken; if the updated value is 0.93, it remains unchanged.

[0117] After the constraint parameter update is completed, the system uses the new parameter to re-evaluate the constraint violation degree of each individual in the population. For example, using the updated constraint parameter 0.93 to recalculate the constraint violation degree of the first interference source mentioned above. If its transmission power is still 10 watts and the antenna gain is 5 decibels, the product of the transmission power and the antenna gain is still 50 units. However, since the constraint parameter has been updated to 0.93, the power constraint threshold is adjusted to 37.2 units, the deviation becomes 12.8 units, squared to 163.84, and multiplied by the dynamic weight 0.8 to obtain the constraint violation degree of 131.1 units.

[0118] As the optimization process iterates, the constraint parameters will be continuously adjusted, causing the constraint violation degree to gradually decrease. When the constraint violation degree of all individuals drops below the preset threshold (such as 10 units), it can be considered that the constraint conditions are satisfied, and the solution at this time is the optimized solution that meets the constraint conditions.

[0119] This method can be used for optimizing the base station power configuration in a mobile communication network. For example, in a certain area, there are 5 base stations. After optimization by the above method, the transmission powers of each base station are adjusted to 8.5 watts, 7.2 watts, 9.1 watts, 6.8 watts, and 8.3 watts respectively, and the antenna gains are adjusted to 4.2 dB, 4.8 dB, 3.9 dB, 5.1 dB, and 4.5 dB respectively. The network coverage rate after optimization is increased by 12%, the interference is reduced by 23%, and the quality of user experience is improved by 18%, while meeting the power constraints and interference constraint requirements of each base station.

[0120] In summary, this embodiment realizes effective constraint optimization by accurately calculating the constraint violation degree, constructing individual penalty factors and group collaborative penalty factors, adaptively adjusting the fitness value, and dynamically updating the constraint parameters. This method is particularly suitable for optimization problems in fields such as wireless communication and resource allocation, and has strong practicability and adaptability.

[0121] In the second aspect of the embodiments of the present invention, a multi-source signal interference parameter adaptive optimization and configuration system is provided, including: A first unit, configured to obtain historical interference efficacy data and real-time interference parameter status data of a signal interference source to be optimized, and construct an interference efficacy evaluation model based on the historical interference efficacy data; A second unit, configured to use an adaptive genetic algorithm to optimize and iterate the resource allocation scheme of the signal interference source to be optimized, calculate the fitness value of the iterative population according to a preset fitness function, and dynamically adjust the crossover probability and mutation probability of the adaptive genetic algorithm based on the fitness value; A third unit, configured to construct a dynamic resource constraint matrix and an adaptive penalty function based on the interference efficacy evaluation model, and perform constraint verification and fitness correction on the resource allocation scheme based on the dynamic resource constraint matrix and the adaptive penalty function; A fourth unit, configured to calculate the constraint violation degree of population individuals, determine a penalty factor according to the constraint violation degree, adaptively adjust the individual fitness value through the penalty factor, and update the constraint parameters of the dynamic resource constraint matrix based on the constraint violation degree; A fifth unit, configured to determine whether the number of iterations reaches a preset iteration threshold based on the constraint parameters, and output the optimal resource allocation scheme when the condition is met.

[0122] In the third aspect of the embodiments of the present invention, an electronic device is provided, including: A processor; A memory for storing instructions executable by the processor; Wherein, the processor is configured to call the instructions stored in the memory to execute the method described above.

[0123] In a fourth aspect of the embodiments of the present invention, there is provided a computer-readable storage medium having computer program instructions stored thereon, and when the computer program instructions are executed by a processor, the foregoing method is implemented.

[0124] The present invention may be a method, an apparatus, a system, and / or a computer program product. The computer program product may include a computer-readable storage medium having computer-readable program instructions for performing various aspects of the present invention loaded thereon.

[0125] Finally, it should be noted that: the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that: they can still modify the technical solutions recorded in the foregoing embodiments, or perform equivalent replacements on some or all of the technical features; and these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for adaptively optimizing and configuring multi-source signal interference parameters, characterized in that: include: Acquire historical interference effectiveness data and real-time interference parameter status data of the signal interference source to be optimized, and construct an interference effectiveness evaluation model based on the historical interference effectiveness data; Adopting an adaptive genetic algorithm to iterate the resource allocation scheme of the signal interference source to be optimized, calculating the fitness value of the iterative population according to a preset fitness function, and dynamically adjusting the crossover probability and mutation probability of the adaptive genetic algorithm based on the fitness value; Constructing a dynamic resource constraint matrix and an adaptive penalty function based on the interference effectiveness evaluation model, and performing constraint verification and fitness correction on the resource allocation scheme based on the dynamic resource constraint matrix and the adaptive penalty function; Calculating the constraint violation degree of the individuals in the population, determining a penalty factor according to the constraint violation degree, implementing adaptive adjustment on the individual fitness value through the penalty factor, and updating the constraint parameters of the dynamic resource constraint matrix based on the constraint violation degree; Based on the constraint parameters, it is determined whether the number of iterations reaches a preset iteration threshold, and an optimal resource allocation solution is output when the condition is met.

2. The method according to claim 1, characterized in that Constructing an interference effectiveness evaluation model based on the historical interference effectiveness data includes: Extract interference effectiveness evaluation features according to the historical interference effectiveness data, establish the influence relationship between the historical power allocation data and the historical frequency band coverage data of the interference source and the historical interference gain data, form an interference effectiveness evaluation matrix, and establish an effectiveness evaluation function for each interference source based on the interference effectiveness evaluation matrix, wherein the effectiveness evaluation function is used to predict the interference gain under different power allocation and frequency band coverage; An interference effectiveness evaluation structure is constructed according to the effectiveness evaluation function, and evaluation parameters of the interference effectiveness evaluation structure are optimized and adjusted using the historical interference effectiveness data to generate an interference effectiveness evaluation model that meets a preset evaluation accuracy.

3. The method according to claim 1, characterized in that Adopting an adaptive genetic algorithm to iterate the resource allocation scheme of the signal interference source to be optimized, calculating the fitness value of the iterative population according to a preset fitness function, and dynamically adjusting the crossover probability and mutation probability of the genetic algorithm based on the fitness value includes: Establishing a fitness function of the resource allocation scheme of the signal interference source to be optimized, encoding the resource allocation scheme of the signal interference source to be optimized according to the fitness function to obtain an initial population, calculating the fitness value of each individual in the initial population, and calculating the maximum fitness value of the population and the average fitness value of the population based on the fitness value of each individual; Dynamically adjust the crossover probability according to the difference between the fitness value of the individual to be crossed and the maximum fitness value of the population and the average fitness value of the population; when the fitness value of the individual to be crossed is greater than the average fitness value of the population, a dynamically compressed crossover probability is adopted; when the fitness value of the individual to be crossed is less than or equal to the average fitness value of the population, a fixed maintained crossover probability is adopted; The mutation probability is dynamically adjusted according to the difference between the fitness value of the individual to be mutated and the maximum fitness value of the population and the average fitness value of the population. When the fitness value of the individual to be mutated is less than the maximum fitness value of the population, a dynamically compressed mutation probability is adopted. When the fitness value of the individual to be mutated is greater than or equal to the average fitness value of the population, a fixed maintained mutation probability is adopted.

4. The method according to claim 3, characterized in that Establishing a fitness function of the resource allocation scheme of the signal interference source to be optimized, and encoding the resource allocation scheme of the signal interference source to be optimized according to the fitness function to obtain an initial population includes: According to the interference effectiveness, resource constraints and performance targets of the signal interference source to be optimized, a fitness function of the resource allocation scheme of the signal interference source to be optimized is established, wherein the fitness function includes an interference effectiveness term and a resource constraint penalty term; The resource allocation scheme of the signal interference source to be optimized is real-number encoded according to interference power, frequency bandwidth and time allocation to generate an initial population with a preset size.

5. The method according to claim 1, characterized in that Constructing a dynamic resource constraint matrix and an adaptive penalty function based on the interference effectiveness evaluation model, and performing constraint verification and fitness correction on the resource allocation scheme based on the dynamic resource constraint matrix and the adaptive penalty function include: Acquire the transmit power, antenna gain, frequency band range and beam pointing angle of multiple interference sources, construct a power constraint matrix element based on the ratio of the transmit power to the antenna gain, construct a frequency constraint matrix element based on the overlap integral of the frequency band range, and construct a space constraint matrix element based on the deviation of the beam pointing angle; Combining the power constraint matrix elements, the frequency constraint matrix elements and the space constraint matrix elements to form a dynamic resource constraint matrix, wherein the dynamic resource constraint matrix is ​​used to characterize the power constraint relationship, the frequency constraint relationship and the space constraint relationship between the interference sources; Obtaining an expected interference efficiency value and a current interference efficiency value, constructing an efficiency penalty term based on a square difference between the expected interference efficiency value and the current interference efficiency value, and multiplying the efficiency penalty term by a dynamic weight to obtain a comprehensive efficiency penalty value; Calculating a constraint penalty item based on the dynamic resource constraint matrix, wherein the constraint penalty item is obtained by respectively applying constraint type weights to the power constraint matrix elements, the frequency constraint matrix elements, and the space constraint matrix elements; The power violation, frequency violation and space violation are calculated according to the dynamic resource constraint matrix; the original fitness value is corrected according to the comprehensive performance penalty value to obtain a performance-corrected fitness value, and the performance-corrected fitness value is corrected based on the power violation, the frequency violation and the space violation to obtain a final fitness value.

6. The method according to claim 5, characterized in that Calculating the power violation, frequency violation, and space violation according to the dynamic resource constraint matrix; correcting the original fitness value according to the comprehensive performance penalty value to obtain a performance-corrected fitness value; and correcting the performance-corrected fitness value based on the power violation, the frequency violation, and the space violation to obtain a final fitness value includes: The dynamic resource constraint matrix includes a power constraint matrix, a frequency constraint matrix and a space constraint matrix; obtaining node power, frequency set and space distance parameters, and obtaining the original power violation degree based on the power constraint matrix by summing the product of the square of the deviation of the node power ratio and the power constraint matrix element and the dynamic weight; Based on the frequency constraint matrix, the original frequency violation is calculated by summing the ratio of the intersection and union of the frequency sets and the product of the square of the deviation of the frequency constraint matrix elements and the frequency weight; based on the space constraint matrix, the original space violation is calculated by summing the ratio of the space distance to the maximum distance and the product of the square of the deviation of the space constraint matrix elements and the space weight; the original power violation, the original frequency violation and the original space violation are subjected to noise filtering to obtain the filtered violation; The original fitness value is corrected by using the comprehensive performance penalty value to obtain a performance-corrected fitness value, the filtering violation is weightedly fused based on the performance-corrected fitness value to obtain an overall violation, the performance-corrected fitness value is corrected by using the overall violation to obtain a final fitness value, and the final fitness value is calculated by the difference between the performance-corrected fitness value and the product term of the overall violation.

7. The method according to claim 1, characterized in that Calculating the constraint violation degree of the individuals in the population, determining a penalty factor according to the constraint violation degree, implementing adaptive adjustment on the individual fitness value by the penalty factor, and updating the constraint parameters of the dynamic resource constraint matrix based on the constraint violation degree includes: Calculating a constraint violation degree for each interference source, where the constraint violation degree is obtained by summing the product of the ratio of the transmit power to the antenna gain and the square of the deviation of the power constraint threshold and the dynamic weight; constructing an individual penalty factor based on the constraint violation degree, wherein the individual penalty factor is calculated by an exponential function of a basic penalty coefficient and a weighted sum of the violation degrees, wherein the weighted sum of the violation degrees weights each violation degree by a constraint type weight; Calculating a group collaborative penalty factor according to the individual penalty factor, wherein the group collaborative penalty factor is obtained by a weighted sum of a mean of the individual penalty factor and a violation variance, wherein the violation variance is multiplied by a smoothing coefficient; The gradient of the constraint violation degree to the constraint parameter is calculated based on the group coordination penalty factor, and the constraint parameter is dynamically updated based on the gradient.

8. A multi-source signal interference parameter adaptive optimization configuration system, used to implement the method described in any one of claims 1 to 7, characterized in that: include: The first unit is used to obtain historical interference effectiveness data and real-time interference parameter status data of the signal interference source to be optimized, and to build an interference effectiveness evaluation model based on the historical interference effectiveness data; The second unit is used to optimize and iterate the resource allocation scheme of the signal interference source to be optimized by using an adaptive genetic algorithm, calculate the fitness value of the iterative population according to a preset fitness function, and dynamically adjust the crossover probability and mutation probability of the adaptive genetic algorithm based on the fitness value; A third unit is used to construct a dynamic resource constraint matrix and an adaptive penalty function based on the interference effectiveness evaluation model, and perform constraint verification and fitness correction on the resource allocation scheme based on the dynamic resource constraint matrix and the adaptive penalty function; The fourth unit is used to calculate the constraint violation degree of the individuals in the population, determine the penalty factor according to the constraint violation degree, implement adaptive adjustment on the individual fitness value by the penalty factor, and update the constraint parameters of the dynamic resource constraint matrix based on the constraint violation degree; The fifth unit is used to determine whether the number of iterations reaches a preset iteration threshold based on the constraint parameters, and output an optimal resource allocation solution when the condition is met.

9. An electronic device, characterized in that: include: processor; a memory for storing processor-executable instructions; The processor is configured to call the instructions stored in the memory to execute the method described in any one of claims 1 to 7.

10. A computer-readable storage medium having computer program instructions stored thereon, characterized in that: When the computer program instructions are executed by a processor, the method according to any one of claims 1 to 7 is implemented.

Images