AI-based low-altitude economic communication server security resource allocation method

Abstract

The invention belongs to the technical field of network communication, and discloses an AI-based low-altitude economic communication server security resource allocation method. The method comprises the following steps: receiving m groups of device feature data, wherein m is an integer greater than 1; analyzing the m groups of equipment feature data, and identifying corresponding job tasks; analyzing the m groups of equipment characteristic data to obtain corresponding communication characteristic data; according to the corresponding job tasks and the communication characteristic data, performing priority division on the m groups of equipment characteristic data; dynamically allocating communication resource data based on the priorities and communication characteristic data corresponding to the m groups of equipment characteristic data, and performing data transmission; monitoring a data transmission process in real time, and judging whether an alarm instruction is generated or not; according to the invention, the low-altitude economic communication server can flexibly adapt to the dynamic change of the network load, quickly respond to the dynamic allocation of the resources, and improve the generalization ability of the application scene, thereby effectively improving the communication efficiency and security.

Description

Technical Field

[0001] The present invention relates to the field of network communication technology, and more specifically, to an AI-based low-altitude economic communication server security resource allocation method. Background Art

[0002] With the rapid development of drone technology and the low-altitude economy, the commercial application of low-altitude airspace is becoming increasingly widespread, such as drone delivery, agricultural monitoring, environmental monitoring and other scenarios; it not only improves logistics and service efficiency, but also provides new impetus for economic growth and social development; however, this is accompanied by higher requirements for data communication and security; low-altitude economic communication servers, as the core link of information transmission, carry a large amount of real-time data streams and instructions; in this environment, servers must have efficient resource allocation capabilities to cope with rapidly changing network needs; at the same time, security has also become a top priority, especially when facing threats such as network attacks and data leaks, how to ensure the security and reliability of communications is a key issue.

[0003] Patent publication number CN115086992A discloses a distributed semantic communication system and a bandwidth resource allocation method and apparatus. The method includes: deploying the fully connected layer of a deep residual network on an edge server, and deploying the neural network before the fully connected layer in the deep residual network on D terminal devices; establishing a bandwidth resource allocation optimization model based on the distributed semantic communication system; constructing and training a deep deterministic policy gradient (DDPG) algorithm model, and determining the optimal resource allocation strategy network in the offline phase based on the bandwidth resource allocation optimization model; when in the online phase, inputting current state information into the optimal resource allocation strategy network to determine the optimal feature compression scheme and bandwidth resource allocation scheme; thereby alleviating the problem of scarce available bandwidth resources in current wireless communication networks.

[0004] However, the DDPG algorithm used in the above technology requires the total amount of bandwidth resources for the entire network to be determined at the beginning of resource allocation. It cannot adapt to changes in network load and lacks flexibility. In addition, the training process of the DDPG algorithm is computationally complex and requires a large amount of data training. It is difficult to quickly obtain the optimal solution in actual networks, which affects real-time responsiveness and reduces resource allocation efficiency and communication performance. In addition, the DDPG algorithm can only be optimized for a specific scenario and is difficult to generalize to other different network environments.

[0005] In view of this, the present invention proposes an AI-based low-altitude economic communication server security resource allocation method to solve the above problems. Summary of the Invention

[0006] In order to overcome the above-mentioned defects of the prior art and achieve the above-mentioned objectives, the present invention provides the following technical solutions: a method for allocating safe resources for low-altitude economic communication servers based on AI, the method comprising: S1: Receive m sets of device feature data, where m is an integer greater than 1; S2: Analyze the characteristic data of m groups of equipment and identify the corresponding work tasks; S3: Analyze the characteristic data of the m groups of devices to obtain corresponding communication characteristic data; S4: Prioritize the m groups of device feature data based on the corresponding job tasks and communication feature data; S5: Dynamically allocate communication resource data based on the priorities and communication characteristic data corresponding to the m groups of device feature data, and perform data transmission; S6: Monitor the data transmission process in real time and determine whether to generate an alarm instruction.

[0007] Furthermore, the device characteristic data includes movement speed, data transmission volume and data transmission efficiency; the device characteristic data corresponds to the low-altitude device one-to-one; the movement speed is the movement speed of the low-altitude device relative to the ground; the data transmission volume is the total amount of data that the low-altitude device needs to transmit to the cloud platform per unit time; the data transmission efficiency is the total amount of data successfully transmitted to the cloud platform by the low-altitude device per unit time; The step of identifying the corresponding job task includes: Step S201: establishing a fuzzy set, dividing each data of the device feature data into multiple fuzzy sets; Step S202: converting each data in the m groups of device feature data into the membership degree of the corresponding fuzzy set by using fuzzification technology; Step S203: defining fuzzy rules; Step S204: matching the fuzzified m groups of equipment feature data with fuzzy rules, performing fuzzy reasoning, and obtaining a fuzzy reasoning result, which is the membership degree of the m groups of equipment feature data corresponding to different work tasks; Step S205: Defuzzify the fuzzy reasoning result to obtain the defuzzified result, which is the job task corresponding to the m groups of equipment feature data; the defuzzification operation is: according to the fuzzy reasoning result, obtain the membership corresponding to each job task of each group of equipment feature data respectively; compare the membership corresponding to each group of equipment feature data, and take the job task corresponding to the membership with the largest value as the job task corresponding to the corresponding equipment feature data, and identify the job task corresponding to the m groups of equipment feature data.

[0008] Furthermore, the method for obtaining corresponding communication characteristic data includes: Communication characteristic data includes bandwidth requirements and power requirements; The bandwidth requirement is expressed as: ; Where, For bandwidth requirements, is the data transmission volume, For data transmission efficiency, is the moving speed, 、 All are preset weight coefficients; The expression for power demand is: ; Where, For power demand, 、 is the preset scale factor.

[0009] Furthermore, the method for prioritizing m groups of device feature data includes: Different digital labels are set for different job tasks and marked as task labels; the task labels and communication characteristic data are used as analysis data; the analysis data corresponding to each set of device feature data are input into the trained priority prediction model respectively to predict the priority coefficient corresponding to each set of device feature data; the priority coefficients corresponding to each set of device feature data are sorted from large to small to obtain a priority sorting table; according to the positive order of the priority sorting table, the corresponding priority numbers are set for the m sets of device feature data in ascending order, and the range of priority numbers is , where 1 is the highest priority and m is the lowest priority; The training process of the priority prediction model includes: Collect q groups of analysis data in advance, set corresponding priority coefficients for each of the q groups of analysis data, where q is an integer greater than 1, and convert the analysis data and the corresponding priority coefficients into a corresponding set of feature vectors; use each set of feature vectors as input to a priority prediction model, and the priority prediction model uses a set of prediction priority coefficients corresponding to each group of analysis data as output, and uses the actual priority coefficients corresponding to each group of analysis data as a prediction target, where the actual priority coefficients are the pre-set priority coefficients corresponding to the analysis data; minimize the sum of the prediction errors of all analysis data as a training target; train the priority prediction model until the sum of the prediction errors converges and the training is stopped; the priority prediction model is a deep neural network model.

[0010] Furthermore, the communication resource data includes bandwidth and power; The step of dynamically allocating communication resource data comprises: Step S501: Define population size , iterative threshold , mobile factor and disturbance factor ; Step S502: construct M sets of data sets, , set different numerical labels for different data sets and mark them as set labels; Step S503: Constructing a population , the position of each individual in the population corresponds to the set label one by one, and individuals include hunters and prey; the number of iterations t corresponding to the population is 0; Step S504: determining an evaluation value function; Step S505: Randomly generate prey, calculate the selection probability of each prey and make a selection; Step S506: determining the global optimal position and calculating the trap depth corresponding to each hunter; Step S507: Calculate the attractiveness corresponding to each hunter; Step S508: updating the position of each hunter and prey; Step S509: Determine whether the number of iterations t is less than the iteration threshold. If so, set , and return to step S505, if not, go to step S510; Step S510: Update the global optimal position, obtain the set tag corresponding to the global optimal position, and dynamically allocate communication resource data according to the data set corresponding to the set tag.

[0011] Furthermore, in step S502, the method for constructing M groups of data sets is as follows: obtaining a total data value, the total data value including the total network bandwidth and the total network power; randomly dividing the total network bandwidth into m bandwidth components, the sum of the m bandwidth components being the total network bandwidth; randomly dividing the total network power into m power components, the sum of the m power components being the total network power; taking the m bandwidth components and the m power components as a group of data sets, wherein one bandwidth component and one power component correspond to one low-altitude device; and so on and so forth, constructing a total of M groups of data sets, each group of data sets being different; In step S503, the population , For the individuals; the position of each individual in the population is defined in a one-dimensional search space, the range of the one-dimensional search space is M sets of data; the population The expression for each individual position in is: Where, is the position of the i-th individual, is the random coefficient of the i-th individual, , ; Randomly divide the population in Individuals as hunters, Individuals as prey, , .

[0012] Furthermore, in step S504, the expression of the evaluation value function is: ; Where, is the evaluation value, For the The priority number corresponding to the group device characteristic data, For the Adjust the transmission efficiency corresponding to the group device characteristic data, For the The adjusted transmission volume corresponding to the group device characteristic data, ; Adjust the transmission efficiency to the data transmission efficiency after the dynamic allocation of communication resource data, and adjust the transmission volume to the data transmission volume after the dynamic allocation of communication resource data; The method for obtaining the adjusted transmission volume and adjusted transmission efficiency is as follows: the data transmission volume, data transmission efficiency, priority number and communication characteristic data are used as research data, and the set labels corresponding to the research data and individual positions are used as evaluation data; the evaluation data is input into the trained adjustment prediction model to predict the corresponding adjustment label; the adjustment label is the digital label corresponding to the adjustment data, and different adjustment data have different corresponding digital labels. The adjustment data includes the adjusted transmission volume and the adjusted transmission efficiency; the training process of the adjustment prediction model is consistent with the training process of the priority prediction model, and both are deep neural network models.

[0013] Furthermore, in step S505, ,The method of calculating the selection probability of each prey and making the selection includes: Calculating populations The evaluation value of each prey in the , and add them up in turn to obtain the total evaluation value, divide the evaluation value of each prey by the total evaluation value to obtain the selection probability of each prey; sort the selection probability of each prey from large to small, and replace each selection probability in positive order with the sum of each selection probability and all selection probabilities before the corresponding selection probability; obtain all selection probabilities after replacement and mark them as replacement probabilities; the prey corresponding to the replacement probability is consistent with the prey corresponding to the selection probability before replacement; take every two adjacent replacement probabilities as an analysis set in positive order; mark the replacement probability with a large value in each analysis set as the first probability, and mark the replacement probability with a small value as the second probability; construct the probability screening range corresponding to each analysis set according to the first probability and the second probability of each analysis set, where the first probability is the maximum value of the probability screening range, the second probability is the minimum value of the probability screening range, and the minimum value is an open interval, and the maximum value is a closed interval; each analysis set corresponds to the prey corresponding to the first probability; where the minimum value of the probability screening range corresponding to the first prey is 0, and the maximum value is the corresponding selection probability; randomly generate indivual A random number between The probability of the random number screening range corresponds to the prey.

[0014] Furthermore, in step S506, the method for determining the global optimal position is: calculating the evaluation value corresponding to each individual, and taking the position of the individual corresponding to the maximum evaluation value as the global optimal position; The expression for the trap depth is: Where, For the The trap depth corresponding to the hunter, is the evaluation value corresponding to the global optimal position, For the The evaluation value corresponding to each hunter is ; In step S507, the expression of attractiveness is: Where, For the The attractiveness corresponding to each hunter, For the The trap depth corresponding to the hunter, ; In step S508, the method for updating the hunter's location includes: ; Where, After the update The position of a hunter, For the The position of a hunter, is the global optimal position, for Random number between Methods for updating prey locations include: ; Where, After the update The location of the prey, For the The location of the prey, , for Random number between In step S510, the method for updating the global optimal position is: calculate the evaluation value corresponding to each individual after the position is updated, and mark it as the first evaluation value; use the evaluation value corresponding to the current global optimal position as the second evaluation value; compare the first evaluation value with the second evaluation value, if the first evaluation value is greater than the second evaluation value, then update the global optimal position to the individual position corresponding to the first evaluation value; if the first evaluation value is less than or equal to the second evaluation value, then do not update the global optimal position.

[0015] Furthermore, the method for determining whether to generate an alarm instruction includes: During the data transmission process, the data transmission volume corresponding to each low-altitude device is collected in real time and marked as the real-time transmission volume. The transmission volume difference is calculated based on the real-time transmission volume and the adjusted transmission volume. The expression of the transmission volume difference is: Where, is the transmission difference, is the real-time transmission volume, To adjust the transmission volume; preset a difference threshold, compare the difference threshold with the transmission volume difference; if the transmission volume difference is greater than or equal to the difference threshold, generate an alarm instruction; if the transmission volume difference is less than the difference threshold, no alarm instruction is generated.

[0016] The technical effects and advantages of the AI-based low-altitude economic communication server security resource allocation method of the present invention are as follows: By comprehensively and dynamically sensing equipment characteristics, identifying work tasks, predicting priorities, intelligently allocating resources, and monitoring safety in real time, intelligent management and safety control of low-altitude communication systems are achieved; low-altitude economic communication servers can flexibly adapt to dynamic changes in network load, quickly respond to dynamic allocation of resources, and improve the generalization ability of application scenarios; thereby effectively improving communication efficiency and security, ensuring the reliability and flexibility of real-time data transmission, reducing the risk of network attacks and data leakage, and effectively responding to the growing communication needs in the low-altitude economic field. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] Figure 1 This is a flow chart of a method for allocating safe resources for a low-altitude economic communication server based on AI according to Example 1 of the present invention; Figure 2 This is a flow chart of the communication resource data dynamic allocation method according to embodiment 1 of the present invention. DETAILED DESCRIPTION

[0018] The following will be combined The drawings in the embodiments of the present invention clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the embodiments described are only part of the embodiments of the present invention, not all of the embodiments. All other embodiments derived by persons of ordinary skill in the art based on the embodiments of the present invention without inventive effort shall fall within the scope of protection of the present invention.

[0019] Example 1 See also Figure 1 As shown, this embodiment provides an AI-based method for allocating safe resources for low-altitude economic communication servers, the method comprising: S1: Receive m sets of device characteristic data, where m is an integer greater than 1.

[0020] Equipment characteristic data includes moving speed, data transmission volume and data transmission efficiency; equipment characteristic data corresponds one-to-one with low-altitude equipment; The moving speed is the speed of the low-altitude device relative to the ground, and is obtained by the device's built-in speed sensor. High-speed low-altitude devices can produce the Doppler effect, causing frequency offset, which reduces the signal quality of data transmission and increases channel attenuation and path loss, thus affecting the communication link evaluation value. Therefore, it is necessary to increase bandwidth to compensate for frequency offset and jitter, and increase power to enhance signal strength and reduce signal transmission path loss. Data transmission volume refers to the total amount of data that low-altitude equipment needs to transmit to the cloud platform per unit time. The cloud platform is a service platform based on cloud computing technology, which provides data transmission, storage, and analysis functions for low-altitude equipment. Data transmission volume is obtained through statistics of the protocol stack within the low-altitude equipment (such as TCP / IP, UDP, MQTT, etc.). The greater the data transmission volume, the greater the corresponding bandwidth and power requirements of the low-altitude equipment. Data transmission efficiency is the total amount of data successfully transmitted from low-altitude devices to the cloud platform per unit time and is an important indicator for measuring the quality of communication links. Data transmission efficiency is obtained through network performance monitoring tools within the cloud platform (such as NetFlow and Wireshark). Low data transmission efficiency indicates a high packet loss rate, and a larger amount of data needs to be retransmitted, necessitating increased bandwidth and power requirements to reduce link congestion, increase transmission rates, minimize interference and signal attenuation, and reduce packet loss rates.

[0021] S2: Analyze the m groups of equipment feature data and identify the corresponding work tasks.

[0022] Task operations such as logistics distribution, agricultural monitoring, environmental testing, map drawing, etc. The steps for identifying the corresponding job tasks include: Step S201: Establish fuzzy sets, and divide each data of the device characteristic data into multiple fuzzy sets; for example, the fuzzy sets corresponding to the moving speed are slow, medium, and fast, etc.; the fuzzy sets corresponding to the data transmission volume are small, medium, and large, etc.; and the fuzzy sets corresponding to the data transmission efficiency are low, medium, and high efficiency, etc.; Step S202: Each data in the m groups of device feature data is converted into a membership of a corresponding fuzzy set using fuzzification technology. Fuzzification is the process of converting an exact numerical value into a membership corresponding to a fuzzy set. Fuzzification technologies include triangular membership function and trapezoidal membership function. For example, if the value of the moving speed is low, it is inferred that the membership of the slow speed is 0.8, the membership of the medium speed is 0.1, and the membership of the fast speed is 0.1. Step S203: Define fuzzy rules. The fuzzy rules are defined based on expert knowledge or literature. For example, if the speed is slow, the transmission volume is small, and the efficiency is high, then the task is inferred to be environmental monitoring; if the speed is medium, the transmission volume is medium, and the efficiency is medium, then the task is inferred to be map drawing. Step S204: Match the fuzzified m groups of equipment feature data with fuzzy rules, perform fuzzy reasoning, and obtain fuzzy reasoning results. The fuzzy reasoning results are the membership degrees of the m groups of equipment feature data corresponding to different work tasks. The fuzzy reasoning method may be, for example, the Mamdani or Sugeno fuzzy reasoning method. For example, the membership degree of the work task for environmental monitoring is 0.2, and the membership degree of the work task for map drawing is 0.8. Step S205: Defuzzify the fuzzy reasoning result to obtain the defuzzified result, which is the job task corresponding to the m groups of equipment feature data; defuzzification is to convert the fuzzy reasoning result into a specific recognition result, and the defuzzification operation is: according to the fuzzy reasoning result, obtain the membership degree corresponding to each job task of each group of equipment feature data respectively; compare the membership degree corresponding to each group of equipment feature data, and take the job task corresponding to the membership degree with the largest value as the job task corresponding to the corresponding equipment feature data, and identify the job task corresponding to the m groups of equipment feature data.

[0023] S3: Analyze the m groups of device characteristic data to obtain corresponding communication characteristic data.

[0024] The method for obtaining corresponding communication characteristic data includes: Communication characteristic data includes bandwidth requirements and power requirements; The bandwidth requirement is expressed as: ; Where, For bandwidth requirements, is the data transmission volume, For data transmission efficiency, is the moving speed, 、 All are preset weight coefficients; The specific value of the weight coefficient in the formula can be set according to actual conditions. The weight coefficient reflects the impact of communication characteristic data on bandwidth requirements. Technical personnel in this field can preset the corresponding weight coefficient based on the actual impact of communication characteristic data on bandwidth requirements, so as to accurately evaluate the bandwidth requirements of low-altitude equipment during data transmission.

[0025] The expression for power demand is: ; Where, For power demand, 、 is the preset proportional coefficient; The specific value of the proportional coefficient in the formula can be set according to actual conditions. The proportional coefficient reflects the impact of communication characteristic data on power requirements. Technical personnel in this field can preset the corresponding proportional coefficient based on the actual impact of communication characteristic data on power requirements, so as to accurately evaluate the power requirements of low-altitude equipment during data transmission.

[0026] S4: Prioritize the m groups of device feature data according to the corresponding job tasks and communication feature data.

[0027] The method for prioritizing m groups of device feature data includes: Different digital labels are set for different job tasks and marked as task labels; the task labels and communication characteristic data are used as analysis data; the analysis data corresponding to each set of device feature data are input into the trained priority prediction model respectively to predict the priority coefficient corresponding to each set of device feature data; the priority coefficients corresponding to each set of device feature data are sorted from large to small to obtain a priority sorting table; according to the positive order of the priority sorting table, the corresponding priority numbers are set for the m sets of device feature data in ascending order, and the range of priority numbers is , where 1 is the highest priority and m is the lowest priority.

[0028] The training process of the priority prediction model includes: Collect q groups of analysis data in advance, set corresponding priority coefficients for each of the q groups of analysis data, where q is an integer greater than 1, and convert the analysis data and the corresponding priority coefficients into a corresponding set of feature vectors; the priority coefficients corresponding to the analysis data are collected by those skilled in the art during the historical priority division process, and each group of analysis data is analyzed in turn based on actual experience to evaluate the corresponding priority coefficients, and the corresponding priority coefficients are set for the q groups of analysis data in turn; Each set of feature vectors is used as the input of the priority prediction model. The priority prediction model takes a set of prediction priority coefficients corresponding to each set of analysis data as output, and the actual priority coefficients corresponding to each set of analysis data as the prediction target. The actual priority coefficients are the pre-set priority coefficients corresponding to the analysis data. The training goal is to minimize the sum of the prediction errors of all analysis data. The calculation formula of the prediction error is: ,in is the prediction error, w is the group number of the eigenvector corresponding to the analysis data, is the prediction priority coefficient corresponding to the wth group of analysis data, is the actual priority coefficient corresponding to the wth group of analysis data; the priority prediction model is trained until the sum of the prediction errors reaches convergence and the training is stopped.

[0029] The above-mentioned priority prediction model is specifically a deep neural network model; it includes an input layer, a hidden layer and an output layer; each hidden layer includes multiple neurons, each neuron is connected to the neurons in the next layer, and the connection contains weights, which determine the importance and influence of data transmission in the neural network; an activation function is applied to each neuron between the hidden layer and the output layer. The activation function introduces nonlinearity, allowing the network to learn more complex patterns and features.

[0030] S5: Based on the priorities and communication characteristic data corresponding to the m groups of device feature data, dynamically allocate communication resource data and perform data transmission.

[0031] Communication resource data includes bandwidth and power; like Figure 2 As shown, the steps of dynamically allocating communication resource data include: Step S501: Define population size , iterative threshold , mobile factor and disturbance factor ; Step S502: construct M sets of data sets, , set different numerical labels for different data sets and mark them as set labels; Step S503: Constructing a population , the position of each individual in the population corresponds to the set label one by one, and individuals include hunters and prey; the number of iterations t corresponding to the population is 0; Step S504: determining an evaluation value function; Step S505: Randomly generate prey, calculate the selection probability of each prey and make a selection; Step S506: determining the global optimal position and calculating the trap depth corresponding to each hunter; Step S507: Calculate the attractiveness corresponding to each hunter; Step S508: updating the position of each hunter and prey; Step S509: Determine whether the number of iterations t is less than the iteration threshold. If so, set , and return to step S505, if not, go to step S510; Step S510: Update the global optimal position, obtain the set tag corresponding to the global optimal position, and dynamically allocate communication resource data according to the data set corresponding to the set tag.

[0032] In the above step S501, the population size , iterative threshold , mobile factor and disturbance factor A person skilled in the art collects multiple sets of different test data in the process of historical dynamic allocation of communication resource data. The test data includes m sets of research data. The research data includes data transmission volume, data transmission efficiency, priority number and communication characteristic data. The research data corresponds to the device characteristic data one by one. For each set of test data, multiple sets of different definition parameters are set. The definition parameters are the population size. , iterative threshold , mobile factor and disturbance factor For the same group of test data corresponding to different groups of definition parameters, the corresponding set labels are obtained through the swarm intelligence algorithm of step S501-step S510, and the definition parameters corresponding to the set label with the largest evaluation value are used as the definition parameters corresponding to the corresponding test data; and the definition parameters corresponding to multiple groups of different test data are obtained by analogy; the mean of the multiple groups of definition parameters (i.e., the population size mean, the iteration threshold mean, the moving factor mean, and the disturbance factor mean) is used as the population size defined in step S501. , iterative threshold , mobile factor and disturbance factor .

[0033] In the above step S502, the method for constructing M groups of data sets is: obtaining a total data value, the total data value includes the total network bandwidth and the total network power, and the total data value is obtained by a person skilled in the art through measurement; randomly dividing the total network bandwidth into m bandwidth components, and the sum of the m bandwidth components is the total network bandwidth; randomly dividing the total network power into m power components, and the sum of the m power components is the total network power; taking the m bandwidth components and the m power components as a group of data sets, where one bandwidth component and one power component corresponds to one low-altitude device; and so on to construct a total of M groups of data sets, each group of data sets is different.

[0034] In the above step S503, the population , For the individuals; the position of each individual in the population is defined in a one-dimensional search space, the range of the one-dimensional search space is M sets of data; the population The expression for each individual position in is: Where, is the position of the i-th individual, is the random coefficient of the i-th individual, , ; Randomly divide the population in Individuals as hunters, Individuals as prey, , .

[0035] In the above step S504, the expression of the evaluation value function is: Where, is the evaluation value, For the The priority number corresponding to the group device characteristic data, For the Adjust the transmission efficiency corresponding to the group device characteristic data, For the The adjusted transmission volume corresponding to the group device characteristic data, ; Adjust the transmission efficiency to the data transmission efficiency after the dynamic allocation of communication resource data, and adjust the transmission volume to the data transmission volume after the dynamic allocation of communication resource data; The method for obtaining the adjusted transmission volume and the adjusted transmission efficiency is as follows: the set labels corresponding to the research data and the individual positions are used as evaluation data; the evaluation data is input into the trained adjustment prediction model to predict the corresponding adjustment labels; the adjustment labels are digital labels corresponding to the adjustment data, and different adjustment data correspond to different digital labels, and the adjustment data include the adjusted transmission volume and the adjusted transmission efficiency; the training process of the adjustment prediction model is consistent with the training process of the priority prediction model, and both are deep neural network models; the adjustment labels corresponding to the evaluation data are collected by technical personnel in this field in the process of historical dynamic allocation of communication resource data, and the communication resource data is allocated according to the data set corresponding to the set label in each set of evaluation data, and the adjustment data of the allocated communication resource data is collected and the adjustment labels corresponding to the adjustment data are used as the adjustment labels of the corresponding evaluation data, and the corresponding adjustment labels are set for the multiple sets of evaluation data in turn.

[0036] In the above step S505, ,The method of calculating the selection probability of each prey and making the selection includes: Calculating populations The evaluation value of each prey in the , and add them up in turn to obtain the total evaluation value, divide the evaluation value of each prey by the total evaluation value to obtain the selection probability of each prey; sort the selection probability of each prey from large to small, and replace each selection probability in positive order with the sum of each selection probability and all selection probabilities before the corresponding selection probability; obtain all selection probabilities after replacement and mark them as replacement probabilities; the prey corresponding to the replacement probability is consistent with the prey corresponding to the selection probability before replacement; take every two adjacent replacement probabilities as an analysis set in positive order; mark the replacement probability with a large value in each analysis set as the first probability, and mark the replacement probability with a small value as the second probability; construct the probability screening range corresponding to each analysis set according to the first probability and the second probability of each analysis set, where the first probability is the maximum value of the probability screening range, the second probability is the minimum value of the probability screening range, and the minimum value is an open interval, and the maximum value is a closed interval; each analysis set corresponds to the prey corresponding to the first probability; where the minimum value of the probability screening range corresponding to the first prey is 0, and the maximum value is the corresponding selection probability; randomly generate indivual A random number between The probability of the random number screening range corresponds to the prey.

[0037] For example, the selection probabilities of the three prey are 0.3, 0.5, and 0.2, respectively. Since there is no value before 0.3, 0.3 is still 0.3 after replacement. Since 0.5 is preceded by 0.3, 0.5 is replaced by 0.5+0.3=0.8. Since 0.2 is preceded by 0.3 and 0.5, 0.2 is replaced by 0.2+0.3+0.5=1. Therefore, the replacement probabilities are 0.3, 0.8, and 1, respectively. 0.3 and 0.8 are an analysis set, and the probability screening range corresponding to the second prey is ; 0.8 and 1 are also an analysis set, then the probability screening range corresponding to the third prey is ; The probability screening range corresponding to the first prey is ; If the randomly generated random number is 0.5, since 0.5 is in In, therefore choose the second prey; In the above step S506, the method for determining the global optimal position is: calculating the evaluation value corresponding to each individual, and taking the position of the individual corresponding to the maximum evaluation value as the global optimal position.

[0038] The expression for the trap depth is: ; Where, For the The trap depth corresponding to the hunter, is the evaluation value corresponding to the global optimal position, For the The evaluation value corresponding to each hunter is .

[0039] In the above step S507, the expression of attractiveness is: Where, For the The attractiveness corresponding to each hunter, For the The trap depth corresponding to the hunter, .

[0040] In step S508, the method for updating the hunter's location includes: ; Where, After the update The position of a hunter, For the The position of a hunter, is the global optimal position, for A random number between .

[0041] Methods for updating prey locations include: ; Where, After the update The location of the prey, For the The location of the prey, , for A random number between .

[0042] In the above step S510, the method for updating the global optimal position is: calculate the evaluation value corresponding to each individual after the position is updated, and mark it as the first evaluation value; use the evaluation value corresponding to the current global optimal position as the second evaluation value; compare the first evaluation value with the second evaluation value, if the first evaluation value is greater than the second evaluation value, then update the global optimal position to the individual position corresponding to the first evaluation value; if the first evaluation value is less than or equal to the second evaluation value, then do not update the global optimal position.

[0043] S6: Monitor the data transmission process in real time and determine whether to generate an alarm instruction.

[0044] Methods for determining whether to generate an alarm instruction include: During the data transmission process, the data transmission volume corresponding to each low-altitude device is collected in real time and marked as the real-time transmission volume. The transmission volume difference is calculated based on the real-time transmission volume and the adjusted transmission volume. The expression of the transmission volume difference is: Where, is the transmission difference, is the real-time transmission volume, To adjust the transmission volume; preset a difference threshold, which is pre-set by technical personnel in this field based on actual conditions; compare the difference threshold with the transmission volume difference; if the transmission volume difference is greater than or equal to the difference threshold, an alarm instruction is generated, indicating that there may be network security incidents such as denial of service attacks, data leaks, malware activities, etc.; if the transmission volume difference is less than the difference threshold, no alarm instruction is generated.

[0045] This embodiment realizes the intelligent management and safety control of the low-altitude communication system through the steps of comprehensively dynamically sensing equipment characteristics, identifying work tasks, predicting priorities, intelligently allocating resources, and real-time monitoring of safety; it enables the low-altitude economic communication server to flexibly adapt to the dynamic changes of network load, quickly respond to the dynamic allocation of resources, and improve the generalization ability of application scenarios; thereby effectively improving communication efficiency and security, ensuring the reliability and flexibility of real-time data transmission, reducing the risk of network attacks and data leakage, and effectively responding to the growing communication needs in the low-altitude economic field.

[0046] Example 2 The present application also provides an electronic device. The electronic device may include one or more processors and one or more memories. The memories may store computer-readable code that, when executed by the one or more processors, may execute the above-described AI-based method for allocating secure resources for a low-altitude economic communication server.

[0047] The method or system according to the embodiment of the present application can also be implemented with the aid of the architecture of the electronic device shown in this application. The electronic device may include a bus, one or more CPUs, ROM, RAM, a communication port connected to a network, input / output, a hard disk, etc. The storage device in the electronic device, such as a ROM or a hard disk, can store an AI-based low-altitude economic communication server security resource allocation method provided in this application. Furthermore, the electronic device may also include a user interface. Of course, the architecture shown in this application is only exemplary. When implementing different devices, one or more components in the electronic device shown in this application may be omitted according to actual needs.

[0048] Example 3 As shown, one embodiment of the present application discloses a computer-readable storage medium. The computer-readable storage medium stores computer-readable instructions. When the computer-readable instructions are executed by a processor, the AI-based low-altitude economic communication server security resource allocation method according to the embodiment of the present application described with reference to the above figures can be executed. The storage medium includes, but is not limited to, volatile memory and / or non-volatile memory. Volatile memory may include, for example, random access memory (RAM) and cache memory. Non-volatile memory may include, for example, read-only memory (ROM), a hard disk, flash memory, etc.

[0049] Furthermore, according to embodiments of the present application, the processes described above with reference to the flowcharts can be implemented as computer software programs. For example, the present application provides a non-transitory machine-readable storage medium storing machine-readable instructions capable of being executed by a processor to perform the steps corresponding to the methods provided herein, such as an AI-based method for allocating secure resources for low-altitude economical communication servers. When executed by a central processing unit (CPU), this computer program performs the functions defined in the methods of the present application.

[0050] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any modifications or substitutions that can be easily conceived by a person skilled in the art within the technical scope disclosed in the present invention should be included in the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be based on the scope of protection of the claims.

[0051] Finally: The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. A method for allocating safe resources for low-altitude economic communication servers based on AI, characterized in that the method include: S1: Receive m sets of device characteristic data, where m is an integer greater than 1; S2: Analyze the m groups of equipment feature data and identify the corresponding work tasks; S3: Analyze the m groups of device characteristic data to obtain corresponding communication characteristic data; S4: Prioritize the m groups of device feature data according to the corresponding job tasks and communication feature data; S5: dynamically allocate communication resource data based on the priorities and communication characteristic data corresponding to the m groups of device feature data, and perform data transmission; S6: Monitor the data transmission process in real time and determine whether to generate an alarm instruction.

2. The AI-based low-altitude economic communication server safety resource allocation method according to claim 1 is characterized in that: The device characteristic data includes moving speed, data transmission volume and data transmission efficiency; the device characteristic data corresponds to the low-altitude device one by one; the moving speed is the moving speed of the low-altitude device relative to the ground; the data transmission volume is the total amount of data that the low-altitude device needs to transmit to the cloud platform per unit time; The data transmission efficiency is the total amount of data successfully transmitted from the low-altitude device to the cloud platform per unit time; The step of identifying the corresponding operation task includes: Step S201: establishing a fuzzy set, dividing each data of the device characteristic data into multiple fuzzy sets; Step S202: converting each data in the m groups of device characteristic data into the membership degree of the corresponding fuzzy set by using fuzzification technology; Step S203: define fuzzy rules; Step S204: matching the fuzzified m groups of equipment feature data with the fuzzy rules, performing fuzzy reasoning, and obtaining a fuzzy reasoning result, which is the membership degree of the m groups of equipment feature data corresponding to different operation tasks; Step S205: Defuzzify the fuzzy reasoning result to obtain the defuzzified result, which is the job task corresponding to the m groups of equipment feature data; the defuzzification operation is: according to the fuzzy reasoning result, respectively obtain the membership degree corresponding to each job task of each group of equipment feature data; compare the membership degree corresponding to each group of equipment feature data, and take the job task corresponding to the membership degree with the largest value as the job task corresponding to the corresponding equipment feature data, and identify the job tasks corresponding to the m groups of equipment feature data.

3. The AI-based low-altitude economic communication server safety resource allocation method according to claim 2 is characterized in that: The method for obtaining corresponding communication characteristic data includes: Communication characteristics data include bandwidth requirements and power requirements; The bandwidth requirement is expressed as: ； In the formula, For bandwidth requirements, is the data transmission volume, For data transmission efficiency, is the moving speed, , All are preset weight coefficients; The expression for power demand is: ; In the formula, For power demand, , is the preset scale factor.

4. The AI-based low-altitude economic communication server safety resource allocation method according to claim 3 is characterized in that: The method for prioritizing m groups of device feature data includes: Different digital labels are set for different job tasks and marked as task labels; the task labels and communication characteristic data are used as analysis data; the analysis data corresponding to each set of device feature data are input into the trained priority prediction model to predict the priority coefficient corresponding to each set of device feature data; the priority coefficients corresponding to each set of device feature data are sorted from large to small to obtain a priority sorting table; according to the positive order of the priority sorting table, the corresponding priority numbers are set for the m sets of device feature data in ascending order, and the range of the priority numbers is , where 1 is the highest priority and m is the lowest priority; The training process of the priority prediction model includes: Collect q groups of analysis data in advance, set corresponding priority coefficients for the q groups of analysis data, q is an integer greater than 1, and convert the analysis data and the corresponding priority coefficients into a corresponding set of feature vectors; use each set of feature vectors as the input of a priority prediction model, the priority prediction model uses a set of prediction priority coefficients corresponding to each group of analysis data as output, and uses the actual priority coefficients corresponding to each group of analysis data as a prediction target, and the actual priority coefficients are the pre-set priority coefficients corresponding to the analysis data; minimize the sum of prediction errors of all analysis data as a training target; train the priority prediction model until the sum of prediction errors converges and stops training; the priority prediction model is a deep neural network model.

5. The AI-based low-altitude economic communication server safety resource allocation method according to claim 4 is characterized in that: The communication resource data includes bandwidth and power; The step of dynamically allocating communication resource data comprises: Step S501: Define population size , Iteration Threshold , mobile factor and disturbance factor ; Step S502: construct M sets of data sets, , set different numerical labels for different data sets and mark them as set labels; Step S503: Constructing a population , the position of each individual in the population corresponds to the set label one by one, and the individuals include hunters and prey; the number of iterations t corresponding to the population is 0; Step S504: determining an evaluation value function; Step S505: Randomly generate prey, calculate the selection probability of each prey and make a selection; Step S506: determining the global optimal position and calculating the trap depth corresponding to each hunter; Step S507: Calculate the attractiveness corresponding to each hunter; Step S508: updating the position of each hunter and prey; Step S509: Determine whether the number of iterations t is less than the iteration threshold. If so, set , and return to step S505, if not, proceed to step S510; Step S510: Update the global optimal position, obtain the set tag corresponding to the global optimal position, and dynamically allocate communication resource data according to the data set corresponding to the set tag.

6. The AI-based low-altitude economic communication server safety resource allocation method according to claim 5 is characterized in that: In step S502, the method for constructing M groups of data sets is: obtaining a total data value, the total data value includes a total network bandwidth and a total network power; randomly dividing the total network bandwidth into m bandwidth components, the sum of the m bandwidth components is the total network bandwidth; randomly dividing the total network power into m power components, the sum of the m power components is the total network power; taking the m bandwidth components and the m power components as a group of data sets, wherein one bandwidth component and one power component correspond to one low-altitude device; and so on, constructing a total of M groups of data sets, each group of data sets is different; In step S503, the population , For the individuals; the position of each individual in the population is defined in a one-dimensional search space, and the range of the one-dimensional search space is M sets of data; the population The expression for each individual position in is: ; In the formula, is the position of the ith individual, is the random coefficient of the ith individual, , ; Randomly divide the population In As a hunter, Individuals as prey, , .

7. The AI-based low-altitude economic communication server safety resource allocation method according to claim 6 is characterized in that: In step S504, the expression of the evaluation value function is: ； In the formula, is the evaluation value, For the The priority number corresponding to the group device characteristic data, For the The adjusted transmission efficiency corresponding to the group equipment characteristic data, For the The adjusted transmission volume corresponding to the group device characteristic data, ; Adjust the transmission efficiency to the data transmission efficiency after the communication resource data is dynamically allocated, and adjust the transmission volume to the data transmission volume after the communication resource data is dynamically allocated; The method for obtaining the adjusted transmission volume and the adjusted transmission efficiency is: taking the data transmission volume, data transmission efficiency, priority number and communication characteristic data as research data, and taking the set label corresponding to the research data and the individual position as evaluation data; inputting the evaluation data into the trained adjustment prediction model to predict the corresponding adjustment label; the adjustment label is the digital label corresponding to the adjustment data, and different adjustment data have different corresponding digital labels, and the adjustment data includes the adjusted transmission volume and the adjusted transmission efficiency; the training process of the adjustment prediction model is consistent with the training process of the priority prediction model, and both are deep neural network models.

8. The AI-based low-altitude economic communication server safety resource allocation method according to claim 7 is characterized in that: In step S505, , the method of calculating the selection probability of each prey and making a selection includes: Calculate population The evaluation value of each prey in the set is added up in sequence to obtain the total evaluation value, and the evaluation value of each prey is divided by the total evaluation value to obtain the selection probability of each prey; the selection probability of each prey is sorted from large to small, and each selection probability is replaced in order by the sum of each selection probability and all selection probabilities before the corresponding selection probability; all selection probabilities after replacement are obtained and marked as replacement probabilities; the prey corresponding to the replacement probability is consistent with the prey corresponding to the selection probability before replacement; every two adjacent replacement probabilities are taken as an analysis set in order; the replacement probability with a large value in each analysis set is marked as the first probability, and the replacement probability with a small value is marked as the second probability; according to the first probability and the second probability of each analysis set, the probability screening range corresponding to each analysis set is constructed, wherein the first probability is the maximum value of the probability screening range, the second probability is the minimum value of the probability screening range, and the minimum value is an open interval, and the maximum value is a closed interval; each analysis set corresponds to the prey corresponding to the first probability; wherein the minimum value of the probability screening range corresponding to the first prey is 0, and the maximum value is the corresponding selection probability; randomly generate indivual A random number between The probability of a random number screening range corresponds to the prey.

9. The AI-based low-altitude economic communication server safety resource allocation method according to claim 8 is characterized in that: In step S506, the method for determining the global optimal position is: calculating the evaluation value corresponding to each individual, and taking the position of the individual corresponding to the maximum evaluation value as the global optimal position; The expression for the trap depth is: ; In the formula, For the The trap depth corresponding to the hunter is is the evaluation value corresponding to the global optimal position, For the The evaluation value of each hunter ; In step S507, the expression of the attractiveness is: ; In the formula, For the The attractiveness corresponding to each hunter is For the The trap depth corresponding to the hunter is ; In step S508, the method for updating the hunter's position includes: ； In the formula, After the update The position of a hunter, For the The position of a hunter, is the global optimal position, for A random number between Methods for updating prey locations include: ； In the formula, After the update The location of the prey, For the The location of the prey, , for A random number between In step S510, the method for updating the global optimal position is: calculate the evaluation value corresponding to each individual after the position is updated, and mark it as the first evaluation value; use the evaluation value corresponding to the current global optimal position as the second evaluation value; compare the first evaluation value with the second evaluation value, if the first evaluation value is greater than the second evaluation value, update the global optimal position to the individual position corresponding to the first evaluation value; if the first evaluation value is less than or equal to the second evaluation value, do not update the global optimal position.

10. The AI-based low-altitude economic communication server safety resource allocation method according to claim 9 is characterized in that: The method for determining whether to generate an alarm instruction comprises: During the data transmission process, the data transmission volume corresponding to each low-altitude device is collected in real time and marked as the real-time transmission volume; the transmission volume difference is calculated based on the real-time transmission volume and the adjusted transmission volume; the expression of the transmission volume difference is: ; In the formula, is the transmission difference, is the real-time transmission volume, To adjust the transmission volume; preset a difference threshold, compare the difference threshold with the transmission volume difference; if the transmission volume difference is greater than or equal to the difference threshold, generate an alarm instruction; if the transmission volume difference is less than the difference threshold, no alarm instruction is generated.

Images