Secure collaborative unloading method in ultra-dense millimeter wave cache network

Abstract

The invention discloses a secure collaborative unloading method in an ultra-dense millimeter wave cache network, and the method specifically comprises the steps: constructing a network architecture, and configuring an optimization problem under multiple constraints in the network architecture; randomly generating an initial solution set according to the optimization problem, defining the initial solution set as an original population, and searching and iterating the original population by adopting an improved sparrow algorithm to obtain a target population; further searching by adopting an improved sparrow algorithm according to position information codes of all individuals in the target population, and outputting historical optimal position information; and performing safe collaborative unloading resource configuration according to the historical optimal position information of the target population. According to the method, the cache configuration is unloaded by the base station, and the task unloading cache configuration is jointly optimized and calculated under the conditions that the security vulnerability cost, the time delay limit, the user calculation capacity and the transmitting power are met, so that the task unloading energy consumption is reduced to the minimum.

Description

Technical Field

[0001] This invention relates to the field of wireless communication technology, and more specifically to a secure collaborative offloading method in ultra-dense millimeter-wave buffered networks. Background Technology

[0002] In recent years, with the rapid development of 5G communication technology, smart devices have been widely used in various fields such as biometrics, virtual reality, autonomous driving, and IoT-based healthcare, bringing about an urgent demand for low latency and high computing power. However, traditional terminal devices are limited by computing resources and battery life, making it difficult to meet the performance requirements of these emerging applications. To alleviate the pressure on terminals, mobile edge computing (MEC) has emerged. Its core concept is to deploy computing resources at the network edge, allowing users to offload tasks to edge servers, thereby reducing latency and improving service quality.

[0003] To shorten the distance between users and computing centers, MEC systems with ultra-dense network architectures have attracted attention. These systems densely deploy small base stations around macro base stations, enabling users to connect to multiple base stations, expanding resource coverage, and improving computing resource utilization. However, this also brings problems such as signal interference and increased energy consumption. Furthermore, task offloading requires cryptographic algorithms, further consuming computing resources. Therefore, ensuring communication quality and data security while rationally allocating resources and controlling interference to reduce energy consumption has become a key issue.

[0004] To address the aforementioned challenges, the Sparrow Search Algorithm (SSA) is an emerging swarm intelligence optimization method with promising application prospects. It simulates the foraging and anti-predation behavior of sparrows, featuring a simple structure, few parameters, and ease of implementation. By dividing sparrows into three roles—discoverer, follower, and watchdog—it balances exploration and exploitation capabilities. Compared to traditional algorithms, it exhibits faster convergence, stronger robustness and adaptability, is suitable for complex optimization problems, and also possesses parallelism and flexibility, easily integrating with other strategies. Introducing it into a secure collaborative offloading architecture for ultra-dense millimeter-wave caching networks holds promise for balancing security, energy consumption, and performance. Therefore, introducing SSA into a secure collaborative offloading architecture for ultra-dense millimeter-wave caching networks is expected to achieve a better balance between security, energy consumption, and performance. Summary of the Invention

[0005] To address the shortcomings of current technologies, this invention provides a secure collaborative offloading method for ultra-dense millimeter-wave cache networks, aiming to overcome the deficiencies of existing methods in computational offloading cache research and improve the algorithm's global search capabilities.

[0006] To achieve the above objectives, the present invention adopts the following technical solution: a secure collaborative offloading method in ultra-dense millimeter-wave caching networks, comprising the following steps: Step S1: Obtain the basic network information for secure collaborative offloading in the ultra-dense millimeter-wave caching network and construct the network architecture. The network architecture includes a system model, a communication model, a caching and computing model, and a security model. Configure the optimization problem under multiple constraints in the network architecture. Step S2: Randomly generate an initial solution set based on the optimization problem, and define the initial solution set as the original population. Calculate the fitness value, and take the individual with the highest fitness as the best initial individual. Improve the sparrow algorithm by randomly perturbing the number of iterations, introducing control parameters, nonlinear weighting factors, and sine and cosine search algorithms. The improved sparrow algorithm searches and iterates on the original population to obtain the target population, and outputs the position information encoding of all individuals in the target population. Step S3: Based on the location information encoding of all individuals in the target population, the improved sparrow algorithm is used to further search, continuously update the location of all individuals, select the better individual between the current best individual and the initial best individual as the historical best individual, and output the historical best position information of the historical best individual. Step S4: Perform secure collaborative unloading of resource configurations based on the historical best position information of the historical best individual in the target population.

[0007] Further, the detailed process of step S2 is as follows: Population initialization is performed based on the optimization problem. The initialization process includes generating a set of random candidate solutions as the initial solution set within the constraints of the optimization problem, defined as the original population. The optimal individual for population initialization is found based on the fitness value. The algorithm iteration process continues: a safety threshold is set, and the sparrow search algorithm is improved by randomly perturbing the number of iterations, introducing control parameters, nonlinear weighting factors, and using sine and cosine search algorithms. The improved sparrow search algorithm consists of three types of sparrow individuals: discoverers, followers, and watchers. Discoverers, followers, and watchers are obtained by sorting them in descending order of fitness value. Discoverers are responsible for extensive searching in the solution space to explore potential optimal regions. The discoverer's position update strategy is adjusted according to the environmental safety level. When the alarm value is less than the safety threshold, the discoverer uses an exponential decay strategy to conduct a large-scale search; when the alarm value is greater than or equal to the safety threshold, random perturbations are introduced to cause position jumps, simulating the behavior of sparrows fleeing dangerous areas; followers update their positions based on the information provided by the discoverer, simulating the foraging behavior of sparrows in a flock. The follower position update strategy incorporates control parameters, nonlinear weighting factors, and sine and cosine search algorithms, adjusting angle variables and dynamic factors to achieve an adaptive balance between global exploration and local development during the follower search process; the watchdog is responsible for monitoring the population status, and the watchdog position update strategy distinguishes based on the individual's current fitness: when an individual is on the edge, it moves towards the current optimal position to avoid danger; when an individual is in the center, it randomly moves towards its neighborhood; finally, the position information encoding of all individuals in the target population is output.

[0008] Furthermore, the detailed process of step S3 is as follows: Based on the location information encoding of all individuals in the target population, first determine whether the number of iterations exceeds the maximum limit. If it does not exceed the maximum limit, perform discoverer search, follower follow, and vigilant monitoring behaviors to update the location information. After updating the locations of the three types of sparrow individuals, use a boundary function to correct out-of-bounds variables, recalculate the population fitness value, compare the new location fitness with the original location fitness, and retain the individual with the higher fitness value as the next generation individual. If the maximum limit of the number of iterations is exceeded, terminate the iteration and output the location encoding of all individuals. After the iteration ends, select the better individual between the current best individual and the initial best individual as the historical best individual, and output the relevant location information.

[0009] Furthermore, in step S1, obtaining basic network information includes: in an ultra-dense millimeter-wave caching network, express A set of indexes for each user. ; express A set of indexes for each base station. ; express A set of indices for each subtask. ; express A set of indices for cryptographic algorithms. ; Get users The computational task of quadruples ,in, Indicates user Calculate the data size of the subtask. Indicates user Number of CPU cycles required to complete the computational subtask Indicates user Calculate the maximum execution time of the subtask. Indicates user subtasks The expected security level is determined; next, the physical locations of all users are obtained, and the K-means clustering algorithm is used to divide all users into groups based on their physical locations. Clusters, among which, users The cluster it belongs to is denoted as Obtain the total bandwidth of the network system. Sub-channel bandwidth ,user With base station Channel gain between and noise power Subsequently, user acquisition Maximum computing resources CPU energy coefficient and base stations Maximum computing resources Finally, acquire users. The set of base stations associated with all subtasks ,in, Indicates user subtasks The associated base station index number, Indicates user The total number of subtasks, m=1,2,...,M.

[0010] Furthermore, the expression for the optimization problem generated in step S1 is as follows: In the above formula, E(·) represents the objective function to be optimized, and X represents the user's association matrix. , Indicates user With base station The correlation index; Z represents the sub-channel selection matrix, , Indicates user Whether to select a sub-channel, where, Indicates each cluster A set of sub-channel indexes U represents the association matrix between the user's subtasks and the base station. , Indicates user subtasks At most one base station Interrelated execution; V represents the choice matrix of the cryptographic algorithm. , Indicates user subtasks Choose a cryptographic algorithm F represents the set of computing capacities that the user is allowed to use. , Indicates user Allowable computational capacity; P represents the transmit power allocation matrix. , Indicates user The transmit power to the associated base station; Indicates user subtasks Processing latency The maximum execution time of this task cannot be exceeded. Constraints; Indicates user Constraint that it can only be associated with a single base station; Indicates user Constraints for selecting a sub-channel within the user's cluster; Indicates user Task At most one base station Constraints enforced above; Indicates user subtasks The constraint that only one cryptographic algorithm can be chosen; Indicates user Overall security vulnerability cost The maximum allowable cost shall not be exceeded. Constraints; Indicates user Constraints on the minimum and maximum allowable computing power for offloading to associated base stations. Indicates user Maximum allowed capacity To represent a sufficiently small value, Indicates user The constraints on the lower and upper limits of the transmit power offloaded to the relevant base station. Indicates the user's maximum allowed power. Indicates user When associated with a single base station, Otherwise, the constraint condition is 0. This indicates that when subchannel s is selected, Otherwise, the constraint condition is 0. Indicates user subtasks When executed on a base station Otherwise, the constraint condition is 0. Indicates the choice of cryptographic algorithm hour, Otherwise, the constraint condition is 0; Indicates user Local processing energy consumption.

[0011] Furthermore, the specific steps of step S2 are as follows: Step S2.1: Initialize the maximum number of iterations using the improved sparrow search algorithm. and set the current iteration number Set to 1; Step S2.2: The number of individual sparrows in the sparrow population is Any individual sparrow in a sparrow population is used Let represent the sparrow populations of the improved sparrow search algorithm; Each individual sparrow Encoded into position vectors , Encoded into position vectors , Encoded into position vectors , Encoded into position vectors , Encoded into position vectors , Encoded into position vectors ,in, Represents all users The number of virtual users for all subtasks m The set of indices, Represents individual sparrows Chinese users The index number of the associated base station, Represents individual sparrows Chinese users The selected channel index number, Represents individual sparrows Chinese users The number of index numbers associated with the base station, Represents individual sparrows Chinese users The selected cryptographic algorithm index number, Represents individual sparrows Chinese users Computing capacity, Represents individual sparrows Chinese users The transmission power; Step S2.3: Initialize the sparrow population and construct the individual sparrows in the population. fitness function : In the improved sparrow search algorithm, the optimization problem is... and Two constraints are transformed into penalty terms embedded in fitness evaluation; any sparrow individual The fitness function is expressed as: in, Indicates user The penalty factor; Indicates user subtasks The penalty factor; The fitness function is used to evaluate the fitness values ​​of all individuals in the sparrow population. The individuals are sorted in descending order of fitness value, and the sparrow with the largest fitness value is recorded as the best sparrow and the sparrow with the smallest fitness value as the worst sparrow. The fitness values ​​are also recorded. Step S2.4: The position of each individual sparrow is considered a valid solution. Within the solution space, multiple sparrows continuously update their positions by simulating the behaviors of discoverers, followers, and watchers; then the process begins iteratively, proceeding through the iteration number... Less than or equal to the maximum number of iterations At that time, sparrow populations were explored, followed, and monitored to simulate different foraging behaviors and update their locations.

[0012] Furthermore, when a sparrow acts as a discoverer, the position update process is represented as follows: A group in Navigation in 3D search space Only sparrows; the position of each sparrow, in this... Represented in the 3D search space as In the formula, , Indicates the first Only sparrows Position in the dimensional search space; Indicates the current sparrow's number There are decision variables; among which They represent the first During round iteration, the base station index selected by the user's mobile device, the channel index selected by the user's mobile device, the associated base station index selected by the user's subtask, the cryptographic algorithm index selected by the user's mobile device, the allowable computing capacity of the user's mobile device, and the transmission power from the user's mobile device to the associated base station are all considered. The first During round iteration, the base station index selected by the user's mobile device, the channel index selected by the user's mobile device, the associated base station index selected by the user's subtask, the cryptographic algorithm index selected by the user's mobile device, the allowable computing capacity of the user's mobile device, and the transmission power from the user's mobile device to the associated base station are all considered. Represents a uniformly random number between 0 and 1; Represents a random number that follows a standard normal distribution; This represents a one-dimensional matrix where each element has a value of 1. Indicates the alarm value, in Variation within the range, Indicates the safety threshold, which is limited to Within the range; when At that time, the surrounding environment was safe, and the discoverer needed to conduct extensive exploration to guide the sparrows to more food; when Immediately release danger signals to prompt the population to quickly adopt anti-predation behaviors, adjust search strategies, and move rapidly to a safe area; This indicates rounding the result down to the nearest integer; exp() represents the exponential function. To improve the effectiveness of the improved sparrow search algorithm, when the alarm value... Update the location of the discoverer as needed: At the start of the iteration, The distribution follows a Cauchy distribution; in the middle stage of the iteration, The distribution lies between the Cauchy and Gaussian distributions; in later iterations, The distribution follows a Gaussian distribution. The discoverer can then search for the region where the current solution is close to the optimal solution of the problem. trnd(·) represents a random perturbation function, which generates random values ​​based on the t-distribution and is used to add random fluctuations to the update of the discoverer's position-related parameters.

[0013] Furthermore, when a sparrow acts as a follower, the position update process is represented as follows: The update formula for followers is optimized by introducing control parameters, nonlinear weighting factors, and sine and cosine search algorithms. Dynamic amplitude control: Introducing control parameters The control parameters consist of the iteration number t and the adjustment factor. Decide, Adaptively adjust the amplitude of the sine and cosine functions; Multidimensional random constraints: This represents a random number that determines the distance the sparrow moves. This represents a random number that controls the influence of the optimal individual on the sparrow's next position. It switches equally between sine and cosine components; Nonlinear Dependency Adjustment: Introducing a Nonlinear Weighting Factor The dynamic balance of position updates to current position information helps prevent followers from getting trapped in local optima. Where e represents the natural constant; The follower's position is updated using the following formula: In the formula, They represent the first During round iteration, the optimal positions of the base station index and channel index selected by the user's mobile device, the associated base station index and cryptographic algorithm index selected by the user's subtask, the optimal positions occupied by the computing capacity allowed by the user's mobile device, and the optimal positions of the transmission power from the user's mobile device to the associated base station are determined.

[0014] Furthermore, the specific process of step S3 is as follows: Step S3.1: When completing a round of position update, the boundary function must be revised after the three different types of sparrows update their positions to correct out-of-bounds variables. After the guard's position is updated, the fitness value of the population must be recalculated. After one round of iteration, the fitness value of all sparrow individuals is calculated and compared, and the individual with the highest fitness is regarded as the current best sparrow individual. Step S3.2: Compare the fitness value of the current best sparrow individual with the fitness value of the historical best sparrow individual. If the current best sparrow individual is better, replace the historical best sparrow individual with the current best sparrow individual, update the historical best position variable, and output its position information code as the input of the next stage algorithm. Otherwise, retain the historical best sparrow individual and increment the current iteration count by 1. Step S3.3: Determine the current iteration number Has the maximum number of iterations been exceeded? If the current iteration number Less than or equal to the maximum number of iterations Then continue to explore, discover, follow, and monitor the sparrow population; if the current iteration number is... Greater than the maximum number of iterations Then, it directly outputs the information encoding of all individual sparrows, including position variables and fitness values.

[0015] The beneficial effects of this invention are as follows: This invention combines multi-strategy balanced search and development to avoid local optima and improve convergence. It can minimize the energy consumption of task offloading while meeting safety, energy consumption, and latency constraints. Attached Figure Description

[0016] To more intuitively demonstrate the effects of the proposed solution, the following is a brief description of the accompanying drawings used in the examples. The accompanying drawings show some of the parameter indicators.

[0017] Figure 1 This is a flowchart of the present invention.

[0018] Figure 2 This is a schematic diagram illustrating the impact of user mobile device density on energy consumption according to the present invention.

[0019] Figure 3This is a schematic diagram illustrating the effect of the number of iterations on the fitness value according to the present invention. Detailed Implementation

[0020] like Figure 1 As shown, the present invention provides a technical solution: a secure collaborative offloading method in an ultra-dense millimeter-wave caching network, comprising the following steps: Step S1: Obtain the basic network information for secure collaborative offloading in the ultra-dense millimeter-wave caching network and construct the network architecture. The network architecture includes a system model, a communication model, a caching and computing model, and a security model. Configure the optimization problem under multiple constraints in the network architecture. Step S2: Randomly generate an initial solution set based on the optimization problem, and define the initial solution set as the original population. Calculate the fitness value, and take the individual with the highest fitness as the best initial individual. Improve the sparrow algorithm by randomly perturbing the number of iterations, introducing control parameters, nonlinear weighting factors, and sine and cosine search algorithms. The improved sparrow algorithm searches and iterates on the original population to obtain the target population, and outputs the position information encoding of all individuals in the target population. Step S3: Based on the location information encoding of all individuals in the target population, the improved sparrow algorithm is used to further search, continuously update the location of all individuals, select the better individual between the current best individual and the initial best individual as the historical best individual, and output the historical best position information of the historical best individual. Step S4: Perform secure collaborative unloading of resource configurations based on the historical best position information of the historical best individual in the target population.

[0021] This study obtains the basic network information for secure collaborative offloading in an ultra-dense millimeter-wave caching network. Based on this information, it further constructs system, communication, caching, computation, and security models within the corresponding network architecture. Within this network architecture framework, it then constructs an optimization problem to achieve the goal of minimizing energy consumption. The specific steps are as follows: Step S1.1: Obtaining basic network information includes: in an ultra-dense millimeter-wave caching network, express A set of indexes for each user. ; express A set of indexes for each base station. ; express A set of indices for each subtask. ; express A set of indices for cryptographic algorithms. ; Get users The computational task of quadruples ,in, Indicates user Calculate the data size of the subtask. Indicates user Number of CPU cycles required to complete the computational subtask Indicates user Calculate the maximum execution time of the subtask. Indicates user subtasks The expected security level is determined; next, the physical locations of all users are obtained, and the K-means clustering algorithm is used to divide all users into groups based on their physical locations. Clusters, among which, users The cluster it belongs to is denoted as Obtain the total bandwidth of the network system. Sub-channel bandwidth ,user With base station Channel gain between and noise power Subsequently, user acquisition Maximum computing resources CPU energy coefficient and base stations Maximum computing resources Finally, acquire users. The set of base stations associated with all subtasks ,in, Indicates user subtasks The associated base station index number, Indicates user The total number of subtasks, m=1,2,...,M.

[0022] Step S1.2: Construct the system model; In a network system, there are Each base station (BS) is equipped with a cache and computing server.

[0023] Each type of task can be handled by the user or offloaded to an edge server for computation. When the latter approach is used, these tasks require confidential processing and security services. To reduce network interference and improve resource utilization, the base stations are divided into clusters based on their physical locations using the K-means clustering algorithm. There are several clusters. Different clusters utilize different spectrums, but base stations within the same cluster have a set of sub-channel indices. From 1 to The same sub-channel (band) indexed, where , It is a rounding function; Indicates system bandwidth. This represents the sub-channel bandwidth. These sub-channels can be used by users associated with base stations in the same cluster via non-orthogonal multiple access (NOMA) in millimeter-wave networks. In this framework, each user can only be associated with one base station (BS), and it can only utilize one sub-channel of the associated base station via NOMA. Clearly, inter-cluster interference does not exist, but intra-cluster interference does.

[0024] Step S1.3: Construct a communication model; In order to accurately represent the path loss in ultra-dense networks, a comprehensive millimeter-wave channel model is adopted.

[0025] Frequency division multiple access (FDMA) is used to distribute the total bandwidth of the network system equally. If there are clusters, then the number of sub-channels in each cluster is... The index set of each cluster sub-channel is denoted as ,in, Represents a sub-channel within any user cluster. This represents the number of sub-channels within any user cluster. This indicates that the result is rounded down to the nearest integer; each user within each cluster utilizes its own occupied subchannel to simultaneously transmit the offloaded portions of its multiple subtasks to multiple base stations using NOMA technology; subsequently, according to Shannon's formula... Calculation in sub-channel Above, users To base station transmission rate ,in, For users At the base station Received self-interference, ; For users At the base station sub-channels The device received interference from different users. , Indicates user Selected sub-channel index number; Indicates user to base station The transmission power, For users to base station The transmission power accounts for the user The proportion of total transmit power; Indicates user Total transmit power; Indicates user to base station Channel gain; This indicates that user k is connected to its associated base station. Channel gain; Indicates noise power; Indicates user Any user within the cluster Selected sub-channel index number; Indicates user In sub-channel The corresponding user set; Represents any user The associated base station index number; Represents any user To its associated base station The transmission power.

[0026] Step S1.4: Constructing the caching and computation model. This invention uses a more complete version of the caching and computation model. Furthermore, unlike other caching decisions, this invention assumes that user tasks are cached in advance based on their popularity. When a user... With base station During association, the first step is to verify whether the base station (BS) has cached its task. Then the user subtasks It is calculated locally, of which Indicates user Task At the base station The unloading (execution) metric at the location, with the following constraints: ; Indicates user subtasks At the base station Execution at the location, This indicates that the action will not be executed. If... Then directly at the selected base station Calculate user subtasks and the results from the base station Uninstall to user In this invention, Indicates user and base stations The related indicators, namely ; Indicates user With base station Related, This indicates that they are not related. Furthermore, For subtasks At the base station The cache metrics at that location, i.e. ; Subtasks Cached at base station middle, This indicates that there is no cache. If , This represents the cache index of user k's subtask m at auxiliary base station n'. If user k's subtask m is executed at auxiliary base station n', then user subtasks The calculation is offloaded from the selected base station to another auxiliary base station via a wired link. If Then the user subtasks Directly at auxiliary base stations Calculate directly. When At that time, the user subtasks The calculation is offloaded to the selected base station. In this case, if At that time, the user will be connected via a wired link. subtasks The calculation is offloaded from another auxiliary base station to the selected base station. If ,user subtasks Initially, it is offloaded to the selected base station, and then transmitted via a wired link to another auxiliary base station for calculation. Finally, when At that time, the user is directly calculated at the auxiliary base station. subtasks .

[0027] In local computing, if Then execute the user locally. subtasks Therefore, users subtasks The local execution time is , here Indicates user Computing power. Users subtasks The local execution energy consumption is ,in It is the energy coefficient. Refers to users Task It can be executed on a maximum of one base station. In edge computing, if... or At that time, it is necessary to include the user Task Data is transmitted to the base station via NOMA. or auxiliary base station The calculation is performed at that location. Therefore, the user subtasks The upload time is: Here, ;user Select sub-channel ,but Otherwise, it is 0. User subtasks Upload power consumption: In addition, if or At that time, it is necessary to include the user subtasks The data is transmitted from the user-selected base station to the auxiliary base station via a wired link; if... ,user subtasks The calculation is offloaded from another auxiliary base station to the user-selected base station via a limited link. subtasks The corresponding return time is: in, Indicates the backhaul rate between base stations, for users subtasks The remote execution time is: (b) refers to the user The fact that it can only connect to one base station It is a subtask At the base station Execution time.

[0028] Step S1.5: Construct a security model; Offloading computing tasks to edge servers makes them vulnerable to attacks, and mobile edge computing networks commonly use cryptographic algorithms for protection. However, improving the robustness of cryptographic algorithms significantly increases energy consumption and latency, and cannot absolutely prevent vulnerabilities, making the quantification of security risks a challenge. Different cryptographic algorithms correspond to different security levels, and encryption operations are computationally expensive, also incurring additional overhead such as latency and energy consumption.

[0029] express A set of indices for cryptographic algorithms. Any cryptographic algorithm The protection level is , The higher the value, the stronger the robustness of the cryptographic algorithm. Additionally, the... The computational power of a cryptographic algorithm to encrypt 1 bit of data is: (Unit: CPU cycles / bit) and the The computational power required for a cryptographic algorithm to decrypt 1 bit of data is (Unit: CPU cycles / bit), the The energy consumption for encrypting and decrypting 1 bit of data is the same using the first cryptographic algorithm, indicating that the first... The energy consumption of a cryptographic algorithm to encrypt / decrypt 1 bit of data is (unit: / bit). This invention is represented as , , .

[0030] In this invention, tasks cached on the base station are encrypted. When certain subtasks need to be uploaded from the user to the base station for computation, the user encrypts the subtasks before transmission. Regardless of whether certain subtasks are cached on the base station or received from the user or another base station, the base station can execute them after decryption. When the user... subtasks Using cryptographic algorithms At that time, the failure probability can be given by the following formula: In the formula For users subtasks The associated security risk coefficient, if the cryptographic algorithm If the security level is at least as high as that required by the subtask, then the cryptographic algorithm... It can effectively protect users subtasks Protected from attacks. Conversely, if the cryptographic algorithm... If the security level is insufficient, there is a certain probability that the user cannot be protected. subtasks .user subtasks The local encryption time is Local encryption power consumption is In the formula For decision variables; if the user subtasks Choose a cryptographic algorithm ,but The value is 1 if the condition is met, and 0 otherwise. Then the user... subtasks The remote decryption time is ,user subtasks The cost of security vulnerabilities is ,in Indicates user subtasks The economic losses caused by failure. Therefore, users The total cost of security vulnerabilities can be determined by Provided.

[0031] Step S1.6: Construct an optimization problem within the system architecture consisting of multiple models described above: The total local energy consumption required to complete all sub-tasks m of the user's mobile device is... ,in, For local execution energy consumption, For local encryption power consumption, For users subtasks Upload power consumption.

[0032] Furthermore, the expression for the optimization problem generated in step S1 is as follows: In the above formula, E(·) represents the objective function to be optimized, and X represents the user's association matrix. , Indicates user With base station The correlation index; Z represents the sub-channel selection matrix, , Indicates user Whether to select a sub-channel, where, Indicates each cluster A set of sub-channel indexes U represents the association matrix between the user's subtasks and the base station. , Indicates user subtasks At most one base station Interrelated execution; V represents the choice matrix of the cryptographic algorithm. , Indicates user subtasks Choose a cryptographic algorithm F represents the set of computing capacities that the user is allowed to use. , Indicates user Allowable computational capacity; P represents the transmit power allocation matrix. , Indicates user The transmit power to the associated base station; Indicates user subtasks Processing latency The maximum execution time of this task cannot be exceeded. Constraints; Indicates user Constraint that it can only be associated with a single base station; Indicates user Constraints for selecting a sub-channel within the user's cluster; Indicates user Task At most one base station Constraints enforced above; Indicates user subtasks The constraint that only one cryptographic algorithm can be chosen; Indicates user Overall security vulnerability cost The maximum allowable cost shall not be exceeded. Constraints; Indicates user Constraints on the minimum and maximum allowable computing power for offloading to associated base stations. Indicates user Maximum allowed capacity To represent a sufficiently small value, this invention uses 10. -20 , Indicates user The constraints on the lower and upper limits of the transmit power offloaded to the relevant base station. Indicates the user's maximum allowed power. Indicates user When associated with a single base station, Otherwise, the constraint condition is 0. This indicates that when subchannel s is selected, Otherwise, the constraint condition is 0. Indicates user subtasks When executed on a base station Otherwise, the constraint condition is 0. Indicates the choice of cryptographic algorithm hour, Otherwise, the constraint condition is 0; Indicates user Local processing energy consumption.

[0033] Furthermore, the specific steps of step S2 are as follows: Step S2.1: Initialize the maximum number of iterations using the improved sparrow search algorithm. and set the current iteration number Set to 1; Step S2.2: The number of individual sparrows in the sparrow population is Any individual sparrow in a sparrow population is used Let represent the sparrow populations of the improved sparrow search algorithm; Each individual sparrow Encoded into position vectors , Encoded into position vectors , Encoded into position vectors , Encoded into position vectors , Encoded into position vectors , Encoded into position vectors ,in, Represents all users The number of virtual users for all subtasks m The set of indices, Represents individual sparrows Chinese users The index number of the associated base station, Represents individual sparrows Chinese users The selected channel index number, Represents individual sparrows Chinese users The number of index numbers associated with the base station, Represents individual sparrows Chinese users The selected cryptographic algorithm index number, Represents individual sparrows Chinese users Computing capacity, Represents individual sparrows Chinese users The transmission power; Step S2.3: Initialize the sparrow population and construct the individual sparrows in the population. fitness function : In the improved sparrow search algorithm, the optimization problem is... and Two constraints are transformed into penalty terms embedded in fitness evaluation; any sparrow individual The fitness function is expressed as: in, Indicates user The penalty factor; Indicates user subtasks The penalty factor; The fitness function is used to evaluate the fitness values ​​of all individuals in the sparrow population. The individuals are sorted in descending order of fitness value, and the sparrow with the largest fitness value is recorded as the best sparrow and the sparrow with the smallest fitness value as the worst sparrow. The fitness values ​​are also recorded. Step S2.4: The position of each individual sparrow is considered a valid solution. Within the solution space, multiple sparrows continuously update their positions by simulating the behaviors of discoverers, followers, and watchers; then the process begins iteratively, proceeding through the iteration number... Less than or equal to the maximum number of iterations At the same time, sparrow populations were explored, tracked, and monitored for different foraging behaviors, and their locations were updated accordingly. The specific process of step S2.4 is as follows: Step S2.4.1: The position of each sparrow is considered as a valid solution. Within the solution space, multiple sparrows continuously update their positions by simulating the behavior of discoverers, followers, and watchers until the population's historical optimal solution is output. Sparrows are highly social and alert birds, exhibiting clear role division and cooperative behavior during foraging. In the improved sparrow search algorithm, individuals dynamically adjust their positions during iterations by simulating sparrow foraging behavior, achieving efficient exploration and development of the solution space. The improved algorithm accelerates convergence and enhances global search capabilities by setting an optimal individual during the initialization phase and comparing the current optimal individual with the historical optimal individual in each iteration. The historical optimal individual refers to the sparrow with the highest fitness value across all iterations, and its position variable is considered the current optimal solution.

[0034] When a sparrow acts as the discoverer, its position update process can be represented as follows: A group in Navigation in 3D search space Only sparrows; the position of each sparrow, in this... Represented in the 3D search space as In the formula, , Indicates the first Only sparrows Position in the dimensional search space; Indicates the current sparrow's number There are decision variables; among which They represent the first During round iteration, the base station index selected by the user's mobile device, the channel index selected by the user's mobile device, the associated base station index selected by the user's subtask, the cryptographic algorithm index selected by the user's mobile device, the allowable computing capacity of the user's mobile device, and the transmission power from the user's mobile device to the associated base station are all considered. The first During round iteration, the base station index selected by the user's mobile device, the channel index selected by the user's mobile device, the associated base station index selected by the user's subtask, the cryptographic algorithm index selected by the user's mobile device, the allowable computing capacity of the user's mobile device, and the transmission power from the user's mobile device to the associated base station are all considered. Represents a uniformly random number between 0 and 1; Represents a random number that follows a standard normal distribution; This represents a one-dimensional matrix where each element has a value of 1. Indicates the alarm value, in Variation within the range, Indicates the safety threshold, which is limited to Within the range; when At that time, the surrounding environment was safe, and the discoverer needed to conduct extensive exploration to guide the sparrows to more food; when Immediately release danger signals to prompt the population to quickly adopt anti-predation behaviors, adjust search strategies, and move rapidly to a safe area; This indicates that the result is rounded down to the nearest integer, and exp(·) represents the exponential function; in addition, the larger coefficient in the discoverer update formula is 500, which makes it converge to the potential optimal solution faster, and helps the algorithm to perform a broader search in the early iterations and a finer search in the later iterations.

[0035] To improve the effectiveness of the improved sparrow search algorithm, when the alarm value... Update the location of the discoverer as needed: All location-related parameters of the discoverer are obtained through "current value". (1 + random perturbation) Updated in the manner of ")". At the start of the iteration, The distribution follows a Cauchy distribution, and the discoverer possesses strong global search capabilities within the search space; in the intermediate stages of iteration, The distribution lies between the Cauchy and Gaussian distributions; in later iterations, The distribution follows a Gaussian distribution, allowing the discoverer to search for the region where the current solution is close to the optimal solution of the problem. trnd(·) represents a random perturbation function, which generates random values ​​based on the t-distribution, adding random fluctuations to the update of the discoverer's position-related parameters. This ensures both the flexibility of the search and adapts to the needs of global exploration and local fine-grained search at different iteration stages.

[0036] To improve the performance of the improved sparrow search algorithm, the follower position update process was refined: If the discoverer's search location is a local optimum, many followers will follow the discoverer into these local optima, causing the entire population to stop moving and resulting in a loss of population diversity. Therefore, to further reduce the probability of local optima, control parameters, nonlinear weighting factors, and sine / cosine search algorithms are introduced to optimize the follower update formula. Dynamic amplitude control: Introducing control parameters The control parameters consist of the iteration number t and the adjustment factor. Decide, The amplitude of the sine and cosine functions is adaptively adjusted, allowing the follower to maintain global exploration capabilities in the early stages of iteration and focus on fine-grained local search in the later stages, thus solving the problem of global-local imbalance in traditional follower search. Multidimensional random constraints: This represents a random number that determines the distance the sparrow moves. This represents a random number that controls the influence of the optimal individual on the sparrow's next position. Equal switching between sine and cosine components injects flexible search diversity into follower position updates; Nonlinear Dependency Adjustment: Introducing a Nonlinear Weighting Factor The dynamic balance of position updates to current position information helps prevent followers from getting trapped in local optima. Where e represents the natural constant; The follower's position is updated using the following formula: In the formula, They represent the first During round iteration, the optimal positions of the base station index and channel index selected by the user's mobile device, the associated base station index and cryptographic algorithm index selected by the user's subtask, the optimal positions occupied by the computing capacity allowed by the user's mobile device, and the optimal positions of the transmission power from the user's mobile device to the associated base station are determined.

[0037] Step S2.4.3: In addition to searching for and capturing prey, sparrows also store excess food for later use. In the improved sparrow search algorithm, this behavior corresponds to the fitness evaluation and selection operation of an individual's location. Specifically, the algorithm compares the individual's fitness value before and after updating its location. If the fitness value of the new location is greater, it accepts that location as the individual's new location; otherwise, it retains the original location.

[0038] The watchdogs are individuals randomly selected from the population. When predators invade, they signal the sparrows to flee to safety. The watchdogs' behavior and location updates are as follows: In the above formula, These represent the optimal user-base station association scheme, the optimal user sub-channel utilization scheme, the optimal subtask execution scheme at the base station, the optimal subtask cryptographic algorithm selection scheme, the optimal user-allowed computational capacity, and the optimal user-to-associated base station transmission power at the t-th iteration, respectively. Let represent the worst user-base station association scheme, the worst user sub-channel utilization scheme, the worst subtask execution scheme at the base station, the worst subtask cryptographic algorithm selection scheme, the allowable computing capacity of the worst user, and the transmission power from the worst user to the associated base station, respectively, in the t-th iteration. This represents the step size coefficient, which follows a standard normal distribution. This indicates the sparrow's current fitness level; and These represent the worst fitness and the best fitness, respectively; when When the sparrow is on the periphery of the population, it indicates that the sparrow is vulnerable to predators; when This indicates that the sparrow is located at the center of the population, is aware of the threat of predators, and needs to immediately turn to other sparrows; , represents a random number; To represent a very small constant, the present invention takes .

[0039] Step S3.1: When completing a round of position update, the boundary function must be revised after the three different types of sparrows update their positions to correct out-of-bounds variables. After the guard's position is updated, the fitness value of the population must be recalculated. After one round of iteration, the fitness value of all sparrow individuals is calculated and compared, and the individual with the highest fitness is regarded as the current best sparrow individual. Step S3.2: Compare the fitness value of the current best sparrow individual with the fitness value of the historical best sparrow individual. If the current best sparrow individual is better, replace the historical best sparrow individual with the current best sparrow individual, update the historical best position variable, and output its position information code as the input of the next stage algorithm. Otherwise, retain the historical best sparrow individual and increment the current iteration count by 1. Step S3.3: Determine the current iteration number Has the maximum number of iterations been exceeded? If the current iteration number Less than or equal to the maximum number of iterations Then continue to explore, discover, follow, and monitor the sparrow population; if the current iteration number is... Greater than the maximum number of iterations Then, it directly outputs the information encoding of all individual sparrows, including position variables and fitness values.

[0040] By iteratively executing the above steps, the improved Sparrow Search Algorithm (ISSA) achieves a good balance between exploration and exploitation, effectively addressing complex optimization problems and gradually approaching the global optimum. The improved ISSA efficiently searches the solution space, ultimately outputting the optimal individual along with its corresponding position variable and fitness value.

[0041] The effects of the embodiments of this invention can be further illustrated through simulation.

[0042] The network model designed in this invention is set as a cellular cell with a diameter of 1km, containing 31 base stations and 10 users near each base station. Each user carries 3 subtasks. The system bandwidth is 20MHz, the maximum user transmit power is 23dBm, the maximum computing power of both base stations and user terminals is 20GHz, and the cache size is 1GB. The sparrow population size is 20, with discoverers, followers, and vigilant sparrows accounting for 40%, 40%, and 20%, respectively. A security threshold is set during the location update of the discoverer. Set it to 0.85, and the maximum number of iterations to 5000.

[0043] Figure 2This invention reveals simulation results showing the impact of sparrow search algorithms on total energy consumption under varying user device density. User density was gradually increased from 10 to 30, primarily comparing three algorithms: Sparrow Search Algorithm (SSA), New Sparrow Search Algorithm (NewSSA), and Improved Sparrow Search Algorithm (ISSA). NewSSA only introduces a nonlinear weighting factor and a sine / cosine search algorithm for followers, without improving the position of the discoverer. The total energy consumption curves show that as density increases, the energy consumption of all three algorithms increases. However, the ISSA curve consistently remains at the bottom. This is because ISSA achieves a better balance between global exploration and local exploitation through exponential decay search of the discoverer, fine-tuning of the follower's sine / cosine curve, and random jumping mechanism of the watcher, thus finding a lower-power offloading strategy.

[0044] Figure 3 This invention reveals the simulation results of the Sparrow Search algorithm on the fitness function under varying iteration counts. The number of iterations was varied from 1 to 5000, and three algorithms—SSA, NewSSA, and ISSA—were tested. The results show that the fitness value of the Sparrow Search algorithm gradually increases, but the ISSA curve remains at the top, exhibiting a better fitness value than the other two algorithms. This demonstrates its advantage of maintaining lower energy consumption and higher resource utilization even under high-load scenarios.

[0045] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A secure collaborative offloading method in ultra-dense millimeter-wave buffer networks, characterized in that, The steps are as follows: Step S1: Obtain the basic network information for secure collaborative offloading in the ultra-dense millimeter-wave caching network and construct the network architecture. The network architecture includes a system model, a communication model, a caching and computing model, and a security model. Configure the optimization problem under multiple constraints in the network architecture. Step S2: Randomly generate an initial solution set based on the optimization problem, and define the initial solution set as the original population. Calculate the fitness value, and take the individual with the highest fitness as the best initial individual. Improve the sparrow algorithm by randomly perturbing the number of iterations, introducing control parameters, nonlinear weighting factors, and sine and cosine search algorithms. The improved sparrow algorithm searches and iterates on the original population to obtain the target population, and outputs the position information encoding of all individuals in the target population. Step S3: Based on the location information encoding of all individuals in the target population, the improved sparrow algorithm is used to further search, continuously update the location of all individuals, select the better individual between the current best individual and the initial best individual as the historical best individual, and output the historical best position information of the historical best individual. Step S4: Perform secure collaborative unloading of resource configurations based on the historical best position information of the historical best individual in the target population.

2. The secure collaborative offloading method in an ultra-dense millimeter-wave buffer network according to claim 1, characterized in that, The detailed process of step S2 is as follows: Population initialization is performed based on the optimization problem. The initialization process includes generating a set of random candidate solutions as the initial solution set within the constraints of the optimization problem, defined as the original population. The optimal individual for population initialization is found based on the fitness value. The algorithm iteration process continues: a safety threshold is set, and the sparrow search algorithm is improved by randomly perturbing the number of iterations, introducing control parameters, nonlinear weighting factors, and using sine and cosine search algorithms. The improved sparrow search algorithm consists of three types of sparrow individuals: discoverers, followers, and vigilants; discoverers, followers, and vigilants are obtained by sorting them in descending order of fitness value. The discoverer is responsible for conducting a broad search in the solution space to explore potential optimal regions. Its position update strategy is adjusted based on the environmental safety level. When the alarm value is less than the safety threshold, the discoverer uses an exponential decay strategy for a large-scale search; when the alarm value is greater than or equal to the safety threshold, random perturbations are introduced to cause position jumps, simulating the behavior of sparrows fleeing dangerous areas. Followers update their positions based on information provided by the discoverer, simulating the foraging behavior of sparrows in a flock. The follower position update strategy incorporates control parameters, nonlinear weighting factors, and sine and cosine search algorithms. By adjusting angle variables and dynamic factors, an adaptive balance between global exploration and local exploitation is achieved during the follower search process. The watcher is responsible for monitoring the population status. The watcher's position update strategy differentiates based on the individual's current fitness: when an individual is on the edge, it moves towards the current optimal region to avoid danger; when an individual is in the center, it randomly moves towards its neighborhood. Finally, the position information encoding of all individuals in the target population is output.

3. The secure collaborative offloading method in an ultra-dense millimeter-wave buffer network according to claim 2, characterized in that, The detailed process of step S3 is as follows: Based on the location information encoding of all individuals in the target population, first determine whether the number of iterations exceeds the maximum limit. If it does not exceed the maximum limit, perform the discoverer search, follower follow, and vigilant monitoring behaviors to update the location information. After updating the locations of the three types of sparrow individuals, use a boundary function to correct out-of-bounds variables, recalculate the population fitness value, compare the new location fitness with the original location fitness, and retain the individual with the higher fitness value as the next generation individual. If the maximum limit of the number of iterations is exceeded, terminate the iteration and output the location encoding of all individuals. After the iteration ends, select the better individual between the current best individual and the initial best individual as the historical best individual and output the relevant location information.

4. The secure collaborative offloading method in an ultra-dense millimeter-wave buffer network according to claim 3, characterized in that, In step S1, obtaining basic network information includes: in an ultra-dense millimeter-wave caching network, express A set of indexes for each user. ; express A set of indexes for each base station. ; express A set of indices for each subtask. ; express A set of indices for cryptographic algorithms. ; Get users The computational task of quadruples ,in, Indicates user Calculate the data size of the subtask. Indicates user Number of CPU cycles required to complete the computational subtask Indicates user Calculate the maximum execution time of the subtask. Indicates user subtasks The expected security level is determined; next, the physical locations of all users are obtained, and the K-means clustering algorithm is used to divide all users into groups based on their physical locations. Clusters, among which, users The cluster it belongs to is denoted as Obtain the total bandwidth of the network system. Sub-channel bandwidth ,user With base station Channel gain between and noise power Subsequently, user acquisition Maximum computing resources CPU energy coefficient and base stations Maximum computing resources Finally, acquire users. The set of base stations associated with all subtasks ,in, Indicates user subtasks The associated base station index number, Indicates user The total number of subtasks, m=1,2,...,M.

5. The secure collaborative offloading method in an ultra-dense millimeter-wave buffer network according to claim 4, characterized in that, The expression for the optimization problem generated in step S1 is as follows: In the above formula, E(·) represents the objective function to be optimized, and X represents the user's association matrix. , Indicates user With base station The correlation index; Z represents the sub-channel selection matrix, , Indicates user Whether to select a sub-channel, where, Indicates each cluster A set of sub-channel indexes U represents the association matrix between the user's subtasks and the base station. , Indicates user subtasks At most one base station Interrelated execution; V represents the choice matrix of the cryptographic algorithm. , Indicates user subtasks Choose a cryptographic algorithm F represents the set of computing capacities that the user is allowed to use. , Indicates user Allowable computational capacity; P represents the transmit power allocation matrix. , Indicates user The transmit power to the associated base station; Indicates user subtasks Processing latency The maximum execution time of this task cannot be exceeded. Constraints; Indicates user Constraint that it can only be associated with a single base station; Indicates user Constraints for selecting a sub-channel within the user's cluster; Indicates user Task At most one base station Constraints enforced above; Indicates user subtasks The constraint that only one cryptographic algorithm can be chosen; Indicates user Overall security vulnerability cost The maximum allowable cost shall not be exceeded. Constraints; Indicates user Constraints on the minimum and maximum allowable computing power for offloading to associated base stations. Indicates user Maximum allowed capacity To represent a sufficiently small value, Indicates user The constraints on the lower and upper limits of the transmit power offloaded to the relevant base station. Indicates the user's maximum allowed power. Indicates user When associated with a single base station, Otherwise, the constraint condition is 0. This indicates that when subchannel s is selected, Otherwise, the constraint condition is 0. Indicates user subtasks When executed on a base station Otherwise, the constraint condition is 0. Indicates the choice of cryptographic algorithm hour, Otherwise, the constraint condition is 0; Indicates user Local processing energy consumption.

6. The secure collaborative offloading method in an ultra-dense millimeter-wave buffer network according to claim 5, characterized in that, The specific steps of step S2 are as follows: Step S2.1: Initialize the maximum number of iterations using the improved sparrow search algorithm. and set the current iteration number Set to 1; Step S2.2: The number of individual sparrows in the sparrow population is Any individual sparrow in a sparrow population is used To indicate; The improved sparrow search algorithm was used to analyze the sparrow population. Each individual sparrow Encoded into position vectors , Encoded into position vectors , Encoded into position vectors , Encoded into position vectors , Encoded into position vectors , Encoded into position vectors ,in, Represents all users The number of virtual users for all subtasks m The set of indices, Represents individual sparrows Chinese users The index number of the associated base station, Represents individual sparrows Chinese users The selected channel index number, Represents individual sparrows Chinese users The number of index numbers associated with the base station, Represents individual sparrows Chinese users The selected cryptographic algorithm index number, Represents individual sparrows Chinese users Computing capacity, Represents individual sparrows Chinese users The transmission power; Step S2.3: Initialize the sparrow population and construct the individual sparrows in the population. fitness function : In the improved sparrow search algorithm, the optimization problem is... and Two constraints are transformed into penalty terms embedded in fitness evaluation; any sparrow individual The fitness function is expressed as: in, Indicates user The penalty factor; Indicates user subtasks The penalty factor; The fitness function is used to evaluate the fitness values ​​of all individuals in the sparrow population. The individuals are sorted in descending order of fitness value, and the sparrow with the largest fitness value is recorded as the best sparrow and the sparrow with the smallest fitness value as the worst sparrow. The fitness values ​​are also recorded. Step S2.4: The position of each individual sparrow is considered a valid solution. Within the solution space, multiple sparrows continuously update their positions by simulating the behaviors of discoverers, followers, and watchers; then the process begins iteratively, proceeding through the iteration number... Less than or equal to the maximum number of iterations At that time, sparrow populations were explored, followed, and monitored to simulate different foraging behaviors and update their locations.

7. The secure collaborative offloading method in an ultra-dense millimeter-wave buffer network according to claim 6, characterized in that, When a sparrow acts as the discoverer, the position update process is represented as follows: A group in Navigation in 3D search space Only sparrows; the position of each sparrow, in this... Represented in the 3D search space as In the formula, , Indicates the first Only sparrows Position in the dimensional search space; Indicates the current sparrow's number There are decision variables; among which They represent the first During round iteration, the base station index selected by the user's mobile device, the channel index selected by the user's mobile device, the associated base station index selected by the user's subtask, the cryptographic algorithm index selected by the user's mobile device, the allowable computing capacity of the user's mobile device, and the transmission power from the user's mobile device to the associated base station are all considered. The first During round iteration, the base station index selected by the user's mobile device, the channel index selected by the user's mobile device, the associated base station index selected by the user's subtask, the cryptographic algorithm index selected by the user's mobile device, the allowable computing capacity of the user's mobile device, and the transmission power from the user's mobile device to the associated base station are all considered. Represents a uniformly random number between 0 and 1; Represents a random number that follows a standard normal distribution; This represents a one-dimensional matrix where each element has a value of 1. Indicates the alarm value, in Variation within the range, Indicates the safety threshold, which is limited to Within the range; when At that time, the surrounding environment was safe, and the discoverer needed to conduct extensive exploration to guide the sparrows to more food; when Immediately release danger signals to prompt the population to quickly adopt anti-predation behaviors, adjust search strategies, and move rapidly to a safe area; This indicates rounding the result down to the nearest integer; exp() represents the exponential function. To improve the effectiveness of the improved sparrow search algorithm, when the alarm value... Update the location of the discoverer as needed: At the start of the iteration, The distribution follows a Cauchy distribution; in the middle stage of the iteration, The distribution lies between the Cauchy and Gaussian distributions; in later iterations, The distribution follows a Gaussian distribution. The discoverer can then search for the region where the current solution is close to the optimal solution of the problem. trnd(·) represents a random perturbation function, which generates random values ​​based on the t-distribution and is used to add random fluctuations to the update of the discoverer's position-related parameters.

8. The secure collaborative offloading method in an ultra-dense millimeter-wave buffer network according to claim 7, characterized in that, When a sparrow acts as a follower, the position update process is represented as follows: The update formula for followers is optimized by introducing control parameters, nonlinear weighting factors, and sine and cosine search algorithms. Dynamic amplitude control: Introducing control parameters The control parameters consist of the iteration number t and the adjustment factor. Decide, Adaptively adjust the amplitude of the sine and cosine functions; Multidimensional random constraints: This represents a random number that determines the distance the sparrow moves. This represents a random number that controls the influence of the optimal individual on the sparrow's next position. It switches equally between sine and cosine components; Nonlinear Dependency Adjustment: Introducing a Nonlinear Weighting Factor The dynamic balance of position updates to current position information helps prevent followers from getting trapped in local optima. Where e represents the natural constant; The follower's position is updated using the following formula: In the formula, They represent the first During round iteration, the optimal positions of the base station index and channel index selected by the user's mobile device, the associated base station index and cryptographic algorithm index selected by the user's subtask, the optimal positions occupied by the computing capacity allowed by the user's mobile device, and the optimal positions of the transmission power from the user's mobile device to the associated base station are determined.

9. The secure collaborative offloading method in an ultra-dense millimeter-wave buffer network according to claim 8, characterized in that, The specific process of step S3 is as follows: Step S3.1: When completing a round of position update, the boundary function must be revised after the three different types of sparrows update their positions to correct out-of-bounds variables. After the guard's position is updated, the fitness value of the population must be recalculated. After one round of iteration, the fitness value of all sparrow individuals is calculated and compared, and the individual with the highest fitness is regarded as the current best sparrow individual. Step S3.2: Compare the fitness value of the current best sparrow individual with the fitness value of the historical best sparrow individual. If the current best sparrow individual is better, replace the historical best sparrow individual with the current best sparrow individual, update the historical best position variable, and output its position information code as the input of the next stage algorithm. Otherwise, retain the historical best sparrow individual and increment the current iteration count by 1. Step S3.3: Determine the current iteration number Has the maximum number of iterations been exceeded? If the current iteration number Less than or equal to the maximum number of iterations Then continue to explore, discover, follow, and monitor the sparrow population; if the current iteration number is... Greater than the maximum number of iterations Then, it directly outputs the information encoding of all individual sparrows, including position variables and fitness values.

Images