A method and system for intelligent network resource allocation based on 5G mobile communication

Abstract

The present invention relates to the technical field of wireless resource allocation, and discloses an intelligent network resource allocation method and system based on 5G mobile communications. The present invention first calculates a theoretical effective network transmission rate; then trains a support vector machine and uses a traveling walking optimization algorithm to adjust and optimize the parameters in the support vector machine to predict the network transmission rate based on the support vector machine model, thereby obtaining a predicted effective network transmission rate; then, the actual effective network transmission rate is obtained from the theoretical effective network transmission rate and the predicted effective network transmission rate; the actual effective network transmission rate and the fairness index are used as indicators for screening users requesting network resources, network resources are allocated and scheduled, and network resource allocation and scheduling priorities are divided based on an Arctic Puffin optimization algorithm to complete network resource allocation and scheduling; finally, a network resource channel efficacy test is performed on a network performance test matrix, and the allocated and scheduled network resources are replaced.

Description

Technical Field

[0001] The present invention relates to the technical field of wireless resource allocation, and specifically to a method and system for intelligent network resource allocation based on 5G mobile communications. Background Art

[0002] Chinese patent CN113163498B discloses a method and device for allocating virtual network resources based on a genetic algorithm under 5G network slicing. The method specifically includes obtaining the resource scale required by the virtual network request node in the virtual network request set, arranging the virtual network request set in descending order according to the resource scale required by the virtual network request node, and obtaining the virtual network request set after sorting; using a genetic algorithm to obtain a virtual network resource allocation strategy set for the virtual network request set obtained after sorting, and realizing virtual network resource allocation by retaining the individual with the largest fitness in the population as an element in the virtual network resource allocation strategy set using the genetic algorithm; the device specifically includes an acquisition module, a calculation module, a sorting module and an allocation module. The network resources and the virtual network request set are obtained by the acquisition module, the number of elements in the virtual network request set is obtained by the calculation module, the node resources are sorted by the sorting module, and the virtual network resource allocation strategy set is obtained by the allocation module.

[0003] The continuous development of 5G communication networks has put forward higher requirements for network resource allocation. Traditional network resource allocation has problems such as unfair resource allocation, network environment redundancy and slow resource allocation speed. At the same time, traditional resource allocation based on 5G communication networks cannot improve the overall performance of the network, and there is no matching resource allocation optimization algorithm. Performance testing is not performed after network resource allocation, resulting in low network resource utilization. Summary of the Invention

[0004] In response to the problems in the related technology, the present invention provides an intelligent network resource allocation method and system based on 5G mobile communication to overcome the above-mentioned technical problems existing in the existing related technology.

[0005] To solve the above technical problems, the present invention is achieved through the following technical solutions:

[0006] The present invention is a method for intelligent network resource allocation based on 5G mobile communications, comprising the following steps:

[0007] S1. Number the 5G base stations and users respectively to generate a network communication matrix, calculate the network data transmission rate of the network communication matrix, and obtain the theoretical effective network transmission rate;

[0008] S2. Set a time series matrix, perform spatial reconstruction on the time series matrix to obtain a network transmission data matrix, establish a support vector machine mathematical model, and use the travel hiking optimization algorithm to adjust and optimize the parameters in the support vector machine. After training the support vector machine, obtain a support vector machine model, and predict the network transmission rate based on the support vector machine model to obtain the predicted effective network transmission rate;

[0009] S3. Calculate the actual effective network transmission rate from the theoretical effective network transmission rate and the predicted effective network transmission rate. Use the actual effective network transmission rate and the fairness index as indicators for screening users requesting network resources. A proportional fair scheduling algorithm is used to allocate and schedule network resources. The network resource allocation and scheduling is optimized based on the Arctic Puffin Optimization Algorithm to find the user with the highest priority for network resource allocation and scheduling, completing the network resource allocation and scheduling.

[0010] S4. Conduct a network resource channel efficacy test on the users who have completed the network resource allocation and scheduling, and evaluate the network resource channel efficacy.

[0011] This invention obtains a network communication matrix by numbering 5G base stations and users, establishes a channel model of the network communication system through the network communication matrix, and calculates the theoretical effective network transmission rate; secondly, the time series matrix is ​​spatially reconstructed to improve the prediction accuracy of the network transmission rate, and the network transmission data of previous years is input into the support vector machine for training to obtain a support vector machine model, and the penalty parameter and kernel width in the support vector machine are adjusted and optimized using the travel hiking optimization algorithm. The algorithm simulates the hiker's behavior by updating the hiker's speed to obtain the position, has high execution efficiency and fast convergence speed, and obtains the predicted effective network transmission rate; then, according to the theoretical effective network transmission rate and The effective network transmission rate is predicted to obtain the actual effective network transmission rate, and combined with the system throughput to obtain the selected network resource users; the sum of the actual effective network transmission rate and the fairness index is used as the priority judgment basis, and the Arctic Puffin optimization algorithm is used to divide high-priority users into low-priority users, and the highest-priority user is found among the high-priority users. The algorithm imitates the aerial flight and underwater foraging behavior of the Arctic Puffin, constantly changing its position to obtain the optimal solution, and introduces Lévy flight, which has strong search capabilities and good iterative performance; then, network resource channels are allocated according to priority to complete network resource allocation and scheduling; finally, the network resource channel efficacy test is conducted, and the allocated and scheduled network resources are replaced.

[0012] Preferably, the S1 comprises the following steps:

[0013] S11, set the 5G base station numbers in sequence to form a 5G base station set A = {a1, a2, a3, ..., a i}, where a iRepresenting a 5G base station numbered i, numbering the users served by the 5G base station, setting the number of users served by the 5G base station to j, the network communication matrix B is generated as follows:

[0014]

[0015] Among them, a ij Indicates the user with 5G base station number i and user number j;

[0016] Select a row of the network communication matrix as the network test set in Indicates that the 5G base station number is For a user with user ID j, the small-scale fading is set to b, the path loss coefficient is set to α, and the user path loss C of the network test set is calculated as follows:

[0017] C = α + 10·log2b;

[0018] Assume that the network allocation scheduling time is t, the network frequency response at time t is b′(t), the network noise power density is c, the network carrier bandwidth is c′, and the network uniform power distribution is The network interference coefficient is β, and the instantaneous signal-to-noise ratio C′(t) of the 5G base stations in the 5G base station set at time t is calculated as follows:

[0019]

[0020] S12. Assume that there are b″ subcarriers in the 5G base station set. The network data transmission rate C″ of the user in the network test set at the 5G base station in the 5G base station set within time t is calculated as follows:

[0021] C″=b″·C′(t);

[0022] Since there are errors in network transmission, the error coefficient is set to χ. The error coefficient is 1 when the network data transmission is successful and 0 when the network data transmission is unsuccessful. The theoretical effective network data transmission rate of the users in the network test set at the 5G base stations in the 5G base station set within time t is recorded as the theoretical effective network transmission rate C″′. The calculation formula is as follows:

[0023] C″′=b″·C′(t)·χ;

[0024] This invention obtains a network communication matrix by numbering 5G base stations and users, establishes a channel model of the network communication system through the network communication matrix, obtains the data transmission rate by calculating the path loss and signal-to-noise ratio, and obtains the theoretical effective network transmission rate after removing the error.

[0025] Preferably, said S2 comprises the following steps:

[0026] S21, introduce time series, set the number of network transmission time series to i′, and obtain the network transmission time series set A2 = {a′1, a′2, a′3, ..., a′ i′}, where a′ i′ Denote the i′th network transmission time series, set the embedding dimension of the network transmission time series in the network transmission time series set to j′, and generate the time series matrix B′ as follows:

[0027]

[0028] Among them, a′ ij′ represents the i′th network transmission time series with embedding dimension j′;

[0029] Select a column in the time series matrix as the network transmission rate set in Indicates the i′th, embedding dimension is The network transmission time series,

[0030] The spatial reconstruction method is used to calculate the standard deviation of the network transmission time series in the network transmission rate set, and then the standard deviation of other network transmission rate sets in the time series matrix is ​​calculated to find the minimum standard deviation d′. The network transmission time series corresponding to the minimum standard deviation is the optimal delay time, which is recorded as d. The optimal embedding window is calculated from the optimal delay time, which is recorded as i″. The optimal embedding dimension is The optimal embedding dimension is used as the dimension of the network transmission data in the network transmission data set; the number of network transmission data is set to i", and a network transmission data set is obtained, the dimension of the network transmission data in the network transmission data set is j", and the network transmission data matrix B" is constructed as follows:

[0031]

[0032] Among them, a′ i″j″ Represents the i″th network transmission data with dimension j″;

[0033] S22: Select the network transmission data of previous years as the support vector machine sample set, denoted as A4 = {a″1, a″2, a″3, ..., a″ e}, where a″ e Represents the e-th network transmission data in previous years, and divides the support vector machine sample set into a support vector machine training set and a support vector machine test set; the output value is set to e′, the actual value is e″, the penalty parameter of the support vector machine is δ, the kernel width is e, and the mathematical model of the support vector machine is established with the following calculation formula:

[0034]

[0035] Among them, minC(δ,e) represents the mathematical model of support vector machine, represents the i1th output value, represents the actual value of the i1th, i1=1,2,3,...,e;

[0036] The vector machine training set is normalized to obtain a normalized vector machine training set, and the normalized vector machine training set is input into the support vector machine. The maximum number of iterations and the current number of iterations are set. When the current number of iterations of the support vector machine is equal to the maximum number of iterations, the iteration is stopped to obtain a trained support vector machine. The support vector machine test set is then input into the trained support vector machine, and the accuracy threshold is set to ω. When the absolute value of the difference between the output value of the trained support vector machine and the actual value of the vector machine training set is less than ω, a support vector machine model is obtained. Otherwise, the penalty parameter and kernel width of the support vector machine are adjusted and optimized until the absolute value of the difference between the output value and the actual value is less than ω. The specific process is as follows:

[0037] S221, using the support vector machine mathematical model as the fitness function, setting a group of travelers, during which the travelers are hiking, setting the altitude of the travelers' hiking as g', the horizontal distance of the travelers' hiking as f', and the terrain inclination angle at the g-th iteration as θ g And θ g ∈[0,50°], then the calculation formula for the traveler's walking slope at the g-th iteration is as follows:

[0038]

[0039] Set the traveler's walking speed at the gth iteration to v g , calculate the fitness function value of the g-th iteration. The formula for calculating the traveler's walking speed at the g-th iteration is as follows:

[0040]

[0041] S222: There is a traveler leader in the traveler group, and the walking speed of the travelers in the g-1th iteration is set to v g-1 , at the gth iteration, the leader of the travelers is located at x g , the traveler's position at the gth iteration is x′ g , the traveler follows the leader to generate a scanning factor ε and ε∈[1,2], φ represents a random number and φ∈[0,1], and then begins to iteratively update the fitness function value. At the g-th iteration, the traveler's walking speed v g The update calculation formula is as follows:

[0042] vg =v g-1 +(x g -ε·x′ g )·φ;

[0043] The traveler follows the leader and receives the signal from the leader to iterate and update the traveler's position. The leader's position is the minimum fitness function value. At this time, the minimum fitness function value is continuously searched, and the fitness function value of the g+1th iteration is used to replace the fitness function value of the gth iteration. The traveler's walking position at the g+1th iteration is set to x' g+1 , the calculation formula for the traveler's walking position at the g+1th iteration is as follows,

[0044] x′ g+1 =x′ g +f g ;

[0045] set up It represents a uniformly distributed number in [0,1]. The upper bound of the initial traveler position is f″, and the lower bound of the initial traveler position is / f″′. Then the initial position of the traveler x′ satisfies the following formula:

[0046]

[0047] The initial position of the traveler is continuously iterated according to the walking speed of the traveler, and the minimum fitness function value is continuously updated. The maximum number of iterations is set to C' and the current number of iterations is set to g. When the current number of iterations is equal to the maximum number of iterations, the iteration is stopped to obtain the position of the traveler leader. The rectangular coordinate values ​​of the traveler leader position correspond to the penalty parameter and kernel width of the support vector machine respectively.

[0048] S23, select a column of the network transmission data matrix as a sample data of the network transmission data test set, and select the network transmission data test set A5={(a″ 11 ,a″ 21 ,a″ 31 ,...,a″ i″1 ),(a″ 12 ,a″ 22 ,a″ 32 ,...,a″ i″2 ),...,(a″ 1j″ ,a″ 2j ″,a″ 3j″ ,...,a″ i″j″ )}, input the network transmission data test set into the support vector machine model to obtain the predicted effective network transmission rate.

[0049] This invention trains a support vector machine by inputting network transmission data from previous years into the machine, and uses a travel hiking optimization algorithm to adjust and optimize the penalty parameter and kernel width parameters in the support vector machine. The algorithm simulates the hiker's behavior by updating the hiker's speed to obtain the position, has high execution efficiency and fast convergence speed, and obtains a support vector machine model to predict the network transmission rate and obtain the predicted effective network transmission rate.

[0050] Preferably, the step S3 includes the following steps:

[0051] S31. Set a first threshold value as ω1. If the absolute value of the difference between the theoretical effective network transmission rate and the predicted effective network transmission rate is less than ω1, the theoretical effective network transmission rate and the predicted effective network transmission rate meet the requirements; otherwise, recalculate the theoretical effective network transmission rate and the predicted effective network transmission rate; and take the average of the theoretical effective network transmission rate and the predicted effective network transmission rate as the actual effective network transmission rate;

[0052] Select the users who request network resources to obtain the user set D = {g1, g2, g3, ..., g h}, where g h Denote the hth user requesting network resources, and calculate the actual effective network transmission rate v and system throughput w of the users requesting network resources in the set of users requesting network resources in turn; set the average response time of requesting network resources to be The system throughput The actual effective network transmission rate and system throughput are used as network resource channel evaluation indicators; an actual effective network transmission rate set and a system throughput set are obtained, and a network transmission rate threshold ξ and a system throughput threshold ψ are set. When the actual effective network transmission rate in the actual effective network transmission rate set is greater than ξ and the system throughput in the system throughput set is greater than ψ, the requesting network resource users corresponding to the actual effective network transmission rate in the actual effective network transmission rate set and the system throughput in the system throughput set are retained; otherwise, the requesting network resource users corresponding to the actual effective network transmission rate in the actual effective network transmission rate set and the system throughput in the system throughput set are deleted, and a filtered network resource user set D′={g′1, g′2, g′3, ..., g′ h′}, where g′ h′ represents the h′th filtered user requesting network resources;

[0053] S32: Set the bandwidth satisfaction of the filtered network resource users in the filtered network resource user set as C″, and the fairness coefficient as γ. Then, the fairness index calculation formula is as follows:

[0054]

[0055] Among them, E represents the fairness index, represents the bandwidth satisfaction of the i2th filtered network resource user in the filtered network resource user set, i2=1,2,3,...,h′;

[0056] A fairness index set is obtained, and a network resource scheduler receives the actual effective network transmission rate set and the fairness index set, calculates the sum of the actual effective network transmission rates in the actual effective network transmission rate set and the fairness indexes in the fairness index set, prioritizes the screened network resource requesting users according to the sum of the actual effective network transmission rates in the actual effective network transmission rate set and the fairness indexes in the fairness index set, and finds the user with the highest network resource allocation scheduling priority. The specific process is as follows:

[0057] S321. The sum of the actual effective network transmission rate in the actual effective network transmission rate set and the fairness index in the fairness index set is used as the fitness function. The number of Arctic puffin populations is set to h′, the upper bound of the k-th Arctic puffin in the Arctic puffin population is set to l, the lower bound of the k-th Arctic puffin in the Arctic puffin population is set to l′, p1 represents a random number and p1∈[0,1], then the initial position y of the k-th Arctic puffin in the Arctic puffin population is set to k The calculation formula is as follows,

[0058] y k =l′+(ll′)·p1;

[0059] When the Arctic puffin is flying and searching for food, it starts to iterate and find the fitness function value. The current number of iterations is set to q. The kth Arctic puffin in the Arctic puffin population has the qth iteration position. Introduce the Lévy flight, denoted as E′, and the random number generated by the Lévy flight is p2. The position of the Arctic puffin in the qth iteration is The kth Arctic puffin's position in the Arctic puffin population at the q+1th iteration The calculation formula is as follows,

[0060]

[0061] S322. In the process of hunting for food, the Arctic puffin continuously iterates to improve its ability to find the value of the fitness function. The speed coefficient η is introduced to adjust the position of the k-th Arctic puffin in the Arctic puffin population to increase the success rate of hunting food. Set p3 to represent a random number and p3∈[0,1]. Then the k-th Arctic puffin in the Arctic puffin population will be iterated again for the q+1th time. as follows,

[0062]

[0063] After the fitness function values ​​of the Arctic puffin population are found, the fitness function values ​​of the Arctic puffin population are sorted from small to large, and the Arctic puffin corresponding to the fitness function values ​​of the first h" Arctic puffin populations are selected as the new Arctic puffin population. The new Arctic puffin population is the user with high priority. The kth Arctic puffin in the new Arctic puffin population has the q+1th iteration position. as follows,

[0064]

[0065] S323, when the Arctic puffin is hunting for food underwater, the k1th Arctic puffin in the new Arctic puffin population is set to the qth iteration position The k2th Arctic puffin's qth iteration position in the new Arctic puffin population is The k3th Arctic puffin in the new Arctic puffin population has the qth iteration position The cooperation factor is λ, p4 represents a random number and p4∈[0,1]. At this time, the maximum fitness function value is continuously iterated. The kth Arctic puffin in the new Arctic puffin population updates its position in the q+1th iteration. The calculation formula is as follows,

[0066]

[0067] As the Arctic puffins intensify their hunting for food underwater, the Arctic puffins in the new Arctic puffin population change their positions to find more food. At this time, the iteration is accelerated to find the maximum fitness function value. The maximum number of iterations is set to Q, the adaptive factor is μ, p5 represents a random number and p5∈[0,1]. The kth Arctic puffin in the new Arctic puffin population updates its position in the q+1th iteration. The calculation formula is as follows,

[0068]

[0069] S324, in the process of escaping predators, the Arctic puffin escapes the local optimum, and μ is set to represent a random number uniformly distributed between 0 and 1. The kth Arctic puffin in the new Arctic puffin population is used as the q+1th iteration position. The fitness function value of replaces the kth Arctic puffin in the new Arctic puffin population and updates the position in the qth iteration The fitness function value is calculated as follows:

[0070]

[0071] Continuously iterate to obtain the final position of the kth Arctic puffin in the new Arctic puffin population at the q+1th iteration When the current number of iterations equals the maximum number of iterations, the iteration is stopped. The rectangular coordinate values ​​of the k-th Arctic puffin position in the new Arctic puffin population correspond to the actual effective network transmission rate in the actual effective network transmission rate set and the fairness index in the fairness index set, respectively, which are recorded as the best actual effective network transmission rate and the best fairness index. The user with the highest priority corresponding to the best actual effective network transmission rate and the best fairness index is designated as the user with the highest priority for network resource allocation scheduling.

[0072] S33. Set a network resource channel set, allocate the first network resource channel in the network resource channel set to the user with the highest priority in network resource allocation scheduling, delete the first network resource channel, obtain a candidate network resource channel set, and then allocate the candidate network resource channels in the candidate network resource channel set to the user with high priority; when the candidate network resource channel set is empty, end the network resource allocation scheduling; otherwise, continue to perform network resource allocation scheduling until the candidate network resource channel set is empty.

[0073] This invention uses the sum of the actual effective network transmission rate and the fairness index as the fitness function, and divides users into high-priority users and low-priority users based on the Arctic Puffin optimization algorithm, and finds the highest-priority user among the high-priority users. The algorithm imitates the Arctic Puffin's aerial flight and underwater foraging behavior, constantly changes its position to obtain the optimal solution, introduces Levy flight, has strong search capabilities and good iterative performance; then, it allocates network resource channels according to priority, completes network resource allocation and scheduling; finally, it conducts a network resource channel efficacy test, and replaces the allocated and scheduled network resources.

[0074] Preferably, the S4 comprises the following steps:

[0075] S41. Set the number of network resource channels to m. The network resource channel parameters include network bandwidth, network delay, network jitter, and network packet loss rate. When a user who has completed network resource allocation and scheduling uses the network resource channel, measure the network bandwidth, network delay, network jitter, and network packet loss rate at n moments to generate a network performance test matrix B″′ as follows:

[0076]

[0077] Among them, r mn represents the network bandwidth of the mth network resource channel at time n, s mn represents the network delay of the mth network resource channel at time n, u mn represents the network jitter of the mth network resource channel at time n, w mn represents the network packet loss rate of the mth network resource channel at time n;

[0078] S42. Calculate the efficacy scores of the sample data in the network performance test matrix to obtain an efficacy score matrix, perform weighted averaging on each row of sample data in the efficacy score matrix to obtain a channel index, set the index threshold to σ, and when the channel index is greater than σ, replace the network resource channel corresponding to the channel index and replace the allocated and scheduled network resources; otherwise, do not replace the network resource channel corresponding to the channel index.

[0079] This embodiment also discloses an intelligent network resource allocation system based on 5G mobile communications, which specifically includes: a theoretical effective network transmission rate module, a predicted effective network transmission rate module, a network resource allocation scheduling module, and a network resource channel efficacy testing module;

[0080] The theoretical effective network transmission rate module is used to establish a channel model of the network communication system to calculate the theoretical effective network transmission rate;

[0081] The effective network transmission rate prediction module is used to construct a support vector machine mathematical model to realize network transmission rate prediction and adjust the optimization parameters using a travel walking optimization algorithm;

[0082] The network resource allocation and scheduling module is used to divide priorities and find the highest priority user through the Arctic Puffin optimization algorithm, and perform network resource allocation and scheduling on the users in order of priority;

[0083] The network resource channel efficacy test module is used to evaluate the network resource channel efficacy and replace the allocated and scheduled network resources.

[0084] The present invention has the following beneficial effects:

[0085] 1. This invention obtains a network communication matrix by numbering 5G base stations and users, establishes a channel model of the network communication system through the network communication matrix, obtains the data transmission rate by calculating the path loss and signal-to-noise ratio, and obtains the theoretical effective network transmission rate after removing the error.

[0086] 2. The invention inputs the network transmission data of previous years into the support vector machine for training, and uses the traveling hiking optimization algorithm to adjust and optimize the penalty parameter and kernel width of the support vector machine to obtain a support vector machine model, thereby realizing the prediction of the network transmission rate and obtaining the predicted effective network transmission rate. The algorithm has high execution efficiency and fast convergence speed.

[0087] 3. This invention divides users into high-priority users and low-priority users based on the Arctic Puffin optimization algorithm, and finds the highest-priority user among the high-priority users. The algorithm introduces Levy flight, which has strong search capabilities and good iterative performance; then, network resource channels are allocated according to priority to complete network resource allocation and scheduling; finally, the network resource channel efficacy test is performed, and the allocated and scheduled network resources are replaced.

[0088] Of course, any product implementing the present invention does not necessarily need to achieve all of the advantages described above at the same time. BRIEF DESCRIPTION OF THE DRAWINGS

[0089] In order to more clearly illustrate the technical solutions of the embodiments of the invention, the following briefly introduces the drawings required for describing the embodiments. Obviously, the drawings described below are only some embodiments of the invention. For ordinary technicians in this field, they can also obtain drawings based on these drawings without paying any creative work.

[0090] Figure 1 The present invention provides a flow chart of intelligent network resource allocation by an intelligent network resource allocation system based on 5G mobile communication. DETAILED DESCRIPTION

[0091] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0092] In the description of the present invention, it should be understood that the terms "opening", "upper", "lower", "top", "middle", "inside" and the like indicating orientation or positional relationship are only for the convenience of describing the invention and simplifying the description, and do not indicate or imply that the components or elements referred to must have a specific orientation, be constructed and operate in a specific orientation, and therefore cannot be understood as limiting the invention.

[0093] This embodiment discloses a method for intelligent network resource allocation based on 5G mobile communications, which specifically includes the following contents:

[0094] S1. Number the 5G base stations and users respectively to generate a network communication matrix, calculate the network data transmission rate of the network communication matrix, and obtain the theoretical effective network transmission rate;

[0095] Said S1 comprises the following steps:

[0096] S11, set the 5G base station numbers in sequence to form a 5G base station set A = {a1, a2, a3, ..., ai}, where a i Representing a 5G base station numbered i, numbering the users served by the 5G base station, setting the number of users served by the 5G base station to j, the network communication matrix B is generated as follows:

[0097]

[0098] Among them, a ij Indicates the user with 5G base station number i and user number j;

[0099] Select a row of the network communication matrix as the network test set in Indicates that the 5G base station number is For a user with user ID j, the small-scale fading is set to b, the path loss coefficient is set to α, and the user path loss C of the network test set is calculated as follows:

[0100] C = α + 10·log2b;

[0101] Assume that the network allocation scheduling time is t, the network frequency response at time t is b′(t), the network noise power density is c, the network carrier bandwidth is c′, and the network uniform power distribution is The network interference coefficient is β, and the instantaneous signal-to-noise ratio C′(t) of the 5G base stations in the 5G base station set at time t is calculated as follows:

[0102]

[0103] S12. Assume that there are b″ subcarriers in the 5G base station set. The network data transmission rate C″ of the user in the network test set at the 5G base station in the 5G base station set within time t is calculated as follows:

[0104] C″=b″·C′(t);

[0105] Since there are errors in network transmission, the error coefficient is set to χ. The error coefficient is 1 when the network data transmission is successful and 0 when the network data transmission is unsuccessful. The theoretical effective network data transmission rate of the users in the network test set at the 5G base stations in the 5G base station set within time t is recorded as the theoretical effective network transmission rate C″′. The calculation formula is as follows:

[0106] C″′=b″·C′(t)·χ;

[0107] S2. Set a time series matrix, perform spatial reconstruction on the time series matrix to obtain a network transmission data matrix, establish a support vector machine mathematical model, and use the travel hiking optimization algorithm to adjust and optimize the parameters in the support vector machine. After training the support vector machine, obtain a support vector machine model, and predict the network transmission rate based on the support vector machine model to obtain the predicted effective network transmission rate;

[0108] The S2 comprises the following steps:

[0109] S21, introduce time series, set the number of network transmission time series to i′, and obtain the network transmission time series set A2 = {a′1, a′2, a′3, ..., a′ i′}, where a′ i′ Denote the i′th network transmission time series, set the embedding dimension of the network transmission time series in the network transmission time series set to j′, and generate the time series matrix B′ as follows:

[0110]

[0111] Among them, a′ ij′ represents the i′th network transmission time series with embedding dimension j′;

[0112] Select a column in the time series matrix as the network transmission rate set in Indicates the i′th, embedding dimension is The network transmission time series is calculated using the spatial reconstruction method. The standard deviation of the network transmission time series in the network transmission rate set is calculated. The standard deviation of other network transmission rate sets in the time series matrix is ​​then calculated to find the minimum standard deviation d′. The network transmission time series corresponding to the minimum standard deviation is the optimal delay time, denoted as d. The optimal embedding window is calculated from the optimal delay time, denoted as i″. The optimal embedding dimension is The optimal embedding dimension is used as the dimension of the network transmission data in the network transmission data set; the number of network transmission data is set to i", and a network transmission data set is obtained, the dimension of the network transmission data in the network transmission data set is j", and the network transmission data matrix B" is constructed as follows:

[0113]

[0114] Among them, a″ i″j″ Represents the i″th network transmission data with dimension j″;

[0115] S22: Select the network transmission data of previous years as the support vector machine sample set, denoted as A4 = {a″1, a″2, a″3, ..., a″ e}, where a″e Represents the e-th network transmission data in previous years, and divides the support vector machine sample set into a support vector machine training set and a support vector machine test set; the output value is set to e′, the actual value is e″, the penalty parameter of the support vector machine is δ, the kernel width is e, and the mathematical model of the support vector machine is established with the following calculation formula:

[0116]

[0117] Among them, minC(δ,e) represents the mathematical model of support vector machine, represents the i1th output value, represents the actual value of the i1th, i1=1,2,3,...,e;

[0118] The vector machine training set is normalized to obtain a normalized vector machine training set, and the normalized vector machine training set is input into the support vector machine. The maximum number of iterations and the current number of iterations are set. When the current number of iterations of the support vector machine is equal to the maximum number of iterations, the iteration is stopped to obtain a trained support vector machine. The support vector machine test set is then input into the trained support vector machine, and the accuracy threshold is set to ω. When the absolute value of the difference between the output value of the trained support vector machine and the actual value of the vector machine training set is less than ω, a support vector machine model is obtained. Otherwise, the penalty parameter and kernel width of the support vector machine are adjusted and optimized until the absolute value of the difference between the output value and the actual value is less than ω. The specific process is as follows:

[0119] S221, using the support vector machine mathematical model as the fitness function, setting a group of travelers, during which the travelers are hiking, setting the altitude of the travelers' hiking as g', the horizontal distance of the travelers' hiking as f', and the terrain inclination angle at the g-th iteration as θ g And θ g ∈[0,50°], then the calculation formula for the traveler's walking slope at the g-th iteration is as follows:

[0120]

[0121] Set the traveler's walking speed at the gth iteration to v g , calculate the fitness function value of the g-th iteration. The formula for calculating the traveler's walking speed at the g-th iteration is as follows:

[0122]

[0123] S222: There is a traveler leader in the traveler group, and the walking speed of the travelers in the g-1th iteration is set to v g-1 , at the gth iteration, the leader of the travelers is located at x g , the traveler's position at the gth iteration is x′ g, the traveler follows the leader to generate a scanning factor ε and ε∈[1,2], φ represents a random number and φ∈[0,1], and then begins to iteratively update the fitness function value. At the g-th iteration, the traveler's walking speed v g The update calculation formula is as follows:

[0124] v g =v g-1 +(x g -ε·x′ g )·φ;

[0125] The traveler follows the leader and receives the signal from the leader to iterate and update the traveler's position. The leader's position is the minimum fitness function value. At this time, the minimum fitness function value is continuously searched, and the fitness function value of the g+1th iteration is used to replace the fitness function value of the gth iteration. The traveler's walking position at the g+1th iteration is set to x' g+1 , the calculation formula for the traveler's walking position at the g+1th iteration is as follows,

[0126] x′ g+1 =x′ g +f g ;

[0127] set up It represents a uniformly distributed number in [0,1]. The upper bound of the initial traveler position is f″, and the lower bound of the initial traveler position is / f″′. Then the initial position of the traveler x′ satisfies the following formula:

[0128]

[0129] The initial position of the traveler is continuously iterated according to the walking speed of the traveler, and the minimum fitness function value is continuously updated. The maximum number of iterations is set to C' and the current number of iterations is set to g. When the current number of iterations is equal to the maximum number of iterations, the iteration is stopped to obtain the position of the traveler leader. The rectangular coordinate values ​​of the traveler leader position correspond to the penalty parameter and kernel width of the support vector machine respectively.

[0130] S23, select a column of the network transmission data matrix as a sample data of the network transmission data test set, and select the network transmission data test set A5={(a″ 11 ,a″ 21 ,a″ 31 ,...,a″ i″1 ),(a″ 12 ,a″ 22 ,a″ 32 ,...,a″ i″2 ),...,(a″ 1j″ ,a″ 2j″ ,a″3j″ ,...,a″ i″j″ )}, inputting the network transmission data test set into a support vector machine model to obtain a predicted effective network transmission rate;

[0131] S3. Calculate the actual effective network transmission rate from the theoretical effective network transmission rate and the predicted effective network transmission rate. Use the actual effective network transmission rate and the fairness index as indicators for screening users requesting network resources. A proportional fairness scheduling algorithm is used to allocate and schedule network resources. The network resource allocation and scheduling is optimized based on the Arctic Puffin Optimization Algorithm to find the user with the highest priority, completing the network resource allocation and scheduling.

[0132] The S3 includes the following steps:

[0133] S31. Set a first threshold value as ω1. If the absolute value of the difference between the theoretical effective network transmission rate and the predicted effective network transmission rate is less than ω1, the theoretical effective network transmission rate and the predicted effective network transmission rate meet the requirements; otherwise, recalculate the theoretical effective network transmission rate and the predicted effective network transmission rate; and take the average of the theoretical effective network transmission rate and the predicted effective network transmission rate that meet the requirements as the actual effective network transmission rate;

[0134] Select the users who request network resources to obtain the user set D = {g1, g2, g3, ..., g h}, where g h Denote the hth user requesting network resources, and calculate the actual effective network transmission rate v and system throughput w of the users requesting network resources in the set of users requesting network resources in turn; set the average response time of requesting network resources to be The system throughput The actual effective network transmission rate and system throughput are used as network resource channel evaluation indicators; an actual effective network transmission rate set and a system throughput set are obtained, and a network transmission rate threshold ξ and a system throughput threshold ψ are set. When the actual effective network transmission rate in the actual effective network transmission rate set is greater than ξ and the system throughput in the system throughput set is greater than ψ, the requesting network resource users corresponding to the actual effective network transmission rate in the actual effective network transmission rate set and the system throughput in the system throughput set are retained; otherwise, the requesting network resource users corresponding to the actual effective network transmission rate in the actual effective network transmission rate set and the system throughput in the system throughput set are deleted, and a filtered network resource user set D′={g1′,g′2,g3′,...,g′ h′}, where g′ h′ represents the h′th filtered user requesting network resources;

[0135] S32: Set the bandwidth satisfaction of the filtered network resource users in the filtered network resource user set as C″, and the fairness coefficient as γ. Then, the fairness index calculation formula is as follows:

[0136]

[0137] Among them, E represents the fairness index, represents the bandwidth satisfaction of the i2th filtered network resource user in the filtered network resource user set, i2=1,2,3,...,h′;

[0138] A fairness index set is obtained, and a network resource scheduler receives the actual effective network transmission rate set and the fairness index set, calculates the sum of the actual effective network transmission rates in the actual effective network transmission rate set and the fairness indexes in the fairness index set, prioritizes the screened network resource requesting users according to the sum of the actual effective network transmission rates in the actual effective network transmission rate set and the fairness indexes in the fairness index set, and finds the user with the highest network resource allocation scheduling priority. The specific process is as follows:

[0139] S321. The sum of the actual effective network transmission rate in the actual effective network transmission rate set and the fairness index in the fairness index set is used as the fitness function. The number of Arctic puffin populations is set to h′, the upper bound of the k-th Arctic puffin in the Arctic puffin population is set to l, the lower bound of the k-th Arctic puffin in the Arctic puffin population is set to l′, p1 represents a random number and p1∈[0,1], then the initial position y of the k-th Arctic puffin in the Arctic puffin population is set to k The calculation formula is as follows,

[0140] y k =l′+(ll′)·p1;

[0141] When the Arctic puffin is flying and searching for food, it starts to iterate and find the fitness function value. The current number of iterations is set to q. The kth Arctic puffin in the Arctic puffin population has the qth iteration position. Introduce the Lévy flight, denoted as E′, and the random number generated by the Lévy flight is p2. The position of the Arctic puffin in the qth iteration is The kth Arctic puffin's position in the Arctic puffin population at the q+1th iteration The calculation formula is as follows,

[0142]

[0143] S322. In the process of hunting for food in flight, the Arctic puffin continuously iterates to improve its ability to find the value of the fitness function. The speed coefficient η is introduced to adjust the position of the k-th Arctic puffin in the Arctic puffin population to increase the success rate of hunting food. Assume that p3 represents a random number and p3∈[0,1], then the k-th Arctic puffin in the Arctic puffin population updates its position in the q+1th iteration. as follows,

[0144]

[0145] After the fitness function values ​​of the Arctic puffin population are found, the fitness function values ​​of the Arctic puffin population are sorted from small to large, and the Arctic puffin corresponding to the fitness function values ​​of the first h" Arctic puffin populations are selected as the new Arctic puffin population. The new Arctic puffin population is the user with high priority. The kth Arctic puffin in the new Arctic puffin population has the q+1th iteration position. as follows,

[0146]

[0147] S323, when the Arctic puffin is hunting for food underwater, the k1th Arctic puffin in the new Arctic puffin population is set to the qth iteration position The k2th Arctic puffin's qth iteration position in the new Arctic puffin population is The k3th Arctic puffin in the new Arctic puffin population has the qth iteration position The cooperation factor is λ, p4 represents a random number and p4∈[0,1]. At this time, the maximum fitness function value is continuously iterated. The kth Arctic puffin in the new Arctic puffin population updates its position in the q+1th iteration. The calculation formula is as follows,

[0148]

[0149] As the Arctic puffins intensify their hunting for food underwater, the Arctic puffins in the new Arctic puffin population change their positions to find more food. At this time, the iteration is accelerated to find the maximum fitness function value. The maximum number of iterations is set to Q, the adaptive factor is μ, p5 represents a random number and p5∈[0,1]. The kth Arctic puffin in the new Arctic puffin population updates its position in the q+1th iteration. The calculation formula is as follows,

[0150]

[0151] S324, in the process of escaping predators, the Arctic puffin escapes the local optimum, and μ is set to represent a random number uniformly distributed between 0 and 1. The kth Arctic puffin in the new Arctic puffin population is used as the q+1th iteration position. The fitness function value of replaces the kth Arctic puffin in the new Arctic puffin population and updates the position in the qth iteration The fitness function value is calculated as follows:

[0152]

[0153] Continuously iterate to obtain the final position of the kth Arctic puffin in the new Arctic puffin population at the q+1th iteration When the current number of iterations equals the maximum number of iterations, the iteration is stopped. The rectangular coordinate values ​​of the k-th Arctic puffin position in the new Arctic puffin population correspond to the actual effective network transmission rate in the actual effective network transmission rate set and the fairness index in the fairness index set, respectively, which are recorded as the best actual effective network transmission rate and the best fairness index. The user with the highest priority corresponding to the best actual effective network transmission rate and the best fairness index is designated as the user with the highest priority for network resource allocation scheduling.

[0154] S33. Set a network resource channel set, allocate a first network resource channel in the network resource channel set to the user with the highest priority in network resource allocation scheduling, delete the first network resource channel, obtain a candidate network resource channel set, and then allocate candidate network resource channels in the candidate network resource channel set to the user with the highest priority; when the candidate network resource channel set is empty, terminate the network resource allocation scheduling; otherwise, continue the network resource allocation scheduling until the candidate network resource channel set is empty;

[0155] S4. Conducting a network resource channel efficacy test on the users who have completed the network resource allocation and scheduling, and evaluating the network resource channel efficacy;

[0156] The S4 comprises the following steps:

[0157] S41. Set the number of network resource channels to m. The network resource channel parameters include network bandwidth, network delay, network jitter, and network packet loss rate. When a user who has completed network resource allocation and scheduling uses the network resource channel, measure the network bandwidth, network delay, network jitter, and network packet loss rate at n moments to generate a network performance test matrix B″′ as follows:

[0158]

[0159] Among them, r mn represents the network bandwidth of the mth network resource channel at time n, s mn represents the network delay of the mth network resource channel at time n, u mn represents the network jitter of the mth network resource channel at time n, w mn represents the network packet loss rate of the mth network resource channel at time n;

[0160] S42. Calculate the efficacy scores of the sample data in the network performance test matrix to obtain an efficacy score matrix, perform weighted averaging on each row of sample data in the efficacy score matrix to obtain a channel index, set the index threshold to σ, and when the channel index is greater than σ, replace the network resource channel corresponding to the channel index and replace the allocated and scheduled network resources; otherwise, do not replace the network resource channel corresponding to the channel index.

[0161] This embodiment also discloses an intelligent network resource allocation system based on 5G mobile communications, which specifically includes: a theoretical effective network transmission rate module, a predicted effective network transmission rate module, a network resource allocation scheduling module, and a network resource channel efficacy testing module;

[0162] The theoretical effective network transmission rate module is used to establish a channel model of the network communication system to calculate the theoretical effective network transmission rate;

[0163] The effective network transmission rate prediction module is used to construct a support vector machine mathematical model to realize network transmission rate prediction and adjust the optimization parameters using a travel walking optimization algorithm;

[0164] The network resource allocation and scheduling module is used to divide priorities and find the highest priority user through the Arctic Puffin optimization algorithm, and perform network resource allocation and scheduling on the users in order of priority;

[0165] The network resource channel efficacy test module is used to evaluate the network resource channel efficacy and replace the allocated and scheduled network resources.

[0166] Throughout this specification, references to terms such as "one embodiment," "example," or "specific example" indicate that the specific features, structures, materials, or characteristics described in conjunction with that embodiment or example are included in at least one embodiment or example of the invention. In this specification, schematic representations of these terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in any one or more embodiments or examples.

[0167] The preferred embodiments of the invention disclosed above are intended only to help illustrate the invention. These preferred embodiments do not exhaust all details, nor do they limit the invention to the specific embodiments described. Obviously, many modifications and variations are possible based on the content of this specification. These embodiments are selected and described in detail in this specification to better explain the principles and practical applications of the invention, thereby enabling those skilled in the art to better understand and utilize the invention.

Claims

1. An intelligent network resource allocation method based on 5G mobile communication, characterized in that: The steps include: S1. Number the 5G base stations and users respectively to generate a network communication matrix, calculate the network data transmission rate of the network communication matrix, and obtain the theoretical effective network transmission rate; S2. Set a time series matrix, spatially reconstruct the time series matrix to obtain a network transmission data matrix, establish a support vector machine mathematical model, adjust and optimize the parameters in the support vector machine, train the support vector machine to obtain a support vector machine model, and predict the network transmission rate based on the support vector machine model to obtain the predicted effective network transmission rate; S3. The actual effective network transmission rate is obtained from the theoretical effective network transmission rate and the predicted effective network transmission rate. The actual effective network transmission rate and the fairness index are used as indicators to screen users requesting network resources. The proportional fair scheduling algorithm is used to allocate and schedule network resources. The network resource allocation and scheduling is optimized to find the user with the highest priority for network resource allocation and scheduling, and the network resource allocation and scheduling is completed. The S3 comprises the following steps: S31. Using the average of the theoretical effective network transmission rate and the predicted effective network transmission rate as the actual effective network transmission rate, and using the actual effective network transmission rate and the system throughput as network resource channel evaluation indicators, screen the users requesting network resources to obtain screened users requesting network resources. S32. Calculate the fairness index of the screened network resource users, prioritize them according to their actual effective network transmission rates and fairness indexes, and find the user with the highest priority for network resource allocation and scheduling; The S32 includes the following steps: S321. The sum of the actual effective network transmission rate and the fairness index is used as the fitness function. The number of Arctic puffin populations is set to h′, the upper bound of the k-th Arctic puffin in the Arctic puffin population is l, the lower bound of the k-th Arctic puffin in the Arctic puffin population is l′, p1 represents a random number and p1∈[0,1], then the initial position y of the k-th Arctic puffin in the Arctic puffin population is k The calculation formula is as follows, y k =l′+(ll′)·p1; When the Arctic puffin is flying and searching for food, it starts to iterate and find the fitness function value. The current number of iterations is set to q. The kth Arctic puffin in the Arctic puffin population has the qth iteration position. Introduce the Lévy flight, denoted as E′, and the random number generated by the Lévy flight is p2. The position of the Arctic puffin in the qth iteration is The kth Arctic puffin's position in the Arctic puffin population at the q+1th iteration The calculation formula is as follows, S322, the Arctic puffin continuously iterates in the process of hunting for food, improving its ability to find the value of the fitness function. The k-th Arctic puffin in the Arctic puffin population iterates again in position q+1. After the fitness function values ​​of the Arctic puffin population are found, the fitness function values ​​of the Arctic puffin population are sorted from small to large, and the Arctic puffin corresponding to the fitness function values ​​of the first h" Arctic puffin populations are selected as the new Arctic puffin population. The new Arctic puffin population is the user with high priority. The kth Arctic puffin in the new Arctic puffin population has the q+1th iteration position. as follows, S323, when the Arctic puffin is hunting for food underwater, it continuously iterates to find the maximum fitness function value. When the Arctic puffin is escaping from predators, it continuously iterates to obtain the final position of the kth Arctic puffin in the new Arctic puffin population at the q+1th iteration. When the current number of iterations equals the maximum number of iterations, the iteration is stopped. The rectangular coordinate values ​​of the k-th Arctic puffin position in the new Arctic puffin population correspond to the actual effective network transmission rate and the fairness index, which are recorded as the best actual effective network transmission rate and the best fairness index, respectively. The user with the highest priority corresponding to the best actual effective network transmission rate and the best fairness index is designated as the user with the highest priority for network resource allocation scheduling. S33, first allocating the network resource channel to the user with the highest priority in network resource allocation scheduling, and then allocating and scheduling the network resource channel to users other than the user with the highest priority in network resource allocation scheduling, until the network resource channel allocation scheduling is completed; S4. Conduct a network resource channel efficacy test on the users who have completed the network resource allocation and scheduling, and evaluate the network resource channel efficacy.

2. The intelligent network resource allocation method based on 5G mobile communication according to claim 1, characterized in that: The S1 comprises the following steps: S11. Number the 5G base stations and users to generate a network communication matrix, calculate the user path loss to obtain the network data transmission rate, and perform error judgment on the network data transmission rate to obtain the theoretical effective network transmission rate.

3. The intelligent network resource allocation method based on 5G mobile communication according to claim 2, characterized in that: The S2 comprises the following steps: S21, introducing a time series, generating a time series matrix, and processing the time series matrix using a spatial reconstruction method to obtain a network transmission data matrix; S22. Select network transmission data from previous years as a support vector machine sample set, divide the support vector machine sample set into a support vector machine training set and a support vector machine test set; set the output value to e′, the actual value to e″, the penalty parameter of the support vector machine to δ, the kernel width to e, and establish the support vector machine mathematical model calculation formula as follows: Among them, minC(δ,e) represents the mathematical model of support vector machine, represents the i1th output value, represents the actual value of the i1th, i1=1,2,3,...,e; The vector machine training set is normalized to obtain a normalized vector machine training set, the normalized vector machine training set is input into the support vector machine, the maximum number of iterations and the current number of iterations are set, and when the current number of iterations of the support vector machine is equal to the maximum number of iterations, the iteration is stopped to obtain a trained support vector machine; the support vector machine test set is then input into the trained support vector machine, the accuracy threshold is set to ω, and when the absolute value of the difference between the output value of the trained support vector machine and the actual value of the vector machine training set is less than ω, a support vector machine model is obtained; otherwise, the penalty parameter and kernel width of the support vector machine are adjusted and optimized until the absolute value of the difference between the output value and the actual value is less than ω.

4. The intelligent network resource allocation method based on 5G mobile communication according to claim 3, characterized in that: A network transmission data test set is selected from the network transmission data matrix, and the network transmission data test set is input into a support vector machine model to obtain a predicted effective network transmission rate.

5. The intelligent network resource allocation method based on 5G mobile communication according to claim 3, characterized in that: Adjusting and optimizing the penalty parameters and kernel width of the support vector machine includes the following steps: S221, using the support vector machine mathematical model as the fitness function, setting a group of travelers, during which the travelers are hiking, setting the altitude of the travelers' hiking as g', the horizontal distance of the travelers' hiking as f', and the terrain inclination angle at the g-th iteration as θ g And θ g ∈[0,50°], then the calculation formula for the traveler's walking slope at the g-th iteration is as follows: Set the traveler's walking speed at the gth iteration to v g , calculate the fitness function value of the g-th iteration. The formula for calculating the traveler's walking speed at the g-th iteration is as follows: S222: There is a traveler leader in the traveler group, and the walking speed of the traveler in the g-1th iteration is set to v g-1 , at the gth iteration, the leader of the travelers is located at x g , the traveler's position at the gth iteration is x′ g , scan factor ε and ε∈[1,2], φ represents a random number and φ∈[0,1], then iteratively update the fitness function value, the traveler's walking speed v at the gth iteration g The update calculation formula is as follows: v g =v g-1 +(x g -e·x′ g )·φ; The position of the leader of the travelers is the minimum fitness function value. At this time, the minimum fitness function value is constantly sought and the calculation formula for the position of the travelers walking at the g+1th iteration is updated as follows: x′ g+1 =x′ g +f g ; set up It represents a uniformly distributed number in [0,1]. The upper bound of the initial traveler position is f″, and the lower bound of the initial traveler position is / f″′. Then the initial position of the traveler x′ satisfies the following formula: The initial position of the traveler is continuously iterated according to the walking speed of the traveler, and the minimum fitness function value is continuously updated. The maximum number of iterations is set to C′ and the current number of iterations is g. When the current number of iterations is equal to the maximum number of iterations, the iteration is stopped to obtain the position of the traveler leader. The rectangular coordinate values ​​of the traveler leader position correspond to the penalty parameter and kernel width of the support vector machine respectively.

6. The intelligent network resource allocation method based on 5G mobile communication according to claim 4, characterized in that: The S4 comprises the following steps: S41. Perform a network resource channel efficacy test on the user who has completed the network resource allocation and scheduling, use network resource channel parameters to measure the network resource channel efficacy, and evaluate the network resource channel efficacy.

7. Implementing an intelligent network resource allocation system based on 5G mobile communication as described in any one of claims 1 to 6, characterized in that: Specifically include: theory Effective network transmission rate module, effective network transmission rate prediction module, network resource allocation and scheduling module and network resource channel efficacy testing module; The theoretical effective network transmission rate module is used to establish a channel model of the network communication system to calculate the theoretical effective network transmission rate; The effective network transmission rate prediction module is used to construct a support vector machine mathematical model to realize network transmission rate prediction and adjust the optimization parameters using a travel walking optimization algorithm; The network resource allocation and scheduling module is used to divide priorities and find the highest priority user through the Arctic Puffin optimization algorithm, and perform network resource allocation and scheduling on the users in order of priority; The network resource channel efficacy test module is used to evaluate the network resource channel efficacy and replace the allocated and scheduled network resources.

Images