A cloud computing-based communication device resource dynamic allocation method and system

Abstract

The application belongs to the technical field of cloud computing and communication resource management, and discloses a communication equipment resource dynamic allocation method and system based on cloud computing, which comprises the following steps: acquiring biological migration tide data, communication network load history data and environmental parameter data, constructing a migration trajectory prediction model and a resource demand prediction model, and generating a resource elastic allocation scheme; performing base station group division on a communication network coverage area, performing resource demand space-time analysis based on migration trajectory prediction results, and calculating a resource utilization rate index; generating a resource scheduling instruction set based on the resource elastic allocation scheme and performing priority sorting, optimizing resource allocation in combination with the resource utilization rate index, and generating a base station parameter configuration instruction; adjusting base station resource configuration according to the configuration instruction, monitoring an operating state, evaluating resource efficiency, performing feedback optimization and applying the resource elastic allocation scheme; and effectively reducing equipment operation and maintenance complexity and manual intervention requirements.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of cloud computing and communication resource management, and more particularly, to a communication device resource dynamic allocation method and system based on cloud computing. BACKGROUND

[0002] With the continuous expansion of global communication network coverage area, a long-neglected problem in the industry gradually emerges: waste of communication resources and sudden congestion on the seasonal migration path of migratory birds. Research has found that on the seasonal migration path of migratory birds, there will be a "communication resource tidal effect" in spring and autumn every year, that is, a large number of micro tracking devices worn by birds simultaneously access the network, resulting in a 300% increase in resource demand of some remote base stations within 48 hours, while the adjacent area has resource idling at the same time. Traditional resource allocation methods focus on human activity patterns and completely ignore this special network load mode caused by biological migration. The existing technology cannot effectively predict this super-short cycle, high spatial heterogeneity resource demand fluctuation, and also lacks a targeted resource dynamic allocation mechanism, resulting in a large amount of bandwidth and computing resources that cannot meet the scientific data transmission demand at the critical moment, while causing resource waste in the adjacent area.

[0003] In view of this, the present application provides a communication device resource dynamic allocation method and system based on cloud computing to solve the above problems. SUMMARY

[0004] In order to overcome the above-mentioned defects of the prior art, in order to achieve the above-mentioned purpose, the present application provides the following technical scheme: a communication device resource dynamic allocation method based on cloud computing, comprising:

[0005] Step S1: obtaining biological migration tide data, communication network load historical data and environmental parameter data; constructing a migration trajectory prediction model according to the biological migration tide data and the environmental parameter data to obtain a migration trajectory prediction result; constructing a resource demand prediction model according to the communication network load historical data and the migration trajectory prediction result to obtain resource load prediction data; generating a resource allocation strategy based on the resource load prediction data to obtain a resource elastic allocation scheme;

[0006] Step S2: dividing the communication network coverage area into base station groups to obtain base station group data; performing resource demand space-time analysis according to the base station group data and the migration trajectory prediction result to obtain a resource demand distribution map; performing resource utilization rate calculation on the resource demand distribution map to obtain a resource utilization rate index;

[0007] Step S3: generating scheduling instructions for the cloud computing resource center based on the resource elasticity deployment scheme, obtaining a resource scheduling instruction set; prioritizing the resource scheduling instruction set, obtaining a hierarchical scheduling sequence; optimizing resource allocation according to the hierarchical scheduling sequence and the resource utilization rate index, obtaining an optimal resource allocation strategy; generating base station configuration parameters based on the optimal resource allocation strategy, obtaining base station parameter configuration instructions;

[0008] Step S4: adjusting resource configuration of each regional base station according to the base station parameter configuration instructions, obtaining base station resource real-time configuration data; monitoring the running state of the base station resource real-time configuration data, obtaining resource running state data; evaluating the resource utilization efficiency according to the resource running state data, obtaining resource performance evaluation results; performing feedback optimization based on the resource performance evaluation results, obtaining optimization parameters; applying the optimization parameters to the resource elasticity deployment scheme to realize closed-loop optimization and deployment of communication equipment resources.

[0009] Further, step S1 includes:

[0010] Step S11: collecting biological migration tide data and analyzing spatio-temporal characteristics, obtaining a migration activity feature library; collecting meteorological environment data and seasonal change data, obtaining environment parameter data;

[0011] Step S12: obtaining communication network region division data; constructing a network topology structure according to the communication network region division data, obtaining network topology data; collecting communication load time series data through a network monitoring system, obtaining communication network load historical data;

[0012] Step S13: constructing a migration trajectory prediction model based on the migration activity feature library and the environment parameter data, obtaining migration trajectory prediction results; extracting migration time series features from the migration trajectory prediction results;

[0013] Step S14: predicting spatial trajectories according to the migration trajectory prediction results and the migration time series features, obtaining migration path probability atlas; dividing the migration path probability atlas into time windows, obtaining time-division migration density distribution data;

[0014] Step S15: mapping the time-division migration density distribution data to the network topology data, obtaining tide load prediction data; constructing a resource demand prediction model according to the communication network load historical data and the tide load prediction data, obtaining source load prediction data;

[0015] Step S16: identifying peaks and analyzing mutations of the source load prediction data, obtaining a resource demand mutation index;

[0016] Step S17: Perform resource pool distribution analysis based on the network topology data to obtain a resource pool capability graph; perform schedulable resource evaluation on the resource pool capability graph to obtain a resource schedulable matrix;

[0017] Step S18: Generate a resource elasticity deployment scheme according to the resource demand mutation index and the resource schedulable matrix.

[0018] Further, the migration trajectory prediction model is constructed based on the migration activity feature library and the environmental parameter data to obtain migration trajectory prediction results, including:

[0019] The migration activity feature library is classified according to the migration behavior mode of the species to obtain species migration type data;

[0020] The migration trigger index is obtained by extracting the environmental trigger factor according to the environmental parameter data;

[0021] A deep learning migration prediction network is constructed based on the species migration type data to obtain a migration prediction base model;

[0022] The migration prediction base model is optimized by using the migration trigger index to obtain a migration trajectory prediction model;

[0023] The migration trajectory prediction results are generated based on the migration trajectory prediction model.

[0024] Further, step S2 includes:

[0025] Step S21: Perform base station geographic distribution analysis according to the communication network region division data to obtain base station geographic location data; perform spatial clustering analysis on the base station geographic location data to obtain base station group data;

[0026] Step S22: Perform communication coverage area calculation according to the base station group data to obtain coverage area data;

[0027] Step S23: Perform region overlap analysis based on the coverage area data and the migration path probability graph to obtain a migration influence coefficient;

[0028] Step S24: Perform resource capability evaluation on the base station group data to obtain group resource capability data;

[0029] Step S25: Generate a resource demand distribution map according to the migration influence coefficient and the group resource capability data;

[0030] Step S26: Perform time series analysis on the resource demand distribution map to obtain resource fluctuation period data; calculate the resource utilization rate according to the resource fluctuation period data to obtain a resource utilization rate index.

[0031] Further, step S3 includes:

[0032] Step S31: quantifying the computing resource demand according to the resource elasticity deployment scheme, obtaining computing resource allocation data; quantifying the storage resource demand according to the resource elasticity deployment scheme, obtaining storage resource allocation data; quantifying the bandwidth resource demand according to the resource elasticity deployment scheme, obtaining bandwidth resource allocation data;

[0033] Step S32: generating a resource scheduling instruction set based on the computing resource allocation data, the storage resource allocation data and the bandwidth resource allocation data;

[0034] Step S33: performing time effectiveness evaluation on the resource scheduling instruction set according to the time-migration density distribution data, obtaining instruction time effectiveness level data;

[0035] Step S34: performing priority sorting on the resource scheduling instruction set based on the instruction time effectiveness level data, obtaining a hierarchical scheduling sequence;

[0036] Step S35: performing cloud resource scheduling simulation according to the hierarchical scheduling sequence and the resource utilization rate index, obtaining resource allocation prediction results; performing resource competition conflict detection on the resource allocation prediction results, obtaining conflict resolution strategies;

[0037] Step S36: generating an optimal resource allocation strategy based on the resource allocation prediction results and the conflict resolution strategies; converting and generating base station parameter configuration instructions according to the optimal resource allocation strategy.

[0038] Further, step S32 includes:

[0039] Step S321: performing virtual machine resource configuration conversion on the computing resource allocation data, obtaining virtual machine configuration instructions;

[0040] Step S322: performing storage unit mapping on the storage resource allocation data, obtaining storage allocation instructions;

[0041] Step S323: performing network channel planning on the bandwidth resource allocation data, obtaining bandwidth allocation instructions;

[0042] Step S324: performing instruction format standardization on the virtual machine configuration instructions, the storage allocation instructions and the bandwidth allocation instructions, obtaining standardized scheduling instructions;

[0043] Step S325: performing instruction set integration on the standardized scheduling instructions, obtaining a resource scheduling instruction set.

[0044] Further, step S34 includes:

[0045] Step S341: constructing an instruction priority evaluation model according to the instruction time effectiveness level data, obtaining a priority evaluation function;

[0046] Step S342: applying a priority evaluation function to the resource scheduling instruction set to obtain an instruction priority value;

[0047] Step S343: performing resource scheduling instruction sorting according to the instruction priority value to obtain priority sequence data;

[0048] Step S344: performing resource dependency relationship analysis on the priority sequence data to obtain a dependency constraint matrix; and adjusting the priority sequence data according to the dependency constraint matrix to obtain a hierarchical scheduling sequence.

[0049] Further, step S4 includes:

[0050] Step S41: performing parsing and conversion on the base station parameter configuration instruction to obtain a base station configuration parameter set; and performing parameter configuration update on each base station according to the base station configuration parameter set to obtain base station parameter update data;

[0051] Step S42: performing configuration validity verification on the base station parameter update data to obtain a configuration verification result;

[0052] Step S43: performing base station resource performance monitoring according to the configuration verification result to obtain resource performance data; and performing resource usage state collection on the resource performance data to obtain resource running state data;

[0053] Step S44: performing energy efficiency index calculation according to the resource running state data to obtain energy efficiency evaluation data; and performing service quality evaluation on the resource running state data to obtain a service quality index;

[0054] Step S45: performing comprehensive resource performance evaluation based on the energy efficiency evaluation data and the service quality index to obtain a resource performance evaluation result;

[0055] Step S46: performing optimization space analysis on the resource performance evaluation result to obtain optimization direction data; generating a parameter tuning strategy according to the optimization direction data to obtain tuning parameters; and feeding back the tuning parameters to the resource elasticity allocation scheme to realize closed-loop optimization of resource configuration.

[0056] Further, step S45 includes:

[0057] Step S451: performing multi-dimensional index standardization on the energy efficiency evaluation data to obtain a standardized energy efficiency index;

[0058] Step S452: performing user experience mapping on the service quality index to obtain experience satisfaction data;

[0059] Step S453: constructing a resource performance evaluation model, and performing weighted fusion on the standardized energy efficiency index and the experience satisfaction data to obtain a comprehensive performance index;

[0060] Step S454: generating resource performance evaluation results according to the comprehensive performance index.

[0061] A cloud computing-based communication device resource dynamic allocation system comprises:

[0062] The migration elasticity prediction module: acquires biological migration tide data, communication network load history data and environmental parameter data; constructs a migration trajectory prediction model according to the biological migration tide data and the environmental parameter data, and obtains migration trajectory prediction results; constructs a resource demand prediction model according to the communication network load history data and the migration trajectory prediction results, and obtains resource load prediction data; generates a resource allocation strategy based on the resource load prediction data, and obtains a resource elasticity allocation scheme;

[0063] The space-time demand analysis module: divides the communication network coverage area into base station groups to obtain base station group data; performs space-time analysis of resource demand based on the base station group data and the migration trajectory prediction results to obtain a resource demand distribution map; and performs resource utilization rate calculation on the resource demand distribution map to obtain a resource utilization rate index;

[0064] The dynamic scheduling optimization module: generates scheduling instructions for the cloud computing resource center based on the resource elasticity allocation scheme to obtain a resource scheduling instruction set; performs priority sorting on the resource scheduling instruction set to obtain a hierarchical scheduling sequence; performs resource allocation optimization based on the hierarchical scheduling sequence and the resource utilization rate index to obtain an optimal resource allocation strategy; and generates base station configuration parameters based on the optimal resource allocation strategy to obtain base station parameter configuration instructions;

[0065] The feedback tuning module: adjusts the resource configuration of the base stations in each region according to the base station parameter configuration instructions to obtain base station resource real-time configuration data; performs running state monitoring on the base station resource real-time configuration data to obtain resource running state data; performs resource utilization efficiency evaluation based on the resource running state data to obtain resource performance evaluation results; performs feedback tuning based on the resource performance evaluation results to obtain tuning parameters; and applies the tuning parameters to the resource elasticity allocation scheme to realize closed-loop optimization and allocation of communication device resources.

[0066] The technical effects and advantages of the cloud computing-based communication device resource dynamic allocation method and system of the present application are as follows:

[0067] The application can accurately predict the space-time distribution characteristics of the communication network load, provide a scientific basis for resource allocation, and improve the foresight and accuracy of resource allocation; by dividing the base station groups of the communication network coverage area, different regions can be analyzed according to the resource demand characteristics, and a more detailed resource demand distribution map can be obtained, the spatial accuracy of resource allocation is provided, and the resource utilization efficiency is improved; by prioritizing the resource scheduling instruction set, limited resources can be reasonably allocated under resource competition, resource scheduling conflicts can be solved, the communication service quality of key regions can be ensured, and meanwhile, the base station parameter configuration instructions generated based on the optimal resource allocation strategy can accurately guide the resource configuration adjustment of the base stations in each region, and the accuracy and execution efficiency of the configuration are improved. BRIEF DESCRIPTION OF DRAWINGS

[0068] Figure 1 A communication equipment resource dynamic allocation method based on cloud computing according to the present application is shown in the figure.

[0069] Figure 2 A communication equipment resource dynamic allocation system based on cloud computing according to the present application is shown in the figure. DETAILED DESCRIPTION

[0070] The technical solutions in the embodiments of the present application will be described clearly and completely below with reference to the accompanying drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor fall within the scope of the present application.

[0071] Embodiment 1

[0072] Please refer to Figure 1 The communication equipment resource dynamic allocation method based on cloud computing according to the present application includes the following steps:

[0073] Step S1: acquiring biological migration tide data, communication network load historical data and environmental parameter data; constructing a migration trajectory prediction model according to the biological migration tide data and the environmental parameter data to obtain a migration trajectory prediction result; constructing a resource demand prediction model according to the communication network load historical data and the migration trajectory prediction result to obtain resource load prediction data; generating a resource allocation strategy based on the resource load prediction data to obtain a resource elastic allocation scheme;

[0074] Step S2: performing base station group division on the communication network coverage area to obtain base station group data; performing resource demand space-time analysis according to the base station group data and the migration trajectory prediction result to obtain a resource demand distribution map; and performing resource utilization rate calculation on the resource demand distribution map to obtain a resource utilization rate index;

[0075] Step S3: generating scheduling instructions for the cloud computing resource center based on the resource elastic deployment scheme to obtain a resource scheduling instruction set; performing priority sorting on the resource scheduling instruction set to obtain a hierarchical scheduling sequence; performing resource allocation optimization according to the hierarchical scheduling sequence and the resource utilization rate index to obtain an optimal resource allocation strategy; and generating base station configuration parameters based on the optimal resource allocation strategy to obtain base station parameter configuration instructions;

[0076] Step S4: performing resource configuration adjustment on the base stations in each region according to the base station parameter configuration instructions to obtain base station resource real-time configuration data; performing running state monitoring on the base station resource real-time configuration data to obtain resource running state data; performing resource utilization efficiency evaluation according to the resource running state data to obtain resource performance evaluation results; performing feedback tuning based on the resource performance evaluation results to obtain tuning parameters; and applying the tuning parameters to the resource elastic deployment scheme to realize closed-loop optimization and deployment of the communication device resources.

[0077] Preferably, step S1 comprises the following steps:

[0078] Step S11: collecting biological migration tide data and performing space-time characteristic analysis to obtain a migration activity feature library; collecting meteorological environment data and seasonal change data to obtain environment parameter data;

[0079] Step S12: obtaining communication network region division data; constructing a network topology structure according to the communication network region division data to obtain network topology data; and collecting communication load time series data through a network monitoring system to obtain communication network load historical data;

[0080] Step S13: constructing a migration trajectory prediction model based on the migration activity feature library and the environment parameter data to obtain migration trajectory prediction results; and performing time series feature extraction on the migration trajectory prediction results to obtain migration time series features;

[0081] Step S14: performing space trajectory prediction according to the migration trajectory prediction results and the migration time series features to obtain a migration path probability atlas; and performing time window division on the migration path probability atlas to obtain time-division migration density distribution data;

[0082] Step S15: mapping the time-division migration density distribution data to the network topology data to obtain tide load prediction data; and constructing a resource demand prediction model according to the communication network load historical data and the tide load prediction data to obtain source load prediction data;

[0083] Step S16: Peak identification and mutation analysis are performed on the source load prediction data to obtain a resource demand mutation index;

[0084] Step S17: Resource pool distribution analysis is performed based on the network topology data to obtain a resource pool capability graph; and schedulable resource evaluation is performed on the resource pool capability graph to obtain a resource schedulable matrix;

[0085] Step S18: A resource elasticity deployment scheme is generated according to the resource demand mutation index and the resource schedulable matrix.

[0086] Specifically, migration tide data is collected and analyzed for spatio-temporal characteristics to obtain a migration activity feature library. The collection methods include satellite tracking data, ring data, observation station records, and radio telemetry data, covering the main migration seasons in the past five years. The migration data collected is analyzed for spatio-temporal clustering, and the density clustering algorithm DBSCAN is used to identify migration paths and key stopover points. The clustering process can be represented as: Cen = DBSCAN(Point, ε, MinPts), where Point represents the set of migration trajectory points, ε represents the clustering radius parameter (usually set to 5 to 10 kilometers, representing the maximum distance for determining the same cluster), MinPts represents the density threshold (usually set to 5 to 10 points, representing the minimum number of points required to form a cluster), and Cen is the set of clustering results. Through clustering analysis, high-frequency channels in the migration path, key stopover areas, and migration intensity variation patterns are identified, and a migration activity feature library is constructed. At the same time, meteorological environmental data and seasonal variation data are collected, including temperature curves, precipitation, wind direction and speed, air pressure changes, and vegetation coverage index, etc. Seasonal indicators form the environmental parameter data set. Obtain communication network regional division data, including base station location coordinates, coverage range, frequency configuration, antenna direction, and other network deployment information. According to the communication network regional division data, the network topology structure is constructed, and a weighted directed graph G(V, E, W) is used to represent it, where V represents the set of network nodes (including base stations, controllers, gateways, and core network devices, etc.), E represents the connection relationship between nodes (including physical links and logical links), and W represents the connection weight matrix (including bandwidth capacity, link delay, transmission distance, etc. indicators). The network topology structure model comprehensively describes the physical and logical connection relationship of the communication network, providing a spatial reference for subsequent resource allocation. Through the network monitoring system, communication load time series data is collected, including the number of user connections, data traffic, resource utilization, and other indicators of each network node at different time points. Time series analysis is performed on the historical data of the communication network load to identify daily load fluctuation patterns, periodic variation rules, and abnormal load events, providing basic data for subsequent load prediction. Based on the migration activity feature library and environmental parameter data, a migration trajectory prediction model is constructed. A deep learning method combining long short-term memory network (LSTM) and attention mechanism is used to capture the time series dependence of migration behavior and the influence weight of environmental factors. The model input includes historical migration data X1 and environmental factor data X2, and the output is the predicted migration behavior Yact. The prediction process can be represented as: Yact = LSTM(X1, X2; θ), where θ represents the model parameter set, including network weights and bias terms. The LSTM network is particularly suitable for processing time series data with long-term dependencies, and can effectively capture the seasonal patterns and environmental response characteristics of bird migration behavior. Through this model, migration trajectory prediction results for different populations are generated, including predicted migration time window, path selection probability, flight speed and direction, and stopover duration.The time series features of the migration trajectory prediction results are extracted, and the periodicity, suddenness and trend features of the migration activities are analyzed by using wavelet transform method. Through multi-scale wavelet analysis, the change features of the migration activities on different time scales are extracted, including daily change, weekly change and seasonal change, and the migration time series features are obtained. The spatial trajectory prediction is performed according to the migration trajectory prediction results and the migration time series features. The particle filtering algorithm is used to simulate the population migration trajectory, which can handle the nonlinear dynamic characteristics and random uncertainty in the migration process. The particle filtering process can be expressed as: pS(x. t |z:t)≈∑w i ×δ(x t -xi t ), wherein x t is a state vector (including position, velocity, direction, etc.) at time t, z:t is an observation sequence at time t, w i is the weight of the i th particle (indicating the probability of the particle representing the true state), and xi tFor the state of the i-th particle at time t, δ is the Dirac function (used for discrete probability representation). By particle filtering algorithm, a large number of possible migration paths are sampled and evaluated to generate a migration path probability atlas, representing the probability distribution of different regions being passed by the bird population at different time points. Time window division is performed on the migration path probability atlas, dividing the prediction time range (usually the migration season) into multiple time windows, each usually 1-3 days. The migration density distribution MD(r, Tim) of each region in each time window is calculated, where r represents the spatial region (usually divided into grid cells), and Tim represents the time window. The migration density distribution reflects the expected number of birds per unit area per unit time, which is a key indicator to quantify the intensity of migration activity. Through time window division, time-dependent migration density distribution data is obtained, accurately describing the spatio-temporal evolution process and dynamic change characteristics of migration activity. The time-dependent migration density distribution data is mapped to the network topology data to establish the mapping relationship between migration activity and communication load. The mapping function can be represented as: L'(v, t) = L(v, t) + α × D(r(v), t), where L'(v, t) represents the predicted load of node v at time t, L(v, t) represents the baseline load (expected load under normal circumstances), r(v) represents the region covered by node v, D(r(v), t) represents the migration density of the region at time t, and α is the mapping coefficient (the influence factor of migration activity on communication load calibrated according to historical data). Through this mapping relationship, the additional communication demand caused by bird migration is quantified, and the tidal load prediction data is generated. According to the historical data of communication network load and the tidal load prediction data, a resource demand prediction model is constructed, and the random forest regression method is used to predict the resource demand of network nodes. The model can be represented as: R(v, t) = RF(L'(v, t), F(v)), where R(v, t) represents the resource demand prediction value of node v at time t (including computing resources, storage resources and network resources), RF represents the random forest model, L'(v, t) represents the predicted load, and F(v) represents the feature vector of node v (including processing capacity, hardware configuration, etc.). The random forest model is composed of multiple decision trees, each tree is trained using a random feature subset, and the final prediction result is obtained by averaging the outputs of all trees, which has good generalization ability and robustness. Through this model, source load prediction data is generated to describe the change trend and fluctuation rule of resource demand of each node in the network in the future time period. Peak identification and mutation analysis are performed on the source load prediction data, and the Z-score-based anomaly detection method is used to identify resource demand mutation points. Cluster analysis is performed on the identified mutation points to calculate the resource demand mutation index: I(v, t) = Z(v, t) × D(r(v), t) × w(v), where w(v) represents the importance weight of node v (determined according to the position and function of the node in the network).The resource demand mutation index comprehensively considers the resource demand change amplitude, migration density influence and node importance, and provides a quantitative index for resource allocation priority. The higher the mutation index, the higher the priority of resource allocation at the time point to prevent the decline of service quality. Based on the network topology data, the resource pool distribution analysis is performed, and the network resources are abstracted as distributed resource pools, each of which contains three types of computing resources, storage resources and network resources. The resource pool capacity can be represented as a matrix Cap, where Cap[j, k] represents the capacity of the kth type of resource in resource pool j (k=1 represents computing resources, k=2 represents storage resources, and k=3 represents network resources). By analyzing the connection relationship and scheduling delay between resource pools, a resource pool capacity atlas is generated to describe the resource distribution and scheduling capability of the entire network. The scheduling delay matrix Delay between resource pools is Delay[j, m], which represents the time delay required for scheduling resources from resource pool j to resource pool m. The resource pool capacity atlas is evaluated for schedulable resources, and the available resource amount of each resource pool at different time points is calculated as Aus(j, t, k) = Cap[j, k]-U(j, t, k), where U(j, t, k) represents the resource usage of the kth type of resource in resource pool j at time t. According to the resource availability and scheduling constraints (based on the scheduling delay matrix, the scheduling waiting time is greater than the time delay), a resource schedulable matrix Max is generated, where Max[j, m, t] represents the maximum schedulable resource amount from resource pool j to resource pool m at time t, and the calculation formula is: Max[j, m, t] = min(Aus(j, t, k), Bac(m, t)), where Bac(m, t) represents the maximum resource receiving capability of resource pool m at time t. The resource schedulable matrix comprehensively describes the feasibility and constraints of resource scheduling in the network, providing a basis for the generation of resource allocation schemes. According to the resource demand mutation index and the resource schedulable matrix, a resource elastic allocation scheme is generated. A multi-objective optimization algorithm is used, and the objective functions include minimizing the resource allocation cost, minimizing the service quality decline risk and maximizing the resource utilization rate. The optimization problem can be represented as: min f() = w1 × CB(Sch) + w2 × Qua(Sch) - w3 × Use(Sch), where Sch represents the resource allocation scheme (including allocation time, source node, target node and resource amount), CB(Sch) represents the allocation cost function (proportional to the scheduling distance and resource amount), Qua(Sch) represents the service quality risk function (proportional to the degree of unmet demand), Use(Sch) represents the resource utilization rate function (proportional to the overall resource utilization efficiency), and w1, w2 and w3 are weight coefficients (determined according to the business priority).The multi-objective optimization problem is solved by mixed integer programming or genetic algorithm under the constraints of resource schedulable matrix and network constraints, to generate a resource elasticity deployment scheme, including deployment schedule (specifying when to trigger resource deployment), resource migration path (specifying from which resource pool to deploy to which resource pool), and allocation strategy (specifying the specific allocation ratio of each type of resource), to realize dynamic optimization configuration of communication resources, and effectively cope with the communication load tidal changes caused by migratory birds migration.

[0087] Preferably, step S2 comprises the following steps:

[0088] Step S21: performing base station geographic distribution analysis according to the communication network region division data to obtain base station geographic position data; performing spatial clustering analysis on the base station geographic position data to obtain base station group data;

[0089] Step S22: performing communication coverage area calculation according to the base station group data to obtain coverage area data;

[0090] Step S23: performing region overlap degree analysis based on the coverage area data and migration path probability atlas to obtain migration influence coefficient;

[0091] Step S24: performing resource capability evaluation on the base station group data to obtain group resource capability data;

[0092] Step S25: generating resource demand distribution map according to the migration influence coefficient and the group resource capability data;

[0093] Step S26: performing time series analysis on the resource demand distribution map to obtain resource fluctuation period data; calculating resource utilization rate according to the resource fluctuation period data to obtain resource utilization rate index.

[0094] Specifically, according to the communication network region division data, the geographic distribution of base stations is analyzed. First, the geographic coordinate data of all base stations in the network region is collected, including longitude, latitude and altitude. For each base station, its geographic position is represented by a vector, and the geographic position data of the base stations in the entire communication network is organized as a set; in addition, the coverage radius, transmit power, antenna direction and other technical parameters of each base station are collected to form a base station technical parameter set. Based on the base station geographic position data and technical parameters, the spatial relationship between base stations is analyzed, including distance, signal coverage overlap degree, etc., to provide a basis for subsequent spatial clustering. The spatial clustering analysis of base station geographic position data is performed by using an improved DBSCAN (Density-Based Spatial Clustering) algorithm. This algorithm can identify spatial clusters of arbitrary shape and can handle noise points. Specifically, for any base station ST, first calculate the distance between other base stations and ST, and use the spherical distance formula to calculate:

[0095] dL(ot, ST) = Rb arccos(sin(ot1) sin(ST1) + cos(ot2) cos(ST2) cos(ot1 - ST1)); wherein dL(ot, ST) represents the base station distance between base station ot and base station ST, R is the earth radius (about 6371 kilometers), ot1 and ot2 are the longitude and latitude coordinates (radian system) of base station ot, respectively, and ST1 and ST2 are the longitude and latitude coordinates (radian system) of base station ST, respectively. The base station ot with a base station distance less than the preset neighborhood radius parameter is classified into the base station ST, and when the number of base stations classified into the base station ST is greater than or equal to the minimum base station parameter, the base station ST is regarded as a core point and used for expansion clustering. Through an iterative expansion process, a plurality of base station groups are finally obtained, each group containing a plurality of base stations with similar geographical positions. According to the base station group data, the communication coverage area is calculated, and an improved Voronoi diagram method is adopted in combination with the transmission power and antenna direction parameters of the base station. For each base station group, the signal coverage range of each base station inside the group is first calculated, and then these coverage areas are combined to obtain the overall coverage range of the group. The signal coverage range of a single base station can be represented by a signal strength contour line, and the signal strength Pua_r can be calculated by the following model:

[0096] wherein Pua_r(d) represents the signal strength P_sen at a distance d from the base station, P_sen is the transmission power, G_sen and G_ace are the transmission and reception antenna gains, respectively, Bc is the signal wavelength, d is the distance, n is the path loss index (usually 2-4, depending on the environment), and π represents the circular constant. By setting a minimum received signal strength threshold P_min, the maximum coverage radius R_max of the base station can be determined:

[0097] Based on the calculated coverage radius, combined with factors such as terrain obstacles and building obstructions, the actual coverage area of each base station is further adjusted to form a more accurate coverage model. Finally, the coverage areas of each base station in the same group are merged to obtain group coverage area data. Based on the coverage area data and the migration path probability map, the area overlap analysis is performed to calculate the overlap degree of each base station group with the migration path of the migratory birds. Assuming that the obtained migration path probability map is represented as Map(x, y, t), where Map(x, y, t) represents the probability of the location (x, y) being passed through by the migratory bird population at time t. For the coverage area of each base station group, the overlap degree with the migration path can be calculated by the following integral: O_k(t) = ∫∫(A_k ∩ Map(x, y, t)) dx dy; where A_k ∩ Map(x, y, t) represents the intersection of the coverage area A_k and the migration path at location (x, y) and time t. The overlap degree O_k(t) represents the number of devices expected to be served by the base station group at time t. Based on the overlap degree data, the migration impact coefficient I_k(t) is further calculated to measure the degree of influence of the migration of migratory birds on the resource demand of the base station group at time t: I_k(t) = O_k(t) ÷ ∫∫A_k dx dy, where ∫∫A_k dx dy represents the total area of the coverage area A_k. The migration impact coefficient I_k(t) has a value range of [0, 1], and the larger the value, the greater the influence of the migration of migratory birds on the group, and the higher the dynamic resource allocation capability required. The resource capacity of the base station group data is evaluated, considering factors such as the hardware configuration, bandwidth resources, computing capacity, etc. of each base station in the group. For a base station group, its resource capacity is composed of the following aspects: bandwidth resources: Wid = ∑(wid_u), where wid_u represents the available frequency spectrum width of base station u in the group; computing resources: Vca = ∑(cal_u), where cal_u represents the computing capacity (such as CPU core number × clock frequency) of base station u in the group; storage resources: Save = ∑(sav_u), where sav_u represents the storage capacity of base station u in the group; energy efficiency: Een = ∑(en_u × pr_u) ÷ ∑(pr_u), where en_i represents the energy efficiency ratio of base station u in the group, and pr_u represents the transmission power of base station u in the group. By weighting and summing these factors, the group resource capacity index is obtained to balance the importance of different resource dimensions. Based on the resource capacity index of each group, the group resource capacity dataset is formed. According to the migration impact coefficient and the group resource capacity data, the resource demand distribution map is generated, which is a spatiotemporal three-dimensional mapping showing the resource demand intensity at different times and in different regions. For each base station group, its resource demand Dnd_k(t) at time t can be calculated by the following formula: Dnd_k(t) = I_k(t) × U_k × ∑R(v, t); where I_k(t) is the migration impact coefficient, and U_k is the basic user capacity (positively correlated with the group resource capacity index).Specifically, U_k can be expressed as: U_k = translate x Rs_k; where translate is a resource conversion coefficient, and Rs_k represents a group resource capability index, which converts resource capability into the number of users that can be served. The resource demand at different locations and different times is mapped to the geographic space to form a resource demand distribution map, which intuitively shows the spatio-temporal variation characteristics of resource demand. The resource demand distribution map is subjected to time series analysis, and wavelet transform and spectrum analysis methods are used to decompose and analyze the time series data of resource demand. For each base station group, the resource demand time series is extracted, and the continuous wavelet transform is performed on the resource demand time series. By analyzing the modulus value of the wavelet transform coefficient, the periodic component in the demand sequence is identified, and the resource fluctuation period data PLZ_k = {p_k1, p_k2,..., p_kr} is obtained, where p_kr represents the period length of the rth periodic component. The resource utilization rate is calculated according to the resource fluctuation period data, and the resource utilization rate index is used to measure the resource use efficiency of the base station group within the period Tcir: where usTar represents the resource utilization rate index, which represents the ratio of resource demand at time t to total resource capability. Ideally, the resource utilization rate should be close to but not exceed 1, and a too high utilization rate can lead to congestion, and a too low utilization rate means resource waste. By analyzing the resource utilization rate index of each group, areas with excess or insufficient resources can be identified, providing decision basis for subsequent dynamic resource allocation. In specific implementation, the resource utilization rate evaluation can be further refined in combination with the resource fluctuation period data. For groups with obvious periodicity, the resource utilization rate variance in different periods is calculated. A larger variance value indicates that the resource utilization rate fluctuates greatly, requiring stronger dynamic allocation capability; a smaller variance means relatively stable resource demand, and a more conservative allocation strategy can be adopted.

[0098] Preferably, step S3 comprises the following steps:

[0099] Step S31: quantifying the resource demand according to the resource elasticity allocation scheme to obtain computing resource allocation data; quantifying the resource demand according to the resource elasticity allocation scheme to obtain storage resource allocation data; quantifying the resource demand according to the resource elasticity allocation scheme to obtain bandwidth resource allocation data;

[0100] Step S32: generating a resource scheduling instruction set based on the computing resource allocation data, the storage resource allocation data, and the bandwidth resource allocation data;

[0101] Step S33: evaluating the time effectiveness of the resource scheduling instruction set according to the time-migration density distribution data to obtain instruction time effectiveness level data;

[0102] Step S34: Priority ranking of the resource scheduling instruction set based on the instruction time limit level data, to obtain a hierarchical scheduling sequence;

[0103] Step S35: Cloud resource scheduling simulation according to the hierarchical scheduling sequence and the resource utilization rate index, to obtain a resource allocation prediction result; resource competition conflict detection on the resource allocation prediction result, to obtain a conflict resolution strategy;

[0104] Step S36: Generation of an optimal resource allocation strategy based on the resource allocation prediction result and the conflict resolution strategy; conversion of the optimal resource allocation strategy to generate a base station parameter configuration instruction.

[0105] Specifically, the resource elasticity allocation scheme is used for calculation resource demand quantification, and a multi-factor weighted calculation model is used for processing. The calculation resource demand quantification formula is:

[0106] C_req = ∑(con_r × Lfz_r × Pcpl_r × Fser_r); wherein, C_req represents the total amount of computing resource demand, with the unit of computing capacity unit (CCU); con_r represents the resource consumption coefficient of the rth type of service, reflecting the consumption characteristics of different service types on computing resources; Lfz_r represents the load prediction value of the rth type of service, including the number of users, the number of connections, etc.; Pcpl_r represents the processing complexity factor of the rth type of service, reflecting the computing complexity of service processing; Fser_r represents the quality of service factor of the rth type of service, related to service response time and reliability requirements, the higher the requirement, the larger the factor. The resource elasticity allocation scheme is stored resource demand quantification, data flow analysis and storage prediction model are adopted. The storage resource demand quantification formula is: S_req = ∑(Dgen_ty × Rda_ty × Tsa_ty) + S_base; wherein, S_req represents the total amount of storage resource demand, with the unit of GB; Dgen_ty represents the predicted generation rate of the tyth type of data, with the unit of GB per hour; Rda_ty represents the redundancy coefficient of the tyth type of data, considering data backup and fault tolerance demand; Tsa_ty represents the storage duration of the tyth type of data, with the unit of hours; S_base represents the basic system storage demand, including the storage space required by operating system, application software and fixed configuration data. The resource elasticity allocation scheme is bandwidth resource demand quantification, network flow model and service quality mapping method are adopted. The bandwidth resource demand quantification formula is: B_req = max(B_up, B_down) × (1 + redun); wherein, B_req represents the total amount of bandwidth resource demand, with the unit of Mbps; B_up represents the uplink bandwidth demand, B_down represents the downlink bandwidth demand; redun represents the bandwidth redundancy coefficient, used to cope with traffic fluctuations and burst conditions, usually taking the value of 0.1 to 0.3. Based on the computing resource allocation data, the storage resource allocation data and the bandwidth resource allocation data, the resource scheduling instruction set is generated. The resource scheduling instruction adopts a standardized format definition: Ins = {Type, Amount, Location, Time, Duration, Priority}; wherein, Type represents the resource type, including computing resource demand, storage resource demand and bandwidth resource demand; Amount represents the resource allocation quantity; Location represents the resource allocation location, which can be a physical server ID or a virtual resource pool identifier; Time represents the execution time; Duration represents the resource occupation duration; Priority represents the initial priority of the instruction. The scheduling instruction set is converted from the resource demand data through the resource mapping algorithm, ensuring that various resource demands can be mapped to specific scheduling operations. According to the time-migration density distribution data, the timeliness of the resource scheduling instruction set is evaluated, and the time sensitivity analysis method is adopted. The higher the timeliness score is, the more urgent the instruction is, and the faster it needs to be executed.The instruction time effectiveness level data is used to prioritize the resource scheduling instruction set, a multi-factor comprehensive sorting algorithm is used to weight and sum the time effectiveness score, resource influence degree and cost benefit ratio of the instruction to obtain a final priority score; the resource influence degree of the instruction reflects the influence degree of the instruction execution on the overall resource state; the cost benefit ratio of the instruction reflects the ratio of the service quality improvement brought by the instruction execution to the resource consumption; according to the calculated priority score, the instructions are sorted from high to low to form a hierarchical scheduling sequence. According to the hierarchical scheduling sequence and the resource utilization rate index, the cloud resource scheduling simulation is carried out, the discrete event simulation method is used to iteratively calculate the resource state at different times to obtain the resource allocation prediction result. The resource allocation prediction result is detected for resource competition conflict, a conflict graph coloring algorithm is used to identify the resource conflict and generate a conflict resolution strategy. The conflict detection standard is that when the instruction InsI and the instruction InsJ satisfy the instruction logic condition: the conflict state between the instruction InsI and the instruction InsJ is 1, otherwise the conflict state between the instruction InsI and the instruction InsJ is 0; the conflict state of 1 indicates that there is a conflict, and 0 indicates that there is no conflict; resource() represents the resource set requested by the instruction, time() represents the execution time period of the instruction, representing an empty set. The conflict resolution strategy includes instruction reordering, resource substitution and task splitting methods. Based on the resource allocation prediction result and the conflict resolution strategy, an optimal resource allocation strategy is generated, and a multi-objective optimization method is used. The optimization objectives include: the comprehensive score of the resource allocation strategy, which represents the resource utilization rate and the service quality satisfaction degree. According to the optimal resource allocation strategy, a base station parameter configuration instruction is generated, and a strategy mapping and parameter conversion method is used. The base station parameter configuration includes wireless resource configuration, power control, antenna parameter and cell coverage adjustment, etc. The mapping process considers the wireless environment characteristics, device capability limitation and network topology structure, to ensure that the generated parameter configuration instruction can be effectively executed and achieve the expected resource allocation effect.

[0107] Preferably, step S4 comprises the following steps:

[0108] Step S41: analyzing and converting the base station parameter configuration instruction to obtain a base station configuration parameter set; updating the parameters of each base station according to the base station configuration parameter set to obtain base station parameter update data;

[0109] Step S42: verifying the configuration validity of the base station parameter update data to obtain a configuration verification result;

[0110] Step S43: monitoring the resource performance of the base station according to the configuration verification result to obtain resource performance data; collecting the resource usage state of the resource performance data to obtain resource running state data;

[0111] Step S44: Perform energy efficiency index calculation according to the resource running state data to obtain energy efficiency evaluation data; perform service quality evaluation on the resource running state data to obtain service quality indexes;

[0112] Step S45: Perform comprehensive resource performance evaluation based on the energy efficiency evaluation data and the service quality indexes to obtain resource performance evaluation results;

[0113] Step S46: Perform optimization space analysis on the resource performance evaluation results to obtain optimization direction data; generate parameter tuning strategies according to the optimization direction data to obtain tuning parameters; feed back the tuning parameters to the resource elastic allocation scheme to realize closed-loop optimization of resource configuration.

[0114] Specifically, when parsing and converting the base station parameter configuration instructions, a hierarchical parsing framework is used to handle different types of configuration instructions. For standardized XML or JSON format configuration instructions, first, a syntax parser is used to extract the instruction structure and parameter values; then, through a parameter mapping table, general parameters are converted into special parameters of specific vendor base stations, and finally, a complete set of base station configuration parameters is formed, which are sent to each base station through SNMP, NETCONF or special API interface to complete parameter configuration update. During the update process, the system records the application state of each parameter and the feedback information of the base station, forming the base station parameter update data, which provides the basis for subsequent verification. When verifying the configuration effectiveness of the base station parameter update data, a multi-stage verification strategy is adopted to ensure the integrity and correctness of the configuration. First, parameter landing verification is performed by reading the current configuration of the base station to check whether the parameters are correctly written, then function verification is performed by sending test traffic or probe signals to detect whether the key functions of the base station are normal, and finally performance verification is performed to check whether the performance indicators after parameter adjustment meet the expectations. Based on the verification results of these three aspects, a complete configuration verification result is formed. When monitoring the resource performance of the base station according to the configuration verification result, a hierarchical monitoring framework is adopted to cover three levels of physical resources, virtual resources and business resources. Physical resource monitoring mainly focuses on hardware level indicators such as CPU usage, memory occupancy, network interface load, power state, etc.; virtual resource monitoring focuses on resource allocation and usage in virtualization environment such as virtual CPU allocation rate, virtual memory usage efficiency, virtual network throughput, etc.; business resource monitoring focuses on resource indicators directly related to communication business such as wireless channel utilization, user connection number, per-user throughput, etc. The monitoring data collection adopts a hierarchical sampling strategy, with high-frequency sampling for key indicators (e.g. 5 seconds per sample), medium-frequency sampling for general indicators (e.g. 30 seconds per sample), and low-frequency sampling for auxiliary indicators (e.g. 5 minutes per sample), and a threshold triggering mechanism is set to increase the sampling frequency when the indicators exceed the preset range. For the collected resource performance data, resource usage state analysis is performed using a moving window statistical method to calculate the statistical characteristics of each resource indicator, including mean, peak, variance, trend, etc., to form a resource load feature vector. At the same time, time series pattern recognition algorithms are used to analyze resource usage patterns to identify periodic load changes and sudden events. Through these analyses, comprehensive resource running state data is obtained. When calculating energy efficiency indicators based on resource running state data, differentiated energy efficiency models are used for different types of resources. For computing resources, the computing throughput (represented by available FLOPS or number of transactions per second) is divided by the power consumption of the computing resources to obtain the computing energy efficiency indicator. For wireless resources, the effective bandwidth utilization (considering the actual amount of data transmitted) is divided by the power consumption of the wireless devices to represent the wireless energy efficiency indicator. For network resources, the network throughput is divided by the power consumption of the network devices to obtain the network resource energy efficiency. By weighting and summing these indicators, the overall energy efficiency evaluation indicator is obtained.When evaluating the quality of service of resource running state data, a multi-dimensional user experience index system is adopted, including three dimensions of access experience, interaction experience and content experience. By comprehensively considering these three dimensions, a weighted summation formula is used to calculate the service quality index. When performing comprehensive resource efficiency evaluation based on energy efficiency evaluation data and service quality index, a multi-objective balance strategy is adopted to seek the best balance point between energy efficiency and service quality. Through this evaluation method, the system can comprehensively consider energy efficiency and service quality, avoid the negative effects that may be caused by single index optimization, and obtain a comprehensive resource efficiency evaluation result. When performing optimization space analysis on the resource efficiency evaluation result, two methods of target gap analysis and resource balance analysis are adopted. The target gap analysis calculates the gap between the current efficiency and the target efficiency: Gdif = (E_target - E_current) ÷ E_target, where Gdif represents the relative gap, E_target represents the target efficiency value, and E_current represents the current efficiency value. When the gap Gdif is greater than a preset threshold, the optimization process is triggered. The resource balance analysis detects the load balancing of various resources and calculates the load balancing degree: B = σ(L) ÷ μ(L). where B represents the load balancing degree, σ(L) and μ(L) represent the standard deviation and average value of the resource load vector L, respectively. When the balancing degree B is lower than the threshold, it is found that the resource allocation is unbalanced and needs to be adjusted. Based on these analysis results, the system identifies the adjustment direction of the specific resource to form the optimization direction data. When generating the parameter tuning strategy according to the optimization direction data, an incremental tuning and historical experience combination method is adopted. The incremental tuning adopts a gradient search strategy to make small adjustments to the parameters and observe the effect. The historical experience uses the past successful tuning records to find a reference scheme through similar scene matching. By comprehensively considering these two methods, the system generates the final tuning parameters and feeds them back to the resource elastic allocation scheme to complete the last link of the closed-loop optimization.

[0115] The embodiment of the application can accurately predict the space-time distribution characteristics of the communication network load by constructing a migration trajectory prediction model based on the biological migration tide data and the environmental parameter data, providing a scientific basis for resource allocation and improving the foresight and accuracy of resource allocation. By dividing the base station groups in the communication network coverage area, the resource demand characteristics of different areas can be analyzed differently, and a more detailed resource demand distribution map can be obtained, providing spatial accuracy for resource allocation and improving resource utilization efficiency. By prioritizing the resource scheduling instruction set, limited resources can be reasonably allocated under resource competition, solving the resource scheduling conflict problem and ensuring the communication service quality of key areas. At the same time, the base station parameter configuration instructions generated based on the optimal resource allocation strategy can accurately guide the resource configuration adjustment of each regional base station, improving the accuracy and execution efficiency of the configuration.

[0116] Embodiment 2;

[0117] Referring to Figure 2 As shown in the description of the embodiment, a cloud computing-based communication device resource dynamic allocation system is provided, which comprises:

[0118] The migration elasticity prediction module: obtains biological migration tide data, communication network load history data and environmental parameter data; constructs a migration trajectory prediction model according to the biological migration tide data and the environmental parameter data to obtain a migration trajectory prediction result; constructs a resource demand prediction model according to the communication network load history data and the migration trajectory prediction result to obtain resource load prediction data; generates a resource allocation strategy based on the resource load prediction data to obtain a resource elasticity allocation scheme;

[0119] The space-time demand analysis module: divides the communication network coverage area into base station groups to obtain base station group data; performs space-time analysis on resource demand based on the base station group data and the migration trajectory prediction result to obtain a resource demand distribution map; and performs resource utilization rate calculation on the resource demand distribution map to obtain a resource utilization rate index;

[0120] The dynamic scheduling optimization module: generates scheduling instructions for the cloud computing resource center based on the resource elasticity allocation scheme to obtain a resource scheduling instruction set; performs priority sorting on the resource scheduling instruction set to obtain a hierarchical scheduling sequence; performs resource allocation optimization according to the hierarchical scheduling sequence and the resource utilization rate index to obtain an optimal resource allocation strategy; and generates base station configuration parameters based on the optimal resource allocation strategy to obtain base station parameter configuration instructions;

[0121] The feedback tuning module: performs resource configuration adjustment on the base stations in each region according to the base station parameter configuration instructions to obtain base station resource real-time configuration data; performs running state monitoring on the base station resource real-time configuration data to obtain resource running state data; performs resource utilization efficiency evaluation according to the resource running state data to obtain resource performance evaluation results; performs feedback tuning based on the resource performance evaluation results to obtain tuning parameters; and applies the tuning parameters to the resource elasticity allocation scheme to realize closed-loop optimization and allocation of communication device resources.

[0122] The various modules are connected through wired and / or wireless means to realize data transmission between the modules.

[0123] Embodiment 3;

[0124] The embodiment discloses an electronic device, which comprises a memory, a processor, and a computer program stored in the memory and executable on the processor, and the processor implements the running mode of the cloud computing-based communication device resource dynamic allocation method provided above when executing the computer program.

[0125] Since the electronic device introduced in the embodiment is the electronic device used in the implementation of the communication device resource dynamic allocation method based on cloud computing in the embodiment of the application, the specific implementation of the electronic device of the embodiment and its various forms can be understood by those skilled in the art based on the communication device resource dynamic allocation method based on cloud computing in the embodiment of the application, so the implementation of the method in the embodiment of the application is not introduced in detail. As long as the electronic device used in the implementation of the communication device resource dynamic allocation method based on cloud computing in the embodiment of the application belongs to the scope of protection of the application.

[0126] The above formulas are dimensionless values calculated, the formulas are obtained by collecting a large amount of data to simulate a formula of the latest real situation, and the preset parameters and threshold values in the formula are set by those skilled in the art according to the actual situation.

[0127] The above is only the preferred embodiment of the application, the protection scope of the application is not limited to the above-mentioned embodiments, and any technical solution belonging to the idea of the application is within the protection scope of the application. It should be noted that for ordinary technical users in the technical field, some improvements and decorations without departing from the principles of the application are also considered as the protection scope of the application.

Claims

1. A method for dynamic allocation of communication equipment resources based on cloud computing, characterized in that, include: Step S1: Obtain biological migration tidal data, historical communication network load data, and environmental parameter data; A migration trajectory prediction model is constructed based on biological migration tidal data and environmental parameter data to obtain migration trajectory prediction results; a resource demand prediction model is constructed based on historical communication network load data and migration trajectory prediction results to obtain resource load prediction data; and a resource allocation strategy is generated based on the resource load prediction data to obtain a resource elastic allocation scheme. Step S2: Divide the communication network coverage area into base station groups to obtain base station group data; perform spatiotemporal analysis of resource demand based on the base station group data and migration trajectory prediction results to obtain a resource demand distribution map; calculate the resource utilization rate of the resource demand distribution map to obtain the resource utilization rate index. Step S3: Generate scheduling instructions for the cloud computing resource center based on the resource elastic allocation scheme to obtain a resource scheduling instruction set; prioritize the resource scheduling instruction set to obtain a hierarchical scheduling sequence; optimize resource allocation based on the hierarchical scheduling sequence and resource utilization indicators to obtain the optimal resource allocation strategy. Based on the optimal resource allocation strategy, base station configuration parameters are generated, and base station parameter configuration instructions are obtained. Step S4: Adjust the resource configuration of base stations in each area according to the base station parameter configuration instructions to obtain real-time base station resource configuration data; The operational status data of base station resources is monitored in real time to obtain resource operational status data. Resource utilization efficiency is evaluated based on resource operation status data to obtain resource effectiveness evaluation results; optimization parameters are obtained based on feedback of resource effectiveness evaluation results. The optimization parameters are applied to the resource elastic allocation scheme to achieve closed-loop optimization allocation of communication equipment resources; Step S1 includes: Step S11: Collect biological migration tidal data and perform spatiotemporal characteristic analysis to obtain a migration activity characteristic database; collect meteorological environmental data and seasonal change data to obtain environmental parameter data; Step S12: Obtain communication network area division data; construct network topology based on communication network area division data to obtain network topology data; collect communication load time-series data through network monitoring system to obtain historical communication network load data; Step S13: Construct a migration trajectory prediction model based on the migration activity feature library and environmental parameter data to obtain the migration trajectory prediction results; extract time-series features from the migration trajectory prediction results to obtain migration time-series features; Step S14: Based on the migration trajectory prediction results and migration time series characteristics, perform spatial trajectory prediction to obtain a migration path probability map; divide the migration path probability map into time windows to obtain time-series migration density distribution data. Step S15: Map the time-sharing migration density distribution data to the network topology data to obtain tidal load prediction data; construct a resource demand prediction model based on the historical data of communication network load and the tidal load prediction data to obtain resource load prediction data. Step S16: Perform peak identification and abrupt change analysis on the resource load forecast data to obtain the resource demand abrupt change index; Step S17: Perform resource pool distribution analysis based on network topology data to obtain a resource pool capability map; evaluate the schedulable resources of the resource pool capability map to obtain a resource schedulable matrix; Step S18: Generate a resource elastic allocation scheme based on the resource demand mutation index and the resource schedulable matrix.

2. The method for dynamic allocation of communication equipment resources based on cloud computing according to claim 1, characterized in that, The migration trajectory prediction model is constructed based on the migration activity feature database and environmental parameter data to obtain the migration trajectory prediction results, including: The species migration behavior pattern was classified into a migration activity feature database to obtain species migration type data. Environmental triggering factors are extracted based on environmental parameter data to obtain the migration triggering index; A deep learning migration prediction network was constructed based on species migration type data to obtain the basic model for migration prediction. The migration triggering index was used to optimize the model parameters of the basic migration prediction model, resulting in a migration trajectory prediction model. Migration trajectory prediction results are generated based on the migration trajectory prediction model.

3. The method for dynamic allocation of communication equipment resources based on cloud computing according to claim 1, characterized in that, Step S2 includes: Step S21: Analyze the geographical distribution of base stations based on the communication network area division data to obtain base station geographical location data; perform spatial cluster analysis on the base station geographical location data to obtain base station group data; Step S22: Calculate the communication coverage area based on the base station group data to obtain the coverage area data; Step S23: Based on the coverage area data and migration path probability map, perform regional overlap analysis to obtain the migration impact coefficient; Step S24: Perform resource capability assessment on the base station spatial group data to obtain group resource capability data; Step S25: Generate a resource demand distribution map based on the migration impact coefficient and group resource capacity data; Step S26: Perform time series analysis on the resource demand distribution map to obtain resource fluctuation cycle data; calculate the resource utilization rate based on the resource fluctuation cycle data to obtain the resource utilization rate index.

4. The method for dynamic allocation of communication equipment resources based on cloud computing according to claim 1, characterized in that, Step S3 includes: Step S31: Quantify the computing resource requirements according to the resource elastic allocation scheme to obtain computing resource allocation data; quantify the storage resource requirements according to the resource elastic allocation scheme to obtain storage resource allocation data; quantify the bandwidth resource requirements according to the resource elastic allocation scheme to obtain bandwidth resource allocation data. Step S32: Generate a resource scheduling instruction set based on computing resource allocation data, storage resource allocation data, and bandwidth resource allocation data; Step S33: Evaluate the timeliness of the resource scheduling instruction set based on the time-sharing migration density distribution data to obtain instruction timeliness level data; Step S34: Prioritize the resource scheduling instruction set based on the instruction timeliness level data to obtain a hierarchical scheduling sequence; Step S35: Perform cloud resource scheduling simulation based on hierarchical scheduling sequence and resource utilization index to obtain resource allocation prediction results; perform resource contention conflict detection on resource allocation prediction results to obtain conflict resolution strategies; Step S36: Generate the optimal resource allocation strategy based on the resource allocation prediction results and conflict resolution strategy; generate base station parameter configuration instructions based on the optimal resource allocation strategy.

5. The method for dynamic allocation of communication equipment resources based on cloud computing according to claim 4, characterized in that, Step S32 includes: Step S321: Perform virtual machine resource configuration conversion on the computing resource allocation data to obtain virtual machine configuration instructions; Step S322: Map storage units to the storage resource allocation data to obtain storage allocation instructions; Step S323: Perform network channel planning on the bandwidth resource allocation data to obtain bandwidth allocation instructions; Step S324: Standardize the instruction formats of virtual machine configuration instructions, storage allocation instructions, and bandwidth allocation instructions to obtain standardized scheduling instructions; Step S325: Integrate the standardized scheduling instructions into a resource scheduling instruction set.

6. The method for dynamic allocation of communication equipment resources based on cloud computing according to claim 4, characterized in that, Step S34 includes: Step S341: Construct an instruction priority evaluation model based on instruction timeliness level data to obtain the priority evaluation function; Step S342: Apply the priority evaluation function to the resource scheduling instruction set to obtain the instruction priority value; Step S343: Sort the resource scheduling instructions according to their priority values ​​to obtain priority sequence data; Step S344: Perform resource dependency analysis on the priority sequence data to obtain the dependency constraint matrix; adjust the priority sequence data according to the dependency constraint matrix to obtain the hierarchical scheduling sequence.

7. The method for dynamic allocation of communication equipment resources based on cloud computing according to claim 1, characterized in that, Step S4 includes: Step S41: Parse and convert the base station parameter configuration command to obtain the base station configuration parameter set; update the parameter configuration of each base station according to the base station configuration parameter set to obtain the base station parameter update data; Step S42: Verify the configuration validity of the base station parameter update data and obtain the configuration verification result; Step S43: Monitor base station resource performance based on configuration verification results to obtain resource performance data; collect resource usage status data from the resource performance data to obtain resource operation status data; Step S44: Calculate energy efficiency indicators based on resource operation status data to obtain energy efficiency assessment data; conduct service quality assessment on resource operation status data to obtain service quality indicators; Step S45: Conduct a comprehensive resource efficiency assessment based on energy efficiency assessment data and service quality indicators to obtain the resource efficiency assessment results; Step S46: Perform optimization space analysis on the resource efficiency assessment results to obtain optimization direction data; generate parameter tuning strategy based on the optimization direction data to obtain tuning parameters; feed the tuning parameters back to the resource elastic allocation scheme to achieve closed-loop optimization of resource allocation.

8. The method for dynamic allocation of communication equipment resources based on cloud computing according to claim 7, characterized in that, Step S45 includes: Step S451: Standardize the energy efficiency assessment data using multi-dimensional indicators to obtain standardized energy efficiency indicators; Step S452: Map service quality indicators to user experience to obtain experience satisfaction data; Step S453: Construct a resource efficiency assessment model, and weight and integrate standardized energy efficiency indicators and experience satisfaction data to obtain a comprehensive efficiency index; Step S454: Generate resource efficiency assessment results based on the comprehensive efficiency index.

9. A cloud computing-based dynamic resource allocation system for communication equipment, characterized in that, include: Migration resilience prediction module: acquires biological migration tidal data, historical communication network load data, and environmental parameter data; A migration trajectory prediction model is constructed based on biological migration tidal data and environmental parameter data to obtain migration trajectory prediction results; a resource demand prediction model is constructed based on historical communication network load data and migration trajectory prediction results to obtain resource load prediction data; and a resource allocation strategy is generated based on the resource load prediction data to obtain a resource elastic allocation scheme. Spatiotemporal demand analysis module: Divides the communication network coverage area into base station groups to obtain base station group data; performs spatiotemporal analysis of resource demand based on base station group data and migration trajectory prediction results to obtain a resource demand distribution map; calculates resource utilization rate on the resource demand distribution map to obtain resource utilization rate index; Dynamic scheduling optimization module: Based on the resource elastic allocation scheme, it generates scheduling instructions for the cloud computing resource center to obtain a resource scheduling instruction set; it prioritizes the resource scheduling instruction set to obtain a hierarchical scheduling sequence; and it optimizes resource allocation based on the hierarchical scheduling sequence and resource utilization indicators to obtain the optimal resource allocation strategy. Based on the optimal resource allocation strategy, base station configuration parameters are generated, and base station parameter configuration instructions are obtained. Feedback and optimization module: Adjusts the resource configuration of base stations in each area according to the base station parameter configuration instructions to obtain real-time base station resource configuration data; The operational status data of base station resources is monitored in real time to obtain resource operational status data. Resource utilization efficiency is evaluated based on resource operation status data to obtain resource performance evaluation results; feedback optimization is performed based on resource performance evaluation results to obtain optimization parameters; the optimization parameters are applied to the resource elastic allocation scheme to achieve closed-loop optimization allocation of communication equipment resources; The process involves acquiring biological migration tidal data, historical communication network load data, and environmental parameter data; constructing a migration trajectory prediction model based on the biological migration tidal data and environmental parameter data to obtain migration trajectory prediction results; and constructing a resource demand prediction model based on the historical communication network load data and migration trajectory prediction results to obtain resource load prediction data. Based on resource load forecast data, a resource allocation strategy is generated, resulting in a resource elastic allocation scheme, including: Collect biological migration tidal data and perform spatiotemporal characteristic analysis to obtain a migration activity characteristic database; collect meteorological and environmental data and seasonal variation data to obtain environmental parameter data; Obtain communication network area division data; construct network topology based on communication network area division data to obtain network topology data; collect communication load time-series data through network monitoring system to obtain historical communication network load data; A migration trajectory prediction model is constructed based on the migration activity feature database and environmental parameter data to obtain the migration trajectory prediction results; temporal features are extracted from the migration trajectory prediction results to obtain migration temporal features; Spatial trajectory prediction is performed based on the migration trajectory prediction results and migration time series characteristics to obtain a migration path probability map; the migration path probability map is divided into time windows to obtain time-series migration density distribution data. Time-sharing migration density distribution data is mapped to network topology data to obtain tidal load prediction data; a resource demand prediction model is constructed based on historical communication network load data and tidal load prediction data to obtain resource load prediction data. Peak identification and abrupt change analysis are performed on resource load forecast data to obtain the resource demand abrupt change index; Resource pool distribution analysis is performed based on network topology data to obtain a resource pool capability map; schedulable resources are evaluated on the resource pool capability map to obtain a resource schedulable matrix; A resource elastic allocation scheme is generated based on the resource demand mutation index and the resource schedulable matrix.

Images