5G Intelligent Optimization Digital Platform and Method Based on Tidal Effect

Abstract

The present invention relates to the field of mobile communication technologies, and discloses a 5G intelligent optimization digital platform and method based on tidal effects. The method includes: fine-grained analysis of crowd movement trajectories, correlation modeling of regional crowd activities and network traffic, extraction of time-varying network topology features, topological elasticity scoring and resource optimization allocation, and execution of adaptive network topology reconstruction. By analyzing anonymized user location data to identify crowd movement patterns, establishing a correlation model between regional activities and network traffic, extracting network topology features to identify the spatio-temporal variation law of traffic, calculating the topological elasticity score to achieve resource optimization allocation, and executing adaptive network topology reconstruction to ensure that the network structure pre-adapts to traffic changes, the present invention realizes the intelligent perception and prediction of tidal effects in 5G networks, adjusts the network topology in advance to adapt to traffic changes, significantly improves the network resource utilization efficiency, reduces the network congestion probability, improves the user experience, and has obvious technical advantages and application values.

Description

Technical Field

[0001] The present invention relates to the field of mobile communication technology, and more specifically, it relates to a 5G intelligent optimization digital platform and method based on tidal effect. Background Art

[0002] With the wide deployment and application of 5G networks, network traffic exhibits significant tidal effect characteristics. The so-called tidal effect refers to the phenomenon that due to the regular changes in population activities (such as commuting to and from work, holding large-scale events, etc.), network traffic shows periodic fluctuations in space and time. This tidal effect poses a huge challenge to network resource allocation: static network configurations are difficult to cope with dynamically changing traffic demands, resulting in an unbalanced state where resources are surplus in some areas while scarce in others.

[0003] Currently, the technologies for solving the tidal effect problem in 5G networks mainly rely on post-event response mechanisms, that is, resource adjustment is carried out after network congestion occurs. This passive optimization method leads to a decline in user experience and low network efficiency. Although some research has begun to focus on predictive network optimization, most methods still have the following deficiencies:

[0004] The correlation analysis between population activities and network traffic is coarse-grained, and the perception of the micro-population movement rules is insufficient;

[0005] Network topology optimization is mainly based on historical traffic data, lacking consideration of external social factors;

[0006] The resource adjustment strategy lacks foresight and cannot cope with traffic changes in advance;

[0007] Most of the optimization algorithms are static algorithms, with limited adaptability and learning ability.

[0008] Therefore, the existing technologies are difficult to effectively cope with the tidal effect challenges in 5G networks, and an intelligent optimization method is needed that can sense changes in population activities, predict network traffic distribution, adjust the network topology in advance to adapt to traffic changes, and achieve efficient utilization of network resources. Summary of the Invention

[0009] The present invention provides a 5G intelligent optimization digital platform and method based on tidal effect, which solves the technical problem that the network topology configuration in related technologies is too static and cannot effectively cope with the rapid change of network traffic distribution caused by tidal effect.

[0010] The present invention provides a 5G intelligent optimization method based on tidal effect, including the following steps:

[0011] Fine-grained analysis of population movement trajectories: Based on anonymized user location data, analyze the population movement patterns in the area and generate a movement trajectory prediction model;

[0012] Regional population activity - network traffic correlation modeling: Predict changes in regional population activities based on multi-source data, and generate network traffic prediction results;

[0013] The steps of regional population activity - network traffic correlation modeling specifically include: obtaining POI information, traffic data, and activity calendar data within the region, and performing data processing and fusion;

[0014] Based on multi-source data, construct a regional activity feature vector, including activity intensity, activity type, and number of participants information at specific time points;

[0015] Construct a dual-mode topology pre-adaptation system including normal mode and event mode, and define an event impact factor to trigger mode switching;

[0016] Based on historical traffic data and regional activity characteristics, train a traffic - activity correlation model, and output the predicted network traffic distribution;

[0017] Time-varying network topology feature extraction: Extract topology features based on the time-varying network graph model and graph embedding algorithm, and identify the spatio-temporal change rules of traffic distribution;

[0018] Topological elasticity scoring and resource optimization allocation: Calculate the adjustment priority and resource allocation ratio based on the topological scoring system to achieve a forward-looking network resource layout;

[0019] Adaptive network topology reconstruction execution: Based on the prediction results and scoring system, perform network topology adjustment in real time to ensure that the network structure pre-adapts to traffic changes.

[0020] Furthermore, the steps of the fine-grained population movement trajectory analysis specifically include:

[0021] Collect anonymized user location data at multiple time points within the region, and perform data cleaning, outlier removal, and spatio-temporal feature extraction;

[0022] Divide the region into grids, count the population density of each grid at different times, and construct a regional movement heat map;

[0023] Apply spatio-temporal trajectory clustering algorithms to analyze user movement trajectories, and identify typical movement patterns and key paths;

[0024] Based on the trajectory clustering results, establish a path traffic prediction function, and predict the traffic of each key path at specific time points in combination with time factors.

[0025] Furthermore, the steps of time-varying network topology feature extraction specifically include:

[0026] Based on historical traffic data, construct a time-varying network graph model to represent the dynamic characteristics of network topology changes over time;

[0027] Apply the time - series graph embedding algorithm to convert the time - varying network graph into a low - dimensional feature representation, and optimize the reconstruction error and temporal consistency;

[0028] Based on the graph embedding results, analyze the typical spatio - temporal patterns of traffic distribution, and extract periodic patterns and abnormal patterns;

[0029] Identify the set of critical links in the network, establish a time - series prediction model for link load, and generate a prediction graph of future network topology load distribution.

[0030] Furthermore, the steps of topological resilience scoring and resource optimization allocation specifically include:

[0031] Establish a network topology resilience scoring system to evaluate the adaptability of the network topology to traffic changes;

[0032] Based on the network topology graph and traffic prediction results, identify the set of critical links and the set of redundant paths;

[0033] For the predicted traffic changes, calculate the adjustment priority scores of each link in the topology;

[0034] Based on the link adjustment priority scores, calculate the optimal allocation ratio of network resources to solve the constrained optimization problem.

[0035] Furthermore, the execution steps of adaptive network topology reconstruction specifically include:

[0036] Construct a network resource re - allocation control system to convert the resource allocation plan into specific resource allocation instructions;

[0037] Based on the resource allocation instructions, achieve dynamic adjustment of link bandwidth, and adopt a smooth transition strategy to avoid service interruption caused by mutations;

[0038] Based on the link bandwidth adjustment results, optimize the network routing strategy to find a set of routing paths to optimize the overall network performance;

[0039] After implementing the topology reconstruction, establish a real - time monitoring and feedback mechanism, and trigger the adaptive adjustment process based on performance deviations.

[0040] Furthermore, the traffic - activity association model adopts a multi - layer perceptron structure, including:

[0041] The input layer receives the regional activity feature vector;

[0042] The hidden layer contains two sub - layers, and introduces a time attention mechanism to assign different weights to the activity features at different time points;

[0043] The output layer outputs the predicted traffic distribution, and the number of neurons is equal to the number of network area divisions.

[0044] Furthermore, the network topology elasticity score consists of the following metrics:

[0045] Link utilization balance, which evaluates the balance of the load distribution across the network links;

[0046] Number of redundant paths, which evaluates the availability of alternative paths in case of link failures in the network;

[0047] Connectivity of critical nodes, which evaluates the connection status of critical nodes in the network.

[0048] Furthermore, the resource allocation instructions include the following types:

[0049] Bandwidth adjustment instruction, which is used to adjust the bandwidth allocation of specific links, where the specific links refer to the high-priority adjustment links obtained based on the identified critical link set and the topology elasticity scoring system;

[0050] Routing update instruction, which is used to update the network routing table and forwarding rules;

[0051] Load balancing instruction, which is used to distribute traffic among multiple paths;

[0052] Resource recovery instruction, which is used to recover resources from low-load areas.

[0053] A 5G intelligent optimization digital platform based on the tidal effect, comprising:

[0054] Data acquisition module, which is used to obtain anonymized user location data and regional activity data;

[0055] Trajectory analysis module, which is used to analyze the crowd movement patterns and generate a mobile trajectory prediction model;

[0056] Traffic prediction module, which is used to establish a correlation model between regional crowd activities and network traffic and predict the traffic distribution;

[0057] Topology analysis module, which is used to extract the network topology features and identify the spatio-temporal variation rules of the traffic distribution;

[0058] Resource optimization module, which is used to calculate the topology elasticity score and the optimal allocation scheme of network resources;

[0059] Topology reconstruction module, which is used to perform network topology adjustment and conduct real-time monitoring and feedback;

[0060] Central control module, which is used to coordinate the work of each functional module and provide a unified management interface.

[0061] The beneficial effects of the present invention are as follows:

[0062] Achieved accurate network traffic prediction: Based on fine-grained population movement trajectory analysis and regional population activity-network traffic correlation modeling, the system accurately predicts the future network traffic distribution, and the prediction accuracy is improved by about 25% compared with traditional methods.

[0063] Improved the utilization efficiency of network resources: Through topology elasticity scoring and resource optimization allocation, accurate scheduling and optimization allocation of network resources are achieved, the network capacity utilization rate is increased by 40%, and resource waste and congestion problems are avoided.

[0064] Significantly reduced network congestion: The predictive topology adjustment system can sense and respond to traffic peaks in advance, reducing the congestion probability in high-load areas by 65% and greatly improving the user experience.

[0065] Enhanced network resilience: The adaptive network topology reconstruction execution system can dynamically respond to changes in the network environment, especially for unconventional events such as large-scale activities, and complete network topology adjustment 13-15 minutes in advance, significantly enhancing network resilience.

[0066] Improved operation and maintenance efficiency: The adaptive adjustment ability of the system changes network maintenance from passive response to active prevention, reduces the frequency of manual intervention, improves the network operation and maintenance efficiency by about 35%, and correspondingly reduces the operation cost. Description of the Drawings

[0067] Figure 1 is a flowchart of a 5G intelligent optimization method based on the tidal effect of the present invention. Detailed Embodiments

[0068] Now, the subject matter described herein will be discussed with reference to exemplary embodiments. It should be understood that discussing these embodiments is only to enable those skilled in the art to better understand and thus implement the subject matter described herein. Without departing from the scope of protection of the content of this specification, changes can be made to the functions and arrangements of the elements discussed. Each example can omit, substitute, or add various processes or components as needed. Additionally, the features described in some examples can also be combined in other examples.

[0069] Embodiment 1, a 5G intelligent optimization method based on the tidal effect, as Figure 1 shown, includes the following steps:

[0070] Step 1: Fine-grained population movement trajectory analysis: Based on anonymized user location data, analyze the population movement patterns in the area to generate a movement trajectory prediction model;

[0071] Specifically, it includes the following steps:

[0072] Anonymized user location data collection and preprocessing: Collect anonymized user location data at multiple time points in the area:

[0073]

[0074] where represents the location data set of the th user, including location coordinates and timestamps. The collected data is preprocessed, including data cleaning, outlier removal, and spatio-temporal feature extraction, to form a standardized location data set .

[0075] Construction of regional movement heat map: Based on the preprocessed location data , the region is divided into grids, and the population density of each grid at time is statistically analyzed to construct a regional movement heat map , where represents the population density value of grid at time . By analyzing the changes in the heat map at different time periods, the main characteristics and patterns of population flow within the region are identified.

[0076] Implementation of trajectory clustering algorithm: Apply the spatio-temporal trajectory clustering algorithm to analyze the user movement trajectories and identify typical movement patterns and key paths. Set the trajectory similarity calculation function , where and represent the th and the th trajectories respectively. Based on the trajectory similarity, use the density clustering algorithm to group similar trajectories into the same category to form a trajectory category set:

[0077]

[0078] where represents the trajectory set of the rd category.

[0079] Construction of path flow prediction model: Based on the trajectory clustering results, establish a path flow prediction function:

[0080]

[0081] where represents the path, represents the prediction time point, represents the number of users, represents that at time user selects path Probability. The prediction function analyzes historical movement trajectory data, combines time factors (weekday / non-weekday, peak period / non-peak period, etc.), and outputs the predicted traffic of each key path at a specific time point.

[0082] In the specific implementation, the path traffic prediction function is calculated based on a sequence prediction algorithm, where the user selects the path probability Implemented through a conditional probability model. For each path , first, count the frequency of users selecting this path at different time points (such as weekday morning rush hour, holidays, etc.) in the historical data to form the basic probability; then, combine the feature vector of the current time point: (including factors such as time type, weather conditions, surrounding activities, etc.), and obtain the real-time selection probability through weighted calculation:

[0083]

[0084] Where is the basic probability, is the time type, are the weight coefficients of each feature.

[0085] Step 2: Regional population activity - network traffic association modeling: Predict changes in regional population activities based on multi-source data and generate network traffic prediction results;

[0086] Specifically, it includes the following steps:

[0087] Multi-source data acquisition and fusion: Obtain multi-source data such as POI (Point of Interest) information, traffic data, activity calendars, etc. within the region, and process and fuse the data. POI information includes:

[0088]

[0089] Where represents the information of the th point of interest, including attributes such as location, type, capacity, etc.;

[0090] Traffic data includes:

[0091]

[0092] Where represents the information of the th traffic route;

[0093] Activity calendars include:

[0094]

[0095] Where represents the Information about a planned activity, including attributes such as time, location, and expected scale.

[0096] Construction of regional activity feature vectors: Based on multi-source data, construct regional activity feature vectors , where represents the region, represents the time point. The feature vector contains information such as the activity intensity, activity type, and number of participants in the region at a specific time point, and is formally expressed as:

[0097]

[0098] where represents the value of the th activity feature dimension.

[0099] Construction of a dual-mode topology pre-adaptation system: Construct a dual-mode topology pre-adaptation system that includes a normal mode and an event mode. In the normal mode, predict traffic changes based on historical patterns; in the event mode, trigger topology adjustment through event impact factors. Define the event impact factor:

[0100]

[0101] where represents the weight coefficient of the th activity, represents the activity scale, represents the activity duration. When the value exceeds the preset threshold , the system switches from the normal mode to the event mode.

[0102] Training of the traffic-activity correlation model: Based on historical traffic data and regional activity characteristics, train the traffic-activity correlation model , which adopts a multi-layer perceptron structure. The input is the regional activity feature vector , and the output is the predicted network traffic distribution . The training objective of the model is to minimize the mean square error between the predicted traffic and the actual traffic:

[0103]

[0104] where represents the actual network traffic in the region at time .

[0105] In a specific implementation, the traffic-activity correlation model adopts a three-layer structure: The input layer receives the regional activity feature vector , containing 32 neurons; the hidden layer contains two sub-layers, with 64 and 32 neurons respectively, using the ReLU activation function; the output layer outputs the predicted traffic distribution , the number of neurons is equal to the number of network area divisions. To improve the model's ability to process time series data, a time attention mechanism is also introduced in the hidden layer, which assigns different weights to the activity features at different time points. The calculation formula is:

[0106]

[0107] where 、 、 are the query, key, and value matrices respectively, obtained by linear transformation of the input features, is the dimension of the key.

[0108] Step 3: Extraction of time-varying network topology features: Based on the time-varying network graph model and graph embedding algorithm, extract topology features to identify the spatio-temporal variation laws of traffic distribution;

[0109] Specifically, it includes the following steps:

[0110] Construction of time-varying network graph model: Based on historical traffic data, construct a time-varying network graph model , where represents the set of network nodes, represents time is the edge set at time represents time is the edge weight matrix at time represents time node and node the link load between. This model can represent the dynamic characteristics of network topology changing with time and capture the spatio-temporal distribution pattern of network traffic under tidal effects.

[0111] Application of graph embedding algorithm: Apply the time series graph embedding algorithm to convert the time-varying network graph into a low-dimensional feature representation. For each time point , calculate the node embedding matrix , where represents the embedding vector of node at time .

[0112] The optimization objective of the embedding algorithm is: minimize the reconstruction error:

[0113]

[0114] At the same time, maximize the temporal consistency:

[0115]

[0116] Analysis and extraction of traffic spatio-temporal patterns: Based on the graph embedding results, apply clustering algorithms to analyze the typical spatio-temporal patterns of traffic distribution. Divide the time series into time windows, calculate the feature matrix for each window , apply the dynamic time warping algorithm to calculate the similarity of features in different time windows, and extract periodic patterns and abnormal patterns. Finally, obtain the spatio-temporal variation law of traffic distribution, including peak period distribution, tidal effect characteristics, etc.

[0117] Identification of critical links and prediction of change trends: Based on the spatio-temporal pattern analysis results, identify the set of critical links in the network:

[0118]

[0119] These links play a key role in the change of traffic distribution. For each critical link, establish a time series prediction model of link load , and predict the link load situation after time. Combine the prediction results of Step 1 and Step 2 to generate a comprehensive prediction model and output a prediction map of the future network topology load distribution .

[0120] Step 4: Topological resilience scoring and resource optimization allocation: Calculate the adjustment priority and resource allocation ratio based on the topological scoring system to achieve a forward-looking network resource layout;

[0121] Specifically, it includes the following steps:

[0122] Construction of the topological resilience scoring system: Establish a network topological resilience scoring system and define a topological resilience scoring function , which is used to evaluate the adaptability of the network topology to traffic changes. The topological resilience score consists of multiple indicators, including link utilization balance , the number of redundant paths , critical node connectivity , etc. The calculation formula of the topological resilience score is:

[0123]

[0124] where , , are weight coefficients, and .

[0125] Identification of critical links and redundant paths: Based on the network topology graph and traffic prediction results, identify the set of critical links:

[0126]

[0127] Redundant path set:

[0128]

[0129] For each critical link , calculate its importance score , which is based on factors such as the traffic-carrying capacity of the link, the importance of the connected nodes, the number of alternative paths, etc.; for each redundant path , calculate its availability score , which takes into account factors such as path length, current load, available bandwidth, etc.

[0130] Topology adjustment priority score calculation: For the predicted traffic changes, calculate the adjustment priority scores of each link in the topology. Define the link adjustment priority score function:

[0131]

[0132] where represents the predicted future link traffic, represents the current link traffic, represents the link capacity, represents the link importance. The higher the priority score, the more the link needs to be adjusted.

[0133] Optimal resource allocation ratio optimization: Based on the link adjustment priority scores, calculate the optimal allocation ratio of network resources.

[0134] Define the resource allocation problem as a constrained optimization problem: Maximize , where represents the adjusted network topology.

[0135] The constraint conditions include the total resource limit , and the capacity requirements of each link .

[0136] Use the Lagrange multiplier method to solve this optimization problem to obtain the optimal resource allocation plan , where represents the amount of resources allocated to the th link.

[0137] Step 5: Adaptive network topology reconstruction execution: Based on the prediction results and the scoring system, perform network topology adjustment in real time to ensure that the network structure pre-adapts to traffic changes;

[0138] Specifically, it includes the following steps:

[0139] Implementation of the network resource reallocation control system: Construct a network resource reallocation control system, which receives the resource allocation plan calculated in step 4 , and convert it into specific resource allocation instructions. Define the resource allocation instruction set:

[0140]

[0141] Each instruction contains information such as the target link, adjustment type, adjustment parameters, etc.

[0142] Link bandwidth dynamic adjustment algorithm: Based on the resource allocation instructions, implement the dynamic adjustment of the link bandwidth.

[0143] Define the bandwidth adjustment function , where represents the target link, represents the bandwidth adjustment amount.

[0144] The bandwidth adjustment process considers the current load status of the link and adopts a smooth transition strategy to avoid service interruption caused by mutations. The mathematical expression of the adjustment process is:

[0145]

[0146] where is a smoothing function to ensure the gradualness of the bandwidth adjustment.

[0147] Routing strategy optimization algorithm: Based on the link bandwidth adjustment results, optimize the network routing strategy.

[0148] Define the routing strategy optimization problem: Under the updated topology, find a set of routing paths , so that the overall network performance is optimal.

[0149] The optimization objectives include minimizing the end-to-end delay, maximizing the throughput, balancing the load, etc. Use a multi-objective optimization algorithm to solve this problem and obtain the optimized routing strategy .

[0150] Topology reconstruction feedback and adaptive adjustment: After implementing the topology reconstruction, establish a real-time monitoring and feedback mechanism to collect actual network performance data and the deviation from the expected performance . Define the performance deviation vector . When the deviation exceeds the threshold , trigger the adaptive adjustment process. The adaptive adjustment process is based on the reinforcement learning algorithm, learns the mapping relationship between the environmental state and the optimal adjustment strategy, and continuously optimizes the topology reconstruction decision model.

[0151] Through the above sub-steps, the adaptive network topology reconstruction execution system realizes the intelligent allocation of network resources and the dynamic optimization of routing strategies, can accurately capture the characteristics of the tidal effect and make topology adjustments in advance, and ensures the stability and efficiency of network performance during the process of traffic changes.

[0152] Technical effects of this embodiment:

[0153] Accurate network traffic prediction ability: Through fine-grained analysis of population movement trajectories and correlation modeling of regional population activities-network traffic, the system can accurately predict the future network traffic distribution, and the prediction accuracy is about 25% higher than that of traditional methods, providing a reliable basis for network topology optimization.

[0154] Efficient utilization of network resources: Based on topology elasticity scoring and resource optimization allocation, accurate scheduling and optimized allocation of network resources are achieved, the network capacity utilization rate is increased by 40%, and resource waste and congestion problems are avoided.

[0155] Significant congestion reduction effect: Through the predictive topology adjustment system, the system can perceive and respond to traffic peaks in advance, and the congestion probability in high-load areas is reduced by 65%, greatly improving the user experience.

[0156] Enhanced network resilience: The adaptive network topology reconstruction execution system can dynamically respond to changes in the network environment. Especially for unconventional events such as large-scale activities, the system can complete network topology adjustment 13 - 15 minutes in advance, significantly enhancing network resilience.

[0157] Improved intelligent operation and maintenance efficiency: The adaptive adjustment ability of the system changes network maintenance from passive response to active prevention, reduces the frequency and degree of manual intervention, the network operation and maintenance efficiency is increased by about 35%, and the operation cost is correspondingly reduced.

[0158] An application example of Embodiment 1 is as follows:

[0159] This method has been practically applied to the 5G network optimization around a certain international convention and exhibition center. The convention and exhibition center covers an area of 120,000 square meters, has 8 main exhibition halls and multiple conference centers, and is an important venue for large-scale exhibitions, international conferences and art performances. The 5G network in this area faces the following challenges:

[0160] Highly dynamic traffic distribution: Different exhibitions and activities attract different types of people, and the concentration of the flow of people changes rapidly according to the activity time arrangement, and the typical tidal effect characteristics are obvious.

[0161] Diversity of traffic demands: During commercial exhibitions, data transmission and video conferencing are mainly carried out; during art performances, high-definition live broadcasts and short video sharing are mainly carried out; during international conferences, web conferencing and data synchronization are mainly carried out, and different activities have large differences in the requirements for network service quality.

[0162] Difficulty in peak prediction: In the short period before and after the start and end of large-scale events, the traffic within the area will increase suddenly, and the traditional static network topology cannot effectively handle it, resulting in a decline in user experience.

[0163] This application scenario covers the following network structure: A total of 32 5G base stations and 102 cells are deployed to form a network covering the convention and exhibition center and its surrounding areas. During peak periods, the number of simultaneously connected users can reach 150,000, and the peak network traffic can reach 48 Gbps. Traditional network optimization methods have experienced congestion many times during large-scale events, and the user complaint rate is high. There is an urgent need to adopt new intelligent optimization methods.

[0164] Deployment process:

[0165] In response to the actual needs of the convention and exhibition center, we carried out system deployment and implementation according to the five core steps of Implementation Method 1.

[0166] Fine-grained analysis of crowd movement trajectories:

[0167] Data collection and preprocessing: 15 data collection points are set in the convention and exhibition center area to collect anonymized user location data based on WiFi probe and base station location information. To ensure user privacy, hash coding and data desensitization technologies are used to process user identifiers. The collection period is 10 seconds per time, and data for 3 months is continuously collected as the training set, including crowd trajectory information during different types of activities.

[0168] Construction of mobile heat maps in the convention and exhibition center area: The convention and exhibition center area is divided into 50×50 grid cells, and each cell covers an area of about 50 square meters. Based on the collected location data, a time-segmented crowd density heat map is constructed. An example of the heat map is shown in Table 1:

[0169] Table 1: Example of crowd density heat maps in the convention and exhibition center at different times (partial data)

[0170]

[0171] Trajectory clustering and identification of key paths: Based on the collected data, 8 main pedestrian flow paths are identified, including the paths from the subway station to each exhibition hall, the connecting paths between exhibition halls, and the paths from exhibition halls to the dining area. An identifier (P1 - P8) is set for each path, and the path selection probabilities at different times are analyzed. The traffic statistics of key paths are shown in Table 2:

[0172] Table 2: Traffic statistics of key paths at different times (weekdays)

[0173]

[0174] Application of Path Flow Prediction Model: The path flow prediction model is trained based on historical data, which comprehensively considers time factors, activity types, and historical flow patterns. The model adopts a tree structure. First, it determines the current time period type (weekday / weekend, peak / off-peak), then further determines whether there are special activities, and finally predicts the flow of each path based on the historical data under the corresponding conditions. Specifically for large-scale exhibitions, a mapping relationship between exhibition types and the distribution of the flow of people is established. For example, the difference in the distribution of the flow of people between the International Consumer Electronics Show and the Medical Devices Show reaches 35%.

[0175] Modeling of the Association between Regional Population Activities and Network Traffic:

[0176] Multi-source Data Acquisition and Fusion: The exhibition schedule, ticket data, booth distribution map, and surrounding traffic information of the convention and exhibition center are obtained to establish a multi-dimensional database. At the same time, network traffic data during different activities are extracted from the historical network monitoring system to establish an activity-flow mapping relationship. An example of the data is shown in Table 3:

[0177] Table 3: Example of Association Data between Activities and Network Traffic

[0178]

[0179] Construction of Activity Feature Vectors: Feature vectors are constructed for each type of activity, including information such as time, scale, type, and population attributes. For example, the feature vector of a certain technology exhibition is: [1 (weekday), 3 (number of continuous days), 35000 (scale), 0.7 (proportion of business people), 0.2 (proportion of technical personnel), 0.1 (others)].

[0180] Deployment of the Dual-mode Topology Pre-adaptation System: The dual-mode system is deployed on the network management platform. In the normal mode, traffic prediction and fine-tuning are performed at 15-minute intervals; in the event mode, more frequent predictions (once every 5 minutes) and larger adjustments are made according to the event impact factor. The system sets the event impact factor threshold to 0.65, and automatically switches to the event mode when a large-scale activity (such as the opening of an exhibition with more than 15,000 participants) is predicted.

[0181] Training and Verification of the Traffic-Activity Association Model: A neural network model is trained based on historical data. The input layer contains 32 nodes (corresponding to activity features), two hidden layers (64 nodes and 32 nodes), and the output layer is 102 nodes (corresponding to the traffic prediction of each cell). The model is trained through 5,000 rounds of iteration, and the prediction accuracy on the validation set reaches 88.5%. The model can identify the network traffic characteristics of different exhibition types. For example, the data download volume of medical exhibitions is 35% higher than that of consumer electronics exhibitions, while the real-time streaming media transmission of consumer electronics exhibitions is 52% higher than that of medical exhibitions.

[0182] Extraction of Time-varying Network Topology Features:

[0183] Construction of Time-Varying Network Graph Model: Based on network topology information and historical traffic data, construct a time-varying network graph model for the convention and exhibition center area. This model consists of 32 nodes (base stations) and 86 edges (links), and the edge weights (link loads) are updated every 15 minutes.

[0184] Graph Embedding and Feature Extraction: Apply the temporal graph neural network algorithm to convert the network topology graph into a low-dimensional feature representation (with a dimension of 16). Based on the extracted features, identify 4 typical network state patterns: ordinary working day pattern, exhibition opening day pattern, regular exhibition day pattern, and large-scale cultural and art activity pattern.

[0185] Analysis of Traffic Spatiotemporal Patterns: Analyze the traffic distribution during different activities within three months and extract the spatiotemporal variation rules. It is found that from 30 minutes before the exhibition opening to 60 minutes after the opening, the traffic growth rate at the south entrance reaches 300%; while from 30 minutes before the exhibition ends to 45 minutes after the end, the traffic growth rates in the exit area and the direction of the transportation hub reach 250%. Based on these rules, a targeted pre-adjustment strategy is designed.

[0186] Identification of Key Links and Prediction of Change Trends: Identify 12 key links that bear the main traffic changes in the tidal effect. Establish a prediction model for each key link, with an average prediction lead time of 18 minutes and an accuracy rate of 83.7%. An example of the key link load prediction results is shown in Table 4:

[0187] Table 4: Example of Key Link Load Prediction Results (Exhibition Opening Day of Science and Technology Exhibition)

[0188]

[0189] Topological Resilience Scoring and Resource Optimization Allocation:

[0190] Implementation of Topological Resilience Scoring System: Establish a topological resilience scoring system, and the scoring indicators include the balance degree of link utilization rate (weight 0.4), availability of redundant paths (weight 0.3), and connectivity of key nodes (weight 0.3). Score the current network topology. The score is 0.72 under normal conditions, while if no adjustment is made during large-scale activities, the score will drop to 0.48, which is lower than the system-set safety threshold of 0.65.

[0191] Calculation of Key Link and Resource Adjustment Plans: For the 12 key links, calculate the resource adjustment priorities and ratios. Part of the priority ranking results is shown in Table 5:

[0192] Table 5: Priority Ranking of Key Link Adjustments (Before the Opening of the Science and Technology Exhibition)

[0193]

[0194] Optimized Calculation of Resource Allocation: Based on link adjustment priorities and network resource constraints, calculate the optimal resource allocation plan. The total resource constraint is 120 units, which needs to be allocated among 32 base stations. The optimization goal is to maximize the network resilience score while ensuring that all predicted loads can be met. The final calculation results show that by dynamically adjusting 30% of the resources in low-load areas to high-load areas, the network resilience score can be maintained above 0.78, ensuring good performance even during the peak period of the exhibition.

[0195] Execution of Adaptive Network Topology Reconfiguration:

[0196] Deployment of Network Resource Reallocation Control System: Deploy a resource reallocation control system on the core network management platform. This system can receive the optimized resource allocation plan and convert it into device-level configuration instructions. The system supports three types of operations: bandwidth adjustment, routing optimization, and load balancing, with the minimum adjustment granularity being 5% resource units.

[0197] Implementation of Bandwidth Dynamic Adjustment: Implement bandwidth dynamic adjustment for critical links. The adjustment adopts a smooth transition strategy and is gradually completed in a 5-minute cycle to avoid interrupting current services. During the test on the opening day of the technology exhibition, the system successfully increased the bandwidth of 3 critical links at the south entrance by 40%, while reducing the bandwidth of 5 links in low-load areas by 15% for resource recovery.

[0198] Execution of Routing Policy Optimization: Based on the predicted traffic distribution, optimize the network routing policy. The system calculates 20 alternative paths and sets the trigger threshold at 85% of the link load. During actual operation, when the load of the L3 link reaches 83%, the system triggers the routing adjustment in advance and diverts approximately 25% of the traffic to the alternative paths, effectively avoiding congestion.

[0199] Feedback Mechanism and Adaptive Adjustment: Establish a real-time monitoring system to collect actual network performance data every 5 minutes and compare it with the expected performance. When the deviation exceeds 15%, trigger the adaptive adjustment process. In the first month of system operation, the number of times of triggering adaptive adjustment was 12, and it decreased to 3 times in the third month, indicating that the prediction ability of the system is continuously improving.

[0200] Verification of Application Effect:

[0201] After the actual deployment of this implementation method in the convention and exhibition center for 6 months, evaluate its application effect through multiple indicators, verifying the technical advantages of the 5G intelligent optimization digital platform and method based on the tidal effect.

[0202] Effect of Network Performance Improvement:

[0203] Before and after the system deployment, for exhibition activities of the same scale and type, the core network performance indicators were compared, and the results are shown in Table 6 as follows:

[0204] Table 6: Comparison of Network Performance Before and after System Deployment (Opening Day of Large-scale Technology Exhibition)

[0205]

[0206] As can be seen from Table 6, after the system deployment, the network performance has been significantly improved. In particular, the number of network congestion occurrences has decreased by 88.9%, and the video service stuttering rate has decreased by 77.3%. These indicators directly reflect the improvement of user experience. At the same time, the average link utilization rate has increased by 47.7%, indicating that the utilization efficiency of network resources has been significantly improved.

[0207] Verification of Prediction Accuracy and Early Warning Time:

[0208] To verify the prediction accuracy and early warning ability of the system, during the 6-month operation period, the prediction results and actual traffic data of different types of activities were collected, and the prediction accuracy rate and average early warning lead time were statistically analyzed, as shown in Table 7:

[0209] Table 7: Statistics of Prediction Accuracy and Early Warning Time for Different Activity Types

[0210]

[0211] As can be seen from Table 7, the system has the highest prediction accuracy rate for regular exhibitions, reaching 92.3%, and also has a relatively high prediction accuracy for the opening of large-scale exhibitions and concerts. Even for emergencies, the prediction accuracy rate reaches 76.8%, and the average early warning lead time is 13.2 minutes, which is sufficient for network adjustment. The comprehensive average prediction accuracy rate of the system is 88.5%, the topology adjustment hit rate is 86.3%, and the average early warning lead time is 20.2 minutes, verifying the prediction ability and early warning effect of the system.

[0212] Analysis of Resource Utilization Efficiency:

[0213] Before and after the system deployment, the changes in resource utilization efficiency are as Figure 1 shown (due to patent document restrictions, a table is used here to replace the chart for display):

[0214] Table 8: Comparison of Network Resource Utilization Distribution (Unit: Number of Base Stations)

[0215]

[0216] As can be seen from Table 8, before the system was deployed, resource utilization showed a polarization. The utilization rate of some base stations was less than 20%, while on peak days, 3 base stations were overloaded. After the system was deployed, the resource utilization was more reasonably distributed. The utilization rates of most base stations were concentrated in the range of 40% - 80%, which not only avoided resource waste but also eliminated the overloaded situation. This result indicates that the system can effectively achieve dynamic allocation of resources and improve the overall resource utilization efficiency.

[0217] Verification of user experience improvement:

[0218] By conducting user satisfaction surveys and analyzing network operation data, the improvement effect of the system on user experience was evaluated. 1000 user satisfaction surveys were conducted before and after the deployment, and the results are shown in Table 9:

[0219] Table 9: Results of user satisfaction surveys before and after system deployment

[0220]

[0221] As can be seen from Table 9, after the system was deployed, user satisfaction increased significantly, especially in terms of network congestion perception, with an increase of 57.7%. The overall satisfaction score increased from 3.0 to 4.3, with an increase of 43.3%, indicating that the system has a significant improvement effect on user experience.

[0222] In addition, during the 6 - month operation of the system, the number of network - related complaints decreased from an average of 68 per month to 12 per month, a decrease of 82.4%. Especially during large - scale exhibitions, the relevant complaints decreased from an average of 35 per exhibition to 5 per exhibition, a decrease of 85.7%.

[0223] Verification of the system's long - term stability:

[0224] To verify the long - term stability and adaptive ability of the system, during the 6 - month operation period, monthly change data of the system's key indicators were collected, as shown in Table 10:

[0225] Table 10: Monthly change data of the system's key indicators

[0226]

[0227] As can be seen from Table 10, as the system operation time increased, the prediction accuracy rate increased from 83.5% to 92.3%, the system response time decreased from 8.5 seconds to 5.0 seconds, and the number of adaptive adjustments decreased from 32 times per month to 8 times per month. This indicates that the system's self - learning ability is continuously improving and its adaptability to the network environment is getting stronger. At the same time, the number of system failures decreased month by month, and zero - failure operation was achieved in the 5th and 6th months, proving the long - term stability of the system.

[0228] Through the above multi-dimensional effect verification, it is fully proved that the 5G intelligent optimization digital platform and method based on the tidal effect provided by this embodiment have technical feasibility and superiority in practical applications. This method can effectively address the problem of rapid changes in network traffic distribution caused by the tidal effect, significantly improve network performance and user experience, and has important application value.

[0229] The above describes the embodiments of the present invention, but these embodiments are not limited to the above specific implementation manners. The above specific implementation manners are merely illustrative rather than restrictive. Under the inspiration of this embodiment, those of ordinary skill in the art can also make more forms of equivalent embodiments, all of which fall within the protection scope of this embodiment.

Claims

1. A 5G intelligent optimization method based on tidal effect, characterized in that: The following steps are involved: Fine-grained crowd movement trajectory analysis: Analyze the movement patterns of people in the area based on anonymized user location data and generate a movement trajectory prediction model; Regional crowd activity-network traffic correlation modeling: predict regional crowd activity changes based on multi-source data and generate network traffic prediction results; The steps of regional crowd activity-network traffic association modeling specifically include: obtaining POI information, traffic data, and activity calendar data in the region, and performing data processing and fusion; Based on multi-source data, a regional activity feature vector is constructed, which contains information on activity intensity, activity type, and number of participants at a specific time point; Build a dual-mode topology pre-adaptive system including normal mode and event mode, and define event impact factors to trigger mode switching; Based on historical traffic data and regional activity characteristics, the traffic-activity association model is trained to output the predicted network traffic distribution; Time-varying network topology feature extraction: Extract topology features based on the time-varying network graph model and graph embedding algorithm to identify the temporal and spatial variation patterns of traffic distribution; Topology elasticity scoring and resource optimization allocation: Based on the topology scoring system, the priority and resource allocation ratio are calculated and adjusted to achieve forward-looking network resource layout; Adaptive network topology reconstruction execution: Based on the prediction results and scoring system, network topology adjustments are performed in real time to ensure that the network structure is pre-adapted to traffic changes.

2. According to a 5G intelligent optimization method based on tidal effect according to claim 1, it is characterized in that: The steps of fine-grained crowd movement trajectory analysis include: Collect anonymized user location data at multiple time points in the region, perform data cleaning, remove outliers, and extract spatiotemporal features; Divide the area into grids, count the crowd density of each grid at different times, and construct a regional mobility heat map; Apply spatiotemporal trajectory clustering algorithm to analyze user movement trajectories and identify typical movement patterns and key paths; Based on the trajectory clustering results, a path flow prediction function is established, and the flow of each key path at a specific time point is predicted by combining the time factor.

3. According to a 5G intelligent optimization method based on tidal effect according to claim 2, it is characterized in that: The time-varying network topology feature extraction step specifically includes: Based on historical traffic data, a time-varying network graph model is constructed to represent the dynamic characteristics of network topology changing over time; Apply the time-series graph embedding algorithm to convert the time-varying network graph into a low-dimensional feature representation, optimizing the reconstruction error and temporal consistency; Based on the graph embedding results, the typical spatiotemporal patterns of traffic distribution are analyzed to extract periodic patterns and abnormal patterns. Identify the critical link set in the network, establish a link load timing prediction model, and generate a future network topology load distribution prediction map.

4. According to claim 3, a 5G intelligent optimization method based on tidal effect is characterized in that: The topology elasticity scoring and resource optimization allocation steps specifically include: Establish a network topology resilience scoring system to evaluate the adaptability of network topology to traffic changes; Based on the network topology and traffic prediction results, identify the critical link set and redundant path set; Calculate the adjustment priority score of each link in the topology based on the predicted traffic changes; Based on the link adjustment priority score, the optimal allocation ratio of network resources is calculated to solve the constrained optimization problem.

5. According to claim 4, a 5G intelligent optimization method based on tidal effect is characterized in that: The steps for executing adaptive network topology reconstruction specifically include: Build a network resource redistribution control system to convert resource allocation plans into specific resource allocation instructions; Based on resource allocation instructions, the link bandwidth is dynamically adjusted, and a smooth transition strategy is adopted to avoid service interruption caused by sudden changes; Based on the link bandwidth adjustment results, the network routing strategy is optimized to find a set of routing paths to optimize the overall network performance; After implementing topology reconstruction, a real-time monitoring and feedback mechanism is established to trigger the adaptive adjustment process based on performance deviations.

6. The 5G intelligent optimization method based on tidal effect according to claim 5 is characterized in that: The traffic-activity association model adopts a multi-layer perceptron structure, including: The input layer receives the regional activity feature vector; The hidden layer contains two sub-layers and introduces a temporal attention mechanism to assign different weights to activity features at different time points; The output layer outputs the predicted traffic distribution, and the number of neurons is equal to the number of network area divisions.

7. The 5G intelligent optimization method based on tidal effect according to claim 6 is characterized in that: The network topology resilience score is composed of the following indicators: Link utilization balance, which evaluates the balance of load distribution on each link in the network; The number of redundant paths, which evaluates the availability of alternative paths in the network in the event of a link failure; Key node connectivity, evaluates the connection status of key nodes in the network.

8. The 5G intelligent optimization method based on tidal effect according to claim 7 is characterized in that: The resource allocation instructions include the following types: A bandwidth adjustment instruction, used to adjust bandwidth allocation of a specific link, wherein the specific link refers to a high priority adjustment link derived from an identified critical link set and a topology resilience scoring system; Routing update instructions, used to update network routing tables and forwarding rules; Load balancing instructions, used to distribute traffic among multiple paths; Resource reclaim instructions are used to reclaim resources from low-load areas.

9. A 5G intelligent optimization digital platform based on tidal effect, which is used to execute a 5G intelligent optimization method based on tidal effect according to any one of claims 1 to 8, comprising: Data collection module, used to obtain anonymized user location data and regional activity data; Trajectory analysis module, used to analyze crowd movement patterns and generate movement trajectory prediction models; Traffic prediction module, used to establish a correlation model between regional crowd activities and network traffic and predict traffic distribution; Topology analysis module, used to extract network topology features and identify the temporal and spatial variation patterns of traffic distribution; Resource optimization module, used to calculate the topology elasticity score and the optimal allocation of network resources; Topology reconstruction module, used to perform network topology adjustment and conduct real-time monitoring and feedback; The central control module is used to coordinate the work of each functional module and provide a unified management interface.

Images