Big data network communication coordination method and device, and storage medium

Abstract

The invention relates to the technical field of big data network communication, and discloses a big data network communication coordination method and device, and a storage medium. The method comprises the following steps: acquiring original communication data, cleaning and classifying the original communication data, and generating an initial coordination scheme; generating a parallel influence data set based on entropy quantization and a PageRank algorithm; generating a bandwidth demand value through normalization processing and dynamic correction; executing multi-level threshold judgment and selecting a content optimization or data compression strategy; integrating the initial scheme and the correction parameters to carry out strategy fusion and simulation evaluation, and generating a final coordination scheme; and monitoring a resource state through a protocol and generating a feedback optimization parameter. According to the invention, the data processing efficiency and the resource coordination capability of network communication are improved, and the system adaptability and the transmission stability are enhanced.

Description

Technical Field

[0001] This invention relates to the field of big data network communication technology, and in particular to a big data network communication coordination method, device and storage medium. Background Technology

[0002] In the field of big data network communication technology, existing solutions related to big data network communication coordination methods, devices, and storage media typically employ static priority allocation or fixed bandwidth scheduling strategies. These solutions suffer from limitations such as dynamic data peak response lag, inaccurate parallel address resource matching, and a single dimension for candidate scheme evaluation. Existing methods often rely on single-dimensional parameter sorting or offline strategy pre-setting, which can easily lead to bandwidth utilization fluctuations and network resource contention in scenarios with multiple senders / receivers communicating concurrently, making it difficult to meet the stable coordination requirements of high-throughput data transmission. Regarding the joint processing of the initial coordination scheme and coordination correction parameters, existing technologies generally lack entropy quantification mechanisms for priority weights and dynamic peak value calculation capabilities. This makes it difficult to form a continuous processing link from data peak sequence—address influence factor value—parallel influence dataset in stages such as parallel data volume sorting, address influence factor matching, and transmission delay simulation. Consequently, the generation of candidate coordination scheme sets and the determination of scheme priority sequences lack systematic dynamic support. Summary of the Invention

[0003] To address the aforementioned technical problems, this invention provides a big data network communication coordination method, comprising: The system acquires raw communication data and performs data cleaning by removing invalid characters, unifying the address format to dotted decimal or lowercase domain names, eliminating duplicate data packets, and filtering abnormal content. It then uses distance measurement algorithms such as edit distance, hash mapping, or machine learning classification models to perform similarity classification and generates an initial coordination scheme by matching the address classification results with a preset benchmark scheme library. Based on the initial coordination scheme, priority weight entropy quantification and dynamic peak calculation are performed to generate a data peak sequence. Parallel data volume sorting and matching with a preset impact factor table are executed, and the PageRank algorithm is used for associated storage processing to generate a parallel impact dataset. The parallel impact dataset is processed by min-max normalization or Z-score standardization, and multiplication and dynamic scene correction based on machine learning model are performed to generate dynamic bandwidth requirement values. Perform multi-level threshold judgment and threshold comparison, select the video encoding parameter adjustment scheme in the content format optimization scheme or the redundant data identification based on content fingerprint in the duplicate data compression strategy, perform parameterized encapsulation, and generate coordinated correction parameters; The initial coordination scheme and coordination correction parameters are integrated, and a rule engine and policy weight mechanism are used for policy fusion processing. Discrete event simulation technology is used to simulate and evaluate transmission delay, and the final coordination scheme is generated. Monitor resource allocation status using SNMP or NetFlow protocols, analyze the effectiveness of strategies, and generate feedback optimization parameters.

[0004] Furthermore, based on the initial coordination scheme, entropy quantification of priority weights and dynamic peak calculation are performed to generate a data peak sequence. Parallel data volume sorting and matching with a preset impact factor table are then executed, and the PageRank algorithm is used for associative storage processing to generate a parallel impact dataset, specifically including: Identify and extract priority identifiers and their associated weight values, which are dynamically adjusted based on content type, business requirements, and historical transmission performance. Collect real-time data traffic information of the current communication content. The real-time data traffic information includes multi-dimensional indicators such as the data packet size, transmission frequency and timestamp of each communication content. Parallel data volume sorting operation is performed, specifically using a multi-dimensional sorting algorithm that combines content priority and real-time traffic as dual indicators to calculate the peak value of parallel communication content; The ranking weights are dynamically adjusted based on preset timeliness and importance rules. After sorting, a data peak sequence is generated. The data peak sequence is an ordered arrangement of content data peak information, which includes content identifier, priority weight and corresponding peak data volume.

[0005] Furthermore, based on the initial coordination scheme, entropy quantification of priority weights and dynamic peak calculation are performed to generate a data peak sequence. Parallel data volume sorting and matching with a preset impact factor table are then performed, and the PageRank algorithm is used for associative storage processing to generate a parallel impact dataset. This also includes: Capture the current communication session state, combine it with the similarity level information in the address classification results, and count the number of parallel communication addresses in each level; The number of parallel addresses is matched with a preset impact factor table, which contains the degree of bandwidth impact corresponding to different address number ranges, and is refined into multiple levels and corresponding impact factor values. Based on the statistically analyzed number of parallel addresses, determine the interval to which it belongs and select the corresponding influence factor value; Fine-tune the selected influence factor values ​​to generate address influence factor values.

[0006] Furthermore, based on the initial coordination scheme, entropy quantification and dynamic peak calculation of priority weights are performed to generate a data peak sequence. Parallel data volume sorting and matching with a preset impact factor table are then executed, and the PageRank algorithm is used for associative storage processing to generate a parallel impact dataset. This further includes: Perform multi-dimensional data association operations, specifically through key-value mapping and timestamp synchronization mechanisms, associate and store the peak values ​​of content data with the corresponding address influence factors to form a structured parallel influence dataset; The associated storage operation first establishes a correspondence between each content identifier and the address influence factor value in the data peak sequence; It employs efficient memory caching and indexing technologies to support rapid association and storage of large-scale data; The parallel impact dataset contains peak data and impact factor values, along with corresponding time windows and address similarity level indicators.

[0007] Furthermore, the parallel impact dataset is processed using min-max normalization or Z-score standardization, followed by product operations and dynamic scene correction based on a machine learning model to generate dynamic bandwidth requirement values, specifically including: The peak data sequence was selected from the parallel impact dataset. The peak data sequence covers the peak traffic data of different communication content in each time window. Normalization is performed by using a preset normalization algorithm to map peak data numerically, and converting different content types and their peak data into a standard numerical range. The normalization algorithm is based on the min-max normalization or Z-score normalization method, and dynamically adjusts the normalization parameters by combining historical peak statistics. Normalization processing, combined with real-time network load information, dynamically corrects the normalization results; After normalization, standardized peak data are generated.

[0008] Furthermore, the parallel impact dataset is processed using min-max normalization or Z-score standardization, followed by product operations and dynamic scene correction based on a machine learning model to generate dynamic bandwidth requirement values. This also includes: The input address influence factor values ​​and standardized peak data are validated for data format and range. The address impact factor value and the standardized peak data are multiplied item by item. For each time window and the corresponding content peak, the product of the peak value and the corresponding address impact factor value is calculated to obtain a preliminary estimate of the bandwidth requirement. A high-precision numerical calculation engine is used to ensure the accuracy of the calculation results, while detecting and automatically correcting calculation anomalies. The calculation results form a set of raw bandwidth demand values, covering bandwidth demand estimates for each content category and time window.

[0009] Furthermore, the parallel impact dataset is processed using min-max normalization or Z-score standardization, followed by product operations and dynamic scene correction based on a machine learning model to generate dynamic bandwidth requirement values, which further includes: Based on the current network operating status, time characteristics, and historical load trends, perform dynamic scenario correction; Collect information on network congestion, node performance metrics, and priority changes as the basis for correction. Dynamic scene correction is based on machine learning models or rule engines to adjust the original bandwidth demand value with weights dynamically calculated based on time period characteristics, network load fluctuations and historical trends. Set upper and lower limits for bandwidth requirements to prevent excessive adjustments that could lead to resource waste or shortage. The correction process takes into account load balancing between nodes and dynamically adjusts the distribution of bandwidth demand values.

[0010] Furthermore, the expression for generating the parallel impact dataset includes: Define an improved Shannon entropy quantification for its weight distribution: in, Prioritize content; Indexed by content type; Total number of content types; For the first The transmission frequency of the content; For the first The size of the data packet containing the class content; Priority weight; Calculate dynamic peak flow using a sliding window mechanism: in, For time window Dynamic peak values ​​of data within the range; A specific moment within a time window; This refers to the current time point; The length of the time window; For a moment No. The instantaneous transmission frequency of the content; For time window Data peak within; This is a function to find the maximum value. Dynamically adjust the values ​​of influencing factors using a linear programming model: in, For similarity ranking index; Similarity level The number of parallel addresses; This is the preset maximum address capacity; This is a network load correction factor; For level Influence factor values; It is an L2 norm; To minimize the objective function; These are constraints; Further time series smoothing processing: in, These are the smoothed impact factor values; It is a smoothing factor; This is the smoothed impact factor value from the previous period; An improved PageRank algorithm is used for associative storage processing to establish association weights: in, For level In time The association weight; For time window Data peak within; The influence factor value of the smoothed similarity level index; For similarity ranking index; The total number of similarity levels; This is the time decay factor; For level In time The association weight.

[0011] Furthermore, a big data network communication coordination device, applied to any of the methods described above, includes: The installation and configuration module is used to obtain network node installation parameters and complete address calibration and limitation. The data acquisition module is used to collect data from the communication port transmission queue and data stream characteristic parameters; The collision detection unit is used to perform collision detection within a preset bandwidth threshold and generate address classification results. The time synchronization module is used to timestamp and calibrate the characteristic parameters of the data stream to generate synchronization time slot labels. The dynamic alignment unit is used to dynamically align the address classification results with the synchronization time slot labels in bandwidth. The event arbitration module is used to determine transmission conflicts based on a preset arbitration strategy table. The control output unit is used to output coordinated correction parameters to the routing control unit; The data update module is used to record the protocol's effective timestamp and update the parallel impact dataset.

[0012] Furthermore, a computer-readable storage medium stores a computer-executable program for causing a computer to perform the big data network communication coordination method as described in any of the preceding claims.

[0013] The key innovations of this invention include: (1) Based on the integration of the initial coordination scheme and the coordination correction parameters, the strategy fusion processing is implemented to generate a set of candidate coordination schemes in order to improve the network resource scheduling efficiency in the scenario of multiple senders / receivers concurrent communication.

[0014] (2) By performing transmission delay simulation evaluation on the candidate coordination scheme set, a scheme priority sequence is generated to ensure more accurate scheme selection support in dynamic data environment.

[0015] (3) Using priority weight entropy quantization and dynamic peak calculation mechanism, the data peak sequence and address influence factor value are associated and stored to generate parallel influence dataset, so as to realize dynamic optimization of bandwidth utilization.

[0016] The following are its main beneficial effects: (1) The strategy of integrating the initial coordination scheme and the coordination correction parameters enables the candidate coordination scheme set to effectively improve the utilization rate and scheduling efficiency of network resources in the scenario of multiple senders / receivers communicating concurrently, and avoids resource conflicts caused by static priority allocation.

[0017] (2) The priority sequence of the scheme generated by the transmission delay simulation evaluation can provide accurate scheme selection in the dynamic data environment, ensure the execution efficiency and response speed of the scheme, and is suitable for the stable coordination requirements of high throughput data transmission.

[0018] (3) The entropy quantization and dynamic peak calculation mechanism of priority weights can dynamically optimize bandwidth utilization by generating parallel impact datasets, which solves the problem of bandwidth utilization fluctuation caused by single-dimensional parameter sorting in the existing technology. Attached Figure Description

[0019] Figure 1 A flowchart illustrating a big data network communication coordination method provided in an embodiment of this application; Figure 2 This is a structural block diagram of a big data network communication coordination device provided in an embodiment of this application. Detailed Implementation

[0020] Example 1: Refer to Figure 1 This is a flowchart illustrating a big data network communication coordination method provided in an embodiment of the present invention. The process may include at least steps S100-S600: S100. Obtain raw communication data, remove invalid characters, unify address format to dotted decimal format or lowercase domain name, remove duplicate data packets and filter abnormal content through preset data cleaning rules, and perform similarity classification using distance measurement algorithms such as edit distance, hash mapping or machine learning classification model. Based on the address classification results, match the preset benchmark scheme library to generate an initial coordination scheme. S200. Based on the initial coordination scheme, the entropy value of priority weights is quantified, dynamic peak calculation is performed, a data peak sequence is generated, parallel data volume sorting is performed, a preset impact factor table is matched, and the PageRank algorithm is used for associated storage processing to generate a parallel impact dataset. S300: The parallel impact dataset is processed by min-max normalization or Z-score standardization, and multiplication and dynamic scene correction based on machine learning model are performed to generate dynamic bandwidth requirement values. S400: Perform multi-level threshold judgment and threshold comparison, select the video encoding parameter adjustment scheme in the content format optimization scheme or the redundant data identification based on content fingerprint in the duplicate data compression strategy, perform parameterized encapsulation, and generate coordinated correction parameters. S500 integrates the initial coordination scheme and coordination correction parameters, uses a rule engine and policy weight mechanism for policy fusion processing, and uses discrete event simulation technology to simulate and evaluate transmission delay, generating the final coordination scheme. The S600 monitors resource allocation status via SNMP or NetFlow protocols, performs policy effect analysis, and generates feedback optimization parameters.

[0021] Step S100 includes at least steps S110-S130: S110. Obtain raw communication data, perform data cleaning processing, and obtain standardized address content data; Specifically, the system uses raw communication data as input, which includes multiple sender and receiver address identifiers and corresponding communication content information. First, using preset data cleaning rules, the address and content fields in the raw communication data undergo format validation and anomaly removal. This includes removing invalid characters, standardizing address formats (IP addresses to dotted decimal notation, domain names to lowercase), removing duplicate data packets, and filtering abnormal content. Further, for the communication content information, content summary generation and sensitive information masking are performed to ensure the standardization and security of data used in subsequent processing stages. During data cleaning, the system employs a multi-threaded pipeline architecture to ensure processing efficiency, while logging and marking abnormal data items for subsequent auditing and correction. After processing, the standardized address and content data is used as output fields for subsequent "address classification processing" in step S120, completing the standardization conversion of the input data and forming a unified format set of address and content data to support subsequent similarity grading and benchmark matching.

[0022] S120. Extract the communication address field from the standardized address content data, perform similarity classification processing, and generate address classification results; Specifically, taking the aforementioned standardized address content data as input, the communication address field is first decomposed into multi-level address units (such as network layer, subnet layer, and terminal identifier layer) through a preset address parsing module. Address feature vectors are then constructed based on the address hierarchy, geographical location, network topology, and other information. Further, a similarity algorithm using distance metrics (edit distance, hash mapping, or machine learning classification model) is applied to classify the address feature vectors based on address similarity grading, forming multiple address categories with similarity levels. During the classification process, the system sets thresholds to control the balance between classification accuracy and recall, performs multiple rounds of iterative optimization for address data with ambiguous boundaries, and dynamically adjusts the classification based on historical results. The classification result includes the similarity level identifier for each address and the corresponding address set. This address classification result is used as an output field and passed to the "baseline scheme matching" step S130 to ensure that the initial coordination scheme can be developed based on accurate address similarity grading, thereby achieving more refined network resource coordination.

[0023] S130. Match the address classification results with a preset benchmark scheme library to generate an initial coordination scheme; Specifically, the aforementioned address classification results are used as input, and a matching operation is performed through a preset benchmark scheme library. This preset benchmark scheme library includes multiple coordination strategies for different address similarity levels and network scenarios, covering bandwidth allocation, priority settings, and resource scheduling rules. The system first retrieves the corresponding preset benchmark scheme from the address classification results based on the similarity level, and then performs optimal matching based on current network status parameters (such as network load and node performance indicators). During the matching process, a rule engine and weighted scoring mechanism are used to comprehensively consider the communication needs and historical execution effects of the address groups, generating an initial coordination scheme that conforms to the current communication environment. This initial coordination scheme clarifies the resource allocation strategy and priority ranking for each address category, providing a foundation for subsequent dynamic bandwidth demand calculations. Finally, the initial coordination scheme is passed as an output field to step S210's "data peak calculation" to support a comprehensive evaluation of the impact of parallel content data and addresses, forming real-time data support for dynamic bandwidth evaluation.

[0024] Step S200 includes at least steps S210-S230: S210. Extract content priority parameters from the initial coordination scheme, sort parallel data volume, and generate data peak sequence; Specifically, taking the initial coordination scheme output in step S130 as input, the system first parses the content priority parameters in the initial coordination scheme. These content priority parameters reflect the priority weights and processing order of different communication contents during network transmission. Specifically, the system uses a preset parsing module to identify and extract the priority identifiers and associated weight values ​​corresponding to various types of communication contents. These weight values ​​are dynamically adjusted based on content type, business requirements, and historical transmission performance. Further, the system collects real-time data traffic information of the current communication content, including multi-dimensional indicators such as data packet size, transmission frequency, and timestamps for each communication content. Based on the content priority parameters and real-time data traffic information, the system performs a parallel data volume sorting operation, specifically employing a multi-dimensional sorting algorithm that combines content priority and real-time traffic indicators to calculate peak values ​​for parallel communication contents. During the sorting process, the system dynamically adjusts the sorting weights according to preset timeliness and importance rules to ensure that the bandwidth requirements of critical content are given priority in the sorting. To adapt to dynamic network environments, the sorting module supports a multi-threaded parallel processing architecture, enabling real-time responses to changes in communication content and ensuring the timeliness and accuracy of the sorting results. After sorting, the system generates a data peak sequence, which is an ordered arrangement of content data peak information, covering content identifiers, priority weights, and corresponding peak data volumes. This data peak sequence is passed as an output field to the "Address Quantity Statistics" call in step S220 for subsequent matching and comprehensive calculation of address influence factors. Furthermore, the generation process of the data peak sequence includes an anomaly detection and removal mechanism, marking and recording abnormal traffic peaks for subsequent auditing and correction.

[0025] S220. Count the number of parallel addresses in the address classification results, match the preset influence factor table, and generate address influence factor values; Specifically, taking the address classification results output in step S120 as input, the system first captures the current communication session status through a real-time monitoring module. Combining this with the similarity level information in the address classification results, it counts the number of parallel communication addresses within each level. The number of parallel communication addresses covers simultaneously active sender and receiver nodes. A dynamic counting mechanism is used during the counting process, and historical data trends are incorporated for smoothing to prevent sudden fluctuations from affecting the statistical results. The system further performs anomaly detection on the statistical data. For abnormal fluctuations or abnormal changes in the number of addresses, anomaly marking is performed and logged for subsequent analysis and processing. Further, the system matches the counted number of parallel addresses with a preset influence factor table. This table contains the bandwidth impact degree corresponding to different address quantity intervals, refined into multiple levels and corresponding influence factor values. During the matching process, the system determines the interval to which the counted parallel addresses belong and selects the corresponding influence factor value. Further, considering current network environment parameters such as network load and node performance indicators, the selected influence factor value is fine-tuned to improve the accuracy and adaptability of the matching. Finally, the system generates address influence factor values ​​as a quantitative representation of the impact of the number of parallel addresses on bandwidth requirements. The address impact factor value is passed as an output field to the generation of the "parallel impact dataset" in step S230 to assist in the accurate calculation of dynamic bandwidth requirements and resource coordination.

[0026] S230. Correlate and store the data peak sequence with the address influence factor value to generate a parallel influence dataset; Specifically, taking the data peak sequence output in step S210 and the address influence factor value output in step S220 as input, the system performs a multi-dimensional data association operation. Specifically, through key-value mapping and timestamp synchronization mechanisms, it associates and stores the content data peaks with their corresponding address influence factors, forming a structured parallel influence dataset. The association storage operation first establishes a correspondence between each content identifier and address influence factor value in the data peak sequence to ensure data consistency and temporal accuracy. The system employs efficient memory caching and indexing technology to support rapid association and storage of large-scale data. The parallel influence dataset not only contains peak data and influence factor values ​​but also includes corresponding time windows and address similarity level identifiers, facilitating refined processing for subsequent dynamic bandwidth calculations. The system further performs integrity verification on the parallel influence dataset to ensure that no associated data is omitted or duplicated. The parallel influence dataset, as an output field, is passed to the "demand bandwidth calculation" call in step S310, supporting multiplication and dynamic adjustment of bandwidth requirements. During the association storage process, the system logs and marks abnormal associations to ensure data quality and the reliability of subsequent processing.

[0027] In another embodiment, using the initial coordination scheme generated in step S130 as input, the system first extracts content priority parameters. These content priority parameters are derived from the transmission frequency in the real-time data traffic information. and data packet size The composition is quantized by the entropy value of the priority weights, specifically by quantifying the weight distribution using an improved Shannon entropy value: in: Content priority; Content type index; Total number of content types; : No. The transmission frequency of the content; : No. The size of the data packet containing the class content; Priority weights are used to characterize the urgency of content during network transmission. Data source → Metrics → Variable mapping: Extracting transmission frequency from real-time data traffic information and data packet size The priority weight is calculated using formula ①. This weight, combined with timestamp information, is used to calculate dynamic peak traffic through a sliding window mechanism: in: Time window Dynamic peak values ​​of data within the range; A specific moment within a time window; : Current time point; : Length of time window; :time No. The instantaneous transmission frequency of the content; Time window Data peak within; : The function to find the maximum value; The output field "Data Peak Sequence" is consumed by the "Address Quantity Statistics" step S220.

[0028] Furthermore, in S220, parallel data volume sorting and matching with a preset influence factor table are performed. Specifically, using the address classification results as input, the number of parallel addresses within each similarity level is counted. The values ​​of influencing factors are dynamically adjusted using a linear programming model. in: Similarity ranking index; Similarity level The number of parallel addresses is counted by the real-time monitoring module; : The preset maximum address capacity; Network load correction factor; :grade The influence factor values ​​are output after constraint optimization. L2 norm (Euclidean distance); Minimize the objective function; : Constraints (subject to); Data source → Metrics → Variable mapping: Extracted from address classification results Combined with the impact factor table and network load parameters The solution obtained by formula ③ is as follows This value is further processed using time series smoothing: in: : Smoothed impact factor values; Smoothing factor, default value is 0.7; The smoothing impact factor value of the previous period; The output field "Address Influence Factor Value" is consumed by the "Parallel Influence Dataset" in step S230.

[0029] Furthermore, in S230, the received data peak sequence and address influence factor value The improved PageRank algorithm is used for associative storage processing; specifically, association weights are established: in: :grade In time The association weight; Time window Data peak within; The impact factor value of the smoothed similarity level index; Similarity ranking index; Total number of similarity levels; Time decay factor, preset to 0.85; :grade In time The association weight; Data source → Metrics → Variable mapping: Extracted from data peak sequence , with address influence factor value The association weight is calculated using formula ⑤. Ultimately, a parallel impact dataset containing timestamps and classification labels is generated.

[0030] The technical effects of this section are as follows: By quantifying the entropy of priority weights and calculating dynamic peak values, the urgency of content transmission can be accurately assessed; by combining linear programming and time series optimization, a dynamic adjustment model for address influence factors can be constructed; and finally, a structured parallel influence dataset can be formed by establishing multi-dimensional data associations through an improved PageRank algorithm.

[0031] Step S300 includes at least steps S310-S330: S310. Extract the data peak sequence from the parallel impact dataset, perform normalization processing, and obtain standardized peak data; Specifically, the parallel impact dataset output in step S230 is used as input. This dataset contains content data peaks, corresponding address impact factor values, and their time window information. First, the system uses a data extraction module to filter out data peak sequences from the parallel impact dataset. These sequences cover peak traffic data for different communication content within each time window. For these peak sequences, the system further performs normalization processing. Specifically, a preset normalization algorithm is used to numerically map the peak data, uniformly converting different content types and their peak data to a standard numerical range. The normalization algorithm is based on min-max normalization or Z-score standardization methods, dynamically adjusting the normalization parameters in conjunction with historical peak statistical data to ensure that the distribution of standardized peak data is reasonable and representative. During processing, the system detects abnormal peak data. For outliers exceeding a reasonable range, truncation or smoothing is performed, and the anomalies are recorded in the log system for subsequent analysis. Furthermore, the normalization processing considers the time-varying characteristics of the network, combining real-time network load information to dynamically correct the normalization results, improving the timeliness and accuracy of the data. After normalization, the system generates standardized peak data. This standardized peak data, as an output field, is passed to the "product operation" call in step S320 for subsequent bandwidth demand calculation and dynamic adjustment. The standardized peak data contains peak information with uniform dimensions, is comparable across content types and time windows, and supports accurate calculation of dynamic bandwidth demands and resource coordination.

[0032] S320. Obtain the address influence factor value, multiply it with the standardized peak data, and generate the original demand bandwidth value. Specifically, the address influence factor value output in step S220 and the standardized peak data output in step S310 are used as inputs. The address influence factor value is a numerical value obtained based on the number of parallel addresses and network environment parameters, reflecting the degree of impact of address concurrency on bandwidth demand. The system first verifies the data format and range of the input address influence factor value and standardized peak data to ensure that they conform to the input specifications for product operations. Further, the product operation performs item-by-item product calculations on the address influence factor value and standardized peak data according to preset calculation rules. Specifically, for each time window and the corresponding content peak, its product with the corresponding address influence factor value is calculated to obtain a preliminary estimate of bandwidth demand. During the calculation process, the system uses a high-precision numerical calculation engine to ensure the accuracy of the calculation results, while detecting and automatically correcting calculation anomalies (such as overflow, non-numerical values). The calculation results form a set of original demand bandwidth values, which covers the bandwidth demand estimates under each content category and time window. The system further verifies the original demand bandwidth values, eliminates outliers and records abnormal events to ensure the reliability of subsequent processing. The original bandwidth demand value is passed as an output field to the "Dynamic Bandwidth Demand Value" generation in step S330, supporting subsequent scenario correction and optimization. This original bandwidth demand value is the core input for dynamic bandwidth demand calculation and directly affects the adjustment of network resource allocation strategies.

[0033] S330. Dynamically modify the original bandwidth demand value to generate a dynamic bandwidth demand value. Specifically, the original bandwidth demand value output in step S320 is used as input. The system combines the current network operating status, time characteristics, and historical load trends to perform dynamic scenario correction. First, the system collects network congestion status, node performance indicators, and priority change information through a real-time monitoring module as the basic parameters for correction. Further, the dynamic scenario correction is based on a machine learning model or rule engine to weight and adjust the original bandwidth demand value. The weights are dynamically calculated based on time period characteristics (such as peak and off-peak periods), network load fluctuations, and historical trends. Specifically, the system imposes upper and lower limits on the bandwidth demand value to prevent over-adjustment leading to resource waste or insufficiency. Further, the correction process considers load balancing among nodes, dynamically adjusting the distribution of the bandwidth demand value to better reflect the current network conditions. The system identifies and processes abnormal data during the dynamic correction process; abnormal situations are recorded and an alarm mechanism is triggered. After correction, a dynamic bandwidth demand value is generated. This dynamic bandwidth demand value is used as an output field and passed to the "threshold comparison processing" in step S410 for determining the bandwidth baseline threshold and triggering subsequent optimization strategies. The dynamic bandwidth requirement value reflects the scenario-based adaptive adjustment of bandwidth demand, supporting refined management and dynamic allocation of network resources.

[0034] Step S400 includes at least steps S410-S430: S410. Obtain the bandwidth baseline value, compare it with the dynamic bandwidth demand value using a threshold, and generate a trigger identifier; Specifically, the dynamic bandwidth demand value output in step S330 is used as input, along with the system's preset bandwidth baseline configuration parameters. The bandwidth baseline is set based on overall network capacity, historical load statistics, and Service Level Agreement (SLA) requirements. The system first verifies the format and range of the input dynamic bandwidth demand value and the bandwidth baseline to ensure both are within a valid value range. Further, a threshold comparison is performed between the bandwidth baseline and the dynamic bandwidth demand value. This threshold comparison follows predefined rules, performing item-by-item comparisons between the dynamic bandwidth demand value and the bandwidth baseline, specifically including single-point peak comparisons and average value comparisons within a time window. During the comparison process, the system considers network latency, fluctuation amplitude, and burst traffic factors, employing a sliding window mechanism to smooth the dynamic bandwidth demand value and reduce the probability of false triggering. Furthermore, the threshold comparison includes multi-level threshold determination, distinguishing between mild, moderate, and severe exceedances, each corresponding to different trigger intensities. The system generates a trigger identifier based on the comparison results. This trigger identifier includes information such as trigger status (triggered or not triggered), trigger level, and trigger timestamp. The generation process of the trigger identifier is accompanied by anomaly detection. For abnormal data or calculation errors, the system performs fault-tolerant processing and records the abnormal events in detail in the log system for subsequent auditing and optimization. The trigger identifier is passed as an output field to the "Optimization Strategy Selection" call in step S420, supporting the dynamic selection and execution of content format optimization or duplicate data compression strategies for bandwidth exceeding limits.

[0035] S420. Select the content format optimization scheme and duplicate data compression strategy according to the trigger identifier, and generate the optimization strategy instruction; Specifically, taking the trigger identifier output in step S410 as input, the system parses the trigger status and level information in the trigger identifier, and, combined with the current network environment configuration and policy library, performs the selection and generation of optimization strategies. The policy library pre-sets multiple content format optimization schemes and duplicate data compression strategies, covering multi-dimensional technologies such as encoding format adjustment, data packet reassembly, redundant data removal, and non-critical data diversion. The system first determines the urgency of the optimization requirement based on the trigger level. For minor triggers, a content format optimization scheme is prioritized, including adjusting video encoding parameters, reducing image resolution, or using a more efficient transmission protocol. For moderate and severe triggers, a duplicate data compression strategy is prioritized, including redundant data identification and compression based on content fingerprints, duplicate data stream separation, and dynamic switching of compression algorithms. Furthermore, the system dynamically matches the most suitable optimization strategy based on the current communication content characteristics and priority information to avoid negatively impacting critical content. The optimization strategy selection process employs a hybrid decision-making mechanism combining a rule engine and a machine learning model to improve the accuracy and adaptability of strategy matching. The system monitors for anomalies during the strategy selection process. In the event of abnormal strategy matching or conflict, it executes strategy rollback or manual intervention alerts. The optimization strategy instruction includes a specific optimization method identifier, parameter configuration, and execution priority, which are passed as output fields to the "coordination correction parameter" generation step S430 to support the parameterized encapsulation and execution adjustment of subsequent coordination schemes.

[0036] S430. Parameterize the optimization strategy instructions to generate coordinated correction parameters; Specifically, the optimization strategy instruction output in step S420 is used as input, and the system performs parameterized encapsulation processing to convert the optimization strategy instruction into a unified set of coordination and correction parameters. The parameterized encapsulation first parses the parameters in the optimization strategy instruction, including optimization type, compression ratio, format conversion rules, and execution sequence, and adjusts the parameter values ​​to adapt to the actual environment based on the current network resource status and node capabilities. During the encapsulation process, the system adopts a modular design, encapsulating the parameters of different optimization strategies in the form of structured data objects, supporting flexible expansion and combination applications. Furthermore, the system performs integrity and consistency verification on the coordination and correction parameters to ensure that there are no conflicts between parameters and that they comply with network device execution specifications. After encapsulation, the coordination and correction parameters are assigned a unique identifier and version number for easy subsequent tracking and management. The system also records abnormal information and adjustment records during the encapsulation process, forming an operation log. The coordination and correction parameters are passed as output fields to the "Adaptive Scheme Generation" call in step S510 to integrate the initial coordination scheme and correction parameters, achieving adaptive network resource coordination and adjustment.

[0037] Step 500 includes at least steps S510-S530: S510. Integrate the initial coordination scheme and coordination correction parameters, perform strategy fusion processing, and generate a set of candidate coordination schemes; The strategy fusion employs a rule engine and strategy weighting mechanism, combining the resource allocation strategy of the initial coordination scheme with optimization measures in the coordination correction parameters to achieve organic integration of multiple strategies. Specifically, the system parses the optimization type, compression ratio, and execution priority in the coordination correction parameters, and adjusts the resource allocation ratio and scheduling order based on the resource requirements and priority ranking of each address category in the initial scheme. During strategy fusion, the system detects and reconciles conflicting strategies, employing priority determination and rollback mechanisms to ensure the rationality and feasibility of the fusion result. The strategy fusion supports dynamic adjustment, updating the fusion weights in real time according to network status to improve scheme adaptability. The system captures and logs abnormal operations during the strategy fusion process for subsequent optimization and troubleshooting. Finally, the generated candidate coordination scheme set is passed as an output field to the "delay evaluation" call in step S520 for subsequent scheme priority ranking.

[0038] S520. Perform transmission delay simulation evaluation on the candidate coordination scheme set and generate a scheme priority sequence; Specifically, the transmission delay simulation evaluation first collects the current network topology, link status, and node performance indicators, and then constructs a network transmission model by combining the resource allocation strategies in the candidate schemes. During the transmission delay simulation evaluation, the system employs discrete event simulation technology to simulate the data transmission process of each scheme under different network conditions, calculating key performance indicators such as transmission delay, packet loss rate, and bandwidth utilization. Furthermore, in the transmission delay simulation evaluation, the system considers the concurrent communication characteristics of multiple senders and receivers, dynamically adjusting transmission paths and resource scheduling strategies to reflect the complexity of the real network environment. The transmission delay simulation evaluation supports multi-threaded parallel computing to improve simulation efficiency. For the simulation results, the system performs a multi-dimensional comprehensive scoring, combining delay, bandwidth utilization, and stability indicators to generate a scheme performance score. Further, based on the scoring results, the system sorts the candidate coordination scheme set according to preset priority rules, forming a scheme priority sequence. The sorting process uses a stable sorting algorithm to ensure the consistency of the order of schemes with the same priority. The system detects and processes abnormal data during the evaluation and sorting process; abnormal events are recorded and alarms are triggered. The priority sequence of the schemes is passed as an output field to the "final coordination scheme" call in step S530, which supports the selection and generation of the optimal scheme.

[0039] S530. Select the optimal solution based on the priority sequence of the solutions and generate the final coordination solution; Taking the priority sequence of schemes output in step S520 as input, the system selects the optimal scheme based on the priority sequence and generates the final coordination scheme. Specifically, the system first parses the sorting results in the priority sequence to select the optimal scheme, which includes selecting the candidate coordination scheme with the highest priority as the preliminary optimal scheme. Further, during the selection process, the system performs an integrity check on the preliminary optimal scheme to verify the rationality of the resource allocation strategy and the validity and compatibility of the execution parameters. The integrity check includes verifying the bandwidth allocation ratio, priority order, and optimization parameters to ensure that the scheme meets the current network environment and service requirements. For any non-compliance found during the check, the system automatically adjusts or calls alternative schemes to ensure the executability of the final scheme. Further, the system dynamically optimizes the selected scheme by combining the current network status and historical execution data, adjusting some parameters to improve the actual performance of the scheme. The dynamic optimization process uses a machine learning model to predict network load trends, adjust resource allocation strategies, and improve the adaptability and efficiency of the scheme. After optimization, the system assigns a unique identifier and version number to the final coordination scheme for easy tracking and management. The system records and processes any abnormal events during the scheme selection and optimization process, generating an operation log. The final coordination scheme is passed as an output field to the "Scheme Execution Instruction" call in step S610, supporting the execution of the coordination scheme by network devices.

[0040] In another embodiment, in S510, an initial coordination scheme and coordination correction parameters are received, and a weighted fusion strategy is used to generate candidate schemes: in: : No. Item weight candidate coordination scheme set; , Strategy index; Initial Scheme Item strategy weights; : Correct the first parameter in Adjustment amount; : Strategy retention coefficient; : Adjust the number of parameters; : Adjust parameter weights; The output field "Candidate Coordination Scheme Set" is consumed by the "Delay Evaluation" step S520.

[0041] In S520, perform the following operations: (1) Scheme analysis and parameter mapping: The candidate scheme set Mapped to network configuration parameters (link bandwidth allocation rate) required for discrete event simulation Queue scheduling weight Node processing rate ), and inject the standard test data stream of S210; (2) Discrete event simulation: Based on the current network topology (from the installation and configuration module) and real-time background traffic (from the data acquisition module), discrete event simulation is used to extract transmission delay. Resource utilization rate Packet loss rate ; (3) Construct a multi-objective scoring function, and the expression for calculating the scheme score is: in, For the plan The rating; For the plan Transmission delay; To prevent small positive numbers from being divided by zero; For the plan Resource utilization rate; Packet loss rate; To prevent small positive numbers from being divided by zero; , These are the weighting coefficients; (4) Generate a priority sequence of schemes: according to Output the data in descending order to S530.

[0042] Step S530 selects the optimal solution by prioritizing and generates the final coordination solution for execution in step S610.

[0043] Technical benefits of this section: Through strategy fusion and dynamic scoring mechanisms, diverse candidate solutions are generated and refined for evaluation; by combining transmission delay simulation and priority ranking, the final coordination solution achieves a balance between latency and resource utilization.

[0044] Step S600 includes at least steps S610-S630: S610: Convert the final coordination scheme into executable instructions for network devices and send them to the communication nodes; The final coordination scheme is taken as input. Specifically, the final coordination scheme is generated in step S530 and includes resource allocation strategies and priority ordering information for concurrent communication environments with multiple senders and multiple receivers. The system first performs format parsing and instruction mapping on the final coordination scheme. The parsing process includes converting the resource allocation parameters, priority weights, and optimization strategies in the scheme into network device executable instructions. Specifically, the network device executable instructions include the parsing module refining the high-level coordination strategy into specific configuration commands and scheduling instructions based on preset instruction templates and protocol specifications, combined with the hardware capabilities and interface standards of the network device. This process involves converting address identifiers, bandwidth allocation ratios, compression parameters, and transmission priorities in the scheme item by item to ensure that the generated scheme execution instructions are fully compatible with the control interface of the network device.

[0045] Furthermore, the system packages the execution instruction set through an instruction encapsulation module, employing a unified message format and transmission protocol to support reliable instruction delivery and confirmation. During encapsulation, the system dynamically adjusts the instruction transmission path and priority based on the network topology and the distribution of communication nodes, using fragmentation and reassembly mechanisms to process large-scale instruction sets, ensuring efficient and stable delivery. The instruction delivery module sends the encapsulated execution instructions to each communication node through a secure channel, supporting multi-threaded concurrent transmission and retransmission mechanisms to ensure instruction integrity and real-time performance. The system monitors the instruction transmission status in real time, recording transmission logs and abnormal events to facilitate subsequent troubleshooting and performance analysis.

[0046] The execution instruction of the scheme is passed as an output field to the "Resource Allocation Monitoring" call in step S620 for subsequent network resource allocation status monitoring and feedback optimization.

[0047] S620 monitors the network resource allocation status of the execution instructions of the monitoring scheme and generates real-time performance data; The scheme execution instructions are taken as input. Specifically, the scheme execution instructions are generated in step S610 and include commands for bandwidth allocation, priority scheduling, and optimization strategies for communication nodes. The system first accesses the network device's status information acquisition interface through the monitoring module to obtain the resource allocation status of communication nodes in real time. The resource allocation includes key performance indicators such as bandwidth utilization, transmission latency, packet loss rate, and node load. Data acquisition adopts a multi-protocol compatible mechanism, supporting SNMP (Simple Network Management Protocol), NetFlow, and customized monitoring interfaces, capable of covering different types of network devices and communication nodes.

[0048] Furthermore, the system preprocesses the collected resource allocation status data, specifically including data format conversion, time synchronization, and anomaly filtering. The data format conversion module maps data collected from different devices to an internal standard format. The time synchronization module calibrates the timestamps of each data source using the Network Time Protocol (NTP). The anomaly filtering module identifies and removes lost data, abnormal peaks, and noise during the collection process, ensuring the accuracy and integrity of the monitoring data. The processed real-time performance data is stored in a high-performance caching system, supporting fast querying and historical data backtracking.

[0049] The system further analyzes real-time performance data from multiple dimensions through a performance analysis module, employing statistical analysis, trend prediction, and anomaly detection algorithms to identify bottlenecks and anomalies in resource allocation. Analysis results include bandwidth utilization fluctuation curves, node load distribution diagrams, and a list of abnormal events. The monitoring module sets threshold alarm mechanisms for key performance indicators, generating alarm events and recording detailed logs when triggered, supporting timely responses from operations and maintenance personnel. The real-time performance data, as an output field, is passed to the "Feedback Optimization Parameters" call in step S630, supporting subsequent strategy effectiveness analysis and feedback adjustments.

[0050] S630: Analyze the strategy effect on real-time performance data and generate feedback optimization parameters; The real-time performance data is analyzed for policy effectiveness to generate feedback optimization parameters. Specifically, the real-time performance data output in step S620 is used as input. The policy effectiveness analysis includes data cleaning and feature extraction of the performance data, removing outliers and noise, and extracting key feature indicators such as bandwidth utilization, latency variation, packet loss rate, and node load. Further, the feedback analysis module combines historical execution data with preset policy objectives, employing machine learning models and a rule engine to comprehensively evaluate the performance indicators, identifying the advantages, disadvantages, and potential improvement areas of the current coordination scheme in actual execution.

[0051] Furthermore, the system generates feedback optimization parameters based on the analysis results, specifically including adjusting bandwidth allocation ratios, optimizing priority ranking, modifying compression strategy parameters, and adjusting scheduling timing. The feedback optimization parameters adopt a structured data format, including parameter name, adjustment range, applicable scope, and effective conditions, supporting fine-grained strategy correction. During the generation process, the system considers the time-varying characteristics of the network and node heterogeneity, dynamically adjusting the weights and priorities of the feedback parameters to avoid over-adjustment leading to system oscillations. The feedback optimization parameters also include version information and a timestamp for easy tracking and backtracking.

[0052] The system transmits the feedback optimization parameters back to the data cleaning and processing module in step S110 via a secure channel, enabling the updating and adjustment of the baseline parameters and forming a closed-loop feedback mechanism. The feedback optimization parameters, as output fields, are passed to the "data cleaning and processing" call in step S110, supporting the dynamic updating and adaptive optimization of the baseline scheme.

[0053] Example 2: Figure 2 A structural block diagram of a big data network communication coordination device according to an embodiment of the present invention is shown. Figure 2 As shown, the structure may include: The installation and configuration module 01 is used to obtain network node installation parameters and complete address calibration. Specifically, it receives network node configuration files from the system initialization phase, parses the installation parameters, including the node's physical location, network topology information, and device identifier. Through the interface configuration tool, it performs address calibration of the nodes, ensuring the uniqueness and accuracy of address identification. During configuration, the system calls a preset calibration algorithm, combining geographical location and topology, to precisely limit the logical address range of the node in the network. After calibration, a node address configuration file is generated, recorded in the configuration database, and transmitted to the data acquisition module as the basic input for address identification information. Simultaneously, configuration logs are retained for subsequent system auditing and configuration updates.

[0054] Data acquisition module 02 is used to collect transmission queue data and data stream characteristic parameters of communication ports. Specifically, it receives the node address configuration file output from the installation and configuration module, and monitors and collects the transmission queue data and data stream characteristic parameters of each communication port in real time through the communication interface with the network device. During the acquisition process, a multi-threaded parallel processing architecture is adopted to ensure the real-time performance and integrity of the data. The system calls a feature extraction algorithm to analyze the collected data and identify characteristic parameters such as the protocol type, packet size, and transmission frequency of the data stream. The generated set of data stream characteristic parameters is recorded as real-time monitoring data and transmitted to the collision detection unit as input for subsequent collision detection and analysis.

[0055] The conflict detection unit 03 is used to perform conflict detection and generate address classification results within a preset bandwidth threshold. Specifically, it receives real-time monitoring data from the data acquisition module, combines it with the system's preset bandwidth threshold, executes a conflict detection algorithm, and identifies potential conflicts in the transmission queue. During the detection process, the system judges the deviation between the bandwidth occupancy and the threshold based on data flow characteristic parameters and generates conflict alarm information. Further, it calls the address resolution module to perform address classification processing on the conflict data, generating address classification results based on address similarity and network topology. This result includes a conflict level identifier for each address and the corresponding address set, which is passed as an output field to the time synchronization module, and the detection details are recorded in the conflict log.

[0056] The time synchronization module 04 is used to calibrate the data stream characteristic parameters with timestamps and generate synchronization time slot labels. Specifically, it receives the address classification results from the collision detection unit, combines them with the timestamp information in the data stream characteristic parameters, and executes a time synchronization algorithm. The system calibrates the timestamps of each data source using the Network Time Protocol (NTP) to ensure the timing consistency of the data streams. During synchronization, the system generates synchronization time slot labels to identify the transmission time slot of each data stream under a unified time base. This synchronization time slot label is output and passed to the dynamic alignment unit for subsequent bandwidth alignment. Simultaneously, the time synchronization log records abnormal events during the synchronization process for subsequent time calibration optimization.

[0057] Dynamic alignment unit 05 is used to dynamically align the address classification results with the synchronization time slot labels. Specifically, it receives the synchronization time slot labels and address classification results from the time synchronization module, calls the bandwidth alignment algorithm, and dynamically adjusts the bandwidth allocation for each address category. During the alignment process, the system combines the synchronization time slot information to identify peak and idle periods of bandwidth usage and dynamically adjusts the allocation priority of bandwidth resources. The generated bandwidth alignment configuration file records the bandwidth allocation strategy for each address category and is passed to the event arbitration module as input to support subsequent arbitration decisions for transmission conflicts. Simultaneously, an alignment log is maintained to record adjustment details and anomalies during the alignment process.

[0058] Event arbitration module 06 is used to determine transmission conflicts based on a preset arbitration policy table. Specifically, it receives a bandwidth alignment configuration file from the dynamic alignment unit, combines it with the system's preset arbitration policy table, and executes a transmission conflict determination algorithm. During the arbitration process, the system determines the transmission priority and resource allocation rationality of each address class based on the bandwidth allocation policy and conflict level information. For identified conflict events, the system calls the rules in the arbitration policy table to generate a transmission conflict arbitration result, including conflict resolution and adjustment suggestions. This arbitration result is transmitted as output to the control output unit, and the determination process and result details are recorded in the arbitration log.

[0059] Control output unit 07 is used to output coordinated correction parameters to the routing control unit. Specifically, it receives the transmission conflict arbitration results from the event arbitration module, parses the adjustment suggestions and solutions, and generates coordinated correction parameters. The system converts the correction parameters into routing control commands through the interface control module, supporting automated adjustment of network devices. The generated coordinated correction parameters are output to the routing control unit as input for routing configuration, ensuring dynamic adjustment and optimization of network resources. Simultaneously, the system records correction logs to track the generation and application of correction parameters, supporting subsequent optimization analysis and strategy adjustments.

[0060] Data update module 08 is used to record the protocol's effective timestamp and update the parallel impact dataset. Specifically, it receives the effective status of the coordinated correction parameters from the control output unit, combines it with the system's protocol execution log, and records the protocol's effective timestamp information. The system calls the update algorithm to update the parallel impact dataset, reflecting the latest network resource allocation and bandwidth usage status. During the update process, the system ensures the integrity and consistency of the dataset, generating an updated parallel impact dataset for subsequent dynamic bandwidth demand calculations. Simultaneously, the update log records any anomalies and data change details during the update process, supporting subsequent data analysis and strategy optimization.

[0061] Furthermore, embodiments of this application also provide a computer-readable storage medium storing a computer-executable program for causing a computer to perform the big data network communication coordination method as described in the above embodiments.

[0062] Those skilled in the art will understand that embodiments of this application may also include computer program products. Therefore, this application may take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this application may take the form of a computer program product embodied on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0063] This application is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), devices, and computer program products according to embodiments of this application. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart... Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0064] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0065] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0066] The above description is only a preferred embodiment of the present invention. It should be noted that for those skilled in the art, several improvements and modifications can be made without departing from the technical principles of the present invention, and these improvements and modifications should also be considered within the scope of protection of the present invention.

Claims

1. A big data network communication coordination method, characterized in that, include: The system acquires raw communication data and performs data cleaning by removing invalid characters, unifying the address format to dotted decimal or lowercase domain names, eliminating duplicate data packets, and filtering abnormal content. It then uses distance measurement algorithms such as edit distance, hash mapping, or machine learning classification models to perform similarity classification and generates an initial coordination scheme by matching the address classification results with a preset benchmark scheme library. Based on the initial coordination scheme, priority weight entropy quantification and dynamic peak calculation are performed to generate a data peak sequence. Parallel data volume sorting and matching with a preset impact factor table are executed, and the PageRank algorithm is used for associated storage processing to generate a parallel impact dataset. The parallel impact dataset is processed by min-max normalization or Z-score standardization, and multiplication and dynamic scene correction based on machine learning model are performed to generate dynamic bandwidth requirement values. Perform multi-level threshold judgment and threshold comparison, select the video encoding parameter adjustment scheme in the content format optimization scheme or the redundant data identification based on content fingerprint in the duplicate data compression strategy, perform parameterized encapsulation, and generate coordinated correction parameters; The initial coordination scheme and coordination correction parameters are integrated, and a rule engine and policy weight mechanism are used for policy fusion processing. Discrete event simulation technology is used to simulate and evaluate transmission delay, and the final coordination scheme is generated. Monitor resource allocation status using SNMP or NetFlow protocols, analyze the effectiveness of strategies, and generate feedback optimization parameters.

2. The big data network communication coordination method according to claim 1, characterized in that, Based on the initial coordination scheme, entropy quantification and dynamic peak calculation of priority weights are performed to generate a data peak sequence. Parallel data volume sorting and matching with a pre-defined impact factor table are then executed, and the PageRank algorithm is used for associative storage processing to generate a parallel impact dataset. Specifically, this includes: Identify and extract priority identifiers and their associated weight values, which are dynamically adjusted based on content type, business requirements, and historical transmission performance. Collect real-time data traffic information of the current communication content. The real-time data traffic information includes multi-dimensional indicators such as the data packet size, transmission frequency and timestamp of each communication content. Parallel data volume sorting operation is performed, specifically using a multi-dimensional sorting algorithm that combines content priority and real-time traffic as dual indicators to calculate the peak value of parallel communication content; The ranking weights are dynamically adjusted based on preset timeliness and importance rules. After sorting, a data peak sequence is generated. The data peak sequence is an ordered arrangement of content data peak information, which includes content identifier, priority weight and corresponding peak data volume.

3. The big data network communication coordination method according to claim 1, characterized in that, Based on the initial coordination scheme, priority weights are quantified by entropy, dynamic peak values ​​are calculated, a data peak sequence is generated, parallel data volume sorting and matching with a preset impact factor table are performed, and the PageRank algorithm is used for associative storage processing to generate a parallel impact dataset. This also includes: Capture the current communication session state, combine it with the similarity level information in the address classification results, and count the number of parallel communication addresses in each level; The number of parallel addresses is matched with a preset impact factor table, which contains the degree of bandwidth impact corresponding to different address number ranges, and is refined into multiple levels and corresponding impact factor values. Based on the statistically analyzed number of parallel addresses, determine the interval to which it belongs and select the corresponding influence factor value; Fine-tune the selected influence factor values ​​to generate address influence factor values.

4. The big data network communication coordination method according to claim 1, characterized in that, Based on the initial coordination scheme, priority weights are quantified by entropy and dynamic peak values ​​are calculated to generate a data peak sequence. Parallel data volume sorting and matching with a pre-defined impact factor table are then performed, and the PageRank algorithm is used for associative storage processing to generate a parallel impact dataset. Further steps include: Perform multi-dimensional data association operations, specifically through key-value mapping and timestamp synchronization mechanisms, associate and store the peak values ​​of content data with the corresponding address influence factors to form a structured parallel influence dataset; The associated storage operation first establishes a correspondence between each content identifier and the address influence factor value in the data peak sequence; It employs efficient memory caching and indexing technologies to support rapid association and storage of large-scale data; The parallel impact dataset contains peak data and impact factor values, along with corresponding time windows and address similarity level indicators.

5. The big data network communication coordination method according to claim 1, characterized in that, The parallel impact dataset is processed using min-max normalization or Z-score standardization, followed by product operations and dynamic scene correction based on a machine learning model to generate dynamic bandwidth requirement values, specifically including: The peak data sequence was selected from the parallel impact dataset. The peak data sequence covers the peak traffic data of different communication content in each time window. Normalization is performed by using a preset normalization algorithm to map peak data numerically, and converting different content types and their peak data into a standard numerical range. The normalization algorithm is based on the min-max normalization or Z-score normalization method, and dynamically adjusts the normalization parameters by combining historical peak statistics. Normalization processing, combined with real-time network load information, dynamically corrects the normalization results; After normalization, standardized peak data are generated.

6. The big data network communication coordination method according to claim 1, characterized in that, The parallel impact dataset is processed using min-max normalization or Z-score standardization, followed by product operations and dynamic scene correction based on a machine learning model to generate dynamic bandwidth requirement values. This also includes: The input address influence factor values ​​and standardized peak data are validated for data format and range. The address impact factor value and the standardized peak data are multiplied item by item. For each time window and the corresponding content peak, the product of the peak value and the corresponding address impact factor value is calculated to obtain a preliminary estimate of the bandwidth requirement. A high-precision numerical calculation engine is used to ensure the accuracy of the calculation results, while detecting and automatically correcting calculation anomalies. The calculation results form a set of raw bandwidth demand values, covering bandwidth demand estimates for each content category and time window.

7. The big data network communication coordination method according to claim 1, characterized in that, The parallel impact dataset is processed using min-max normalization or Z-score standardization, followed by product operations and dynamic scene correction based on a machine learning model to generate dynamic bandwidth requirement values. Further steps include: Based on the current network operating status, time characteristics, and historical load trends, perform dynamic scenario correction; Collect information on network congestion, node performance metrics, and priority changes as the basis for correction. Dynamic scene correction is based on machine learning models or rule engines to adjust the original bandwidth demand value with weights dynamically calculated based on time period characteristics, network load fluctuations and historical trends. Set upper and lower limits for bandwidth requirements to prevent excessive adjustments that could lead to resource waste or shortage. The correction process takes into account load balancing between nodes and dynamically adjusts the distribution of bandwidth demand values.

8. The big data network communication coordination method according to claim 1, characterized in that, The expressions for generating parallel impact datasets include: Define an improved Shannon entropy quantification for its weight distribution: in, Prioritize content; Indexed by content type; Total number of content types; For the first The transmission frequency of the content; For the first The size of the data packet containing the class content; Priority weight; Calculate dynamic peak flow using a sliding window mechanism: in, For time window Dynamic peak values ​​of data within the range; A specific moment within a time window; This refers to the current time point; The length of the time window; For a moment No. The instantaneous transmission frequency of the content; For time window Data peak within; This is a function to find the maximum value. Dynamically adjust the values ​​of influencing factors using a linear programming model: in, For similarity ranking index; Similarity level The number of parallel addresses; This is the preset maximum address capacity; This is a network load correction factor; For level Influence factor values; It is an L2 norm; To minimize the objective function; These are constraints; Further time series smoothing processing: in, These are the smoothed impact factor values; It is a smoothing factor; This is the smoothed impact factor value from the previous period; An improved PageRank algorithm is used for associative storage processing to establish association weights: in, For level In time The association weight; For time window Data peak within; The influence factor value of the smoothed similarity level index; For similarity ranking index; The total number of similarity levels; This is the time decay factor; For level In time The association weight.

9. A big data network communication coordination device, applied to the method of any one of claims 1-8, characterized in that, include: The installation and configuration module is used to obtain network node installation parameters and complete address calibration and limitation. The data acquisition module is used to collect data from the communication port transmission queue and data stream characteristic parameters; The collision detection unit is used to perform collision detection within a preset bandwidth threshold and generate address classification results. The time synchronization module is used to timestamp and calibrate the characteristic parameters of the data stream to generate synchronization time slot labels. The dynamic alignment unit is used to dynamically align the address classification results with the synchronization time slot labels in bandwidth. The event arbitration module is used to determine transmission conflicts based on a preset arbitration strategy table. The control output unit is used to output coordinated correction parameters to the routing control unit; The data update module is used to record the protocol's effective timestamp and update the parallel impact dataset.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer-executable program for causing a computer to perform the big data network communication coordination method as described in any one of claims 1-8.

Images