Search

CN-121984899-A - Intelligent aviation internet flow monitoring system based on multidimensional correlation analysis

CN121984899ACN 121984899 ACN121984899 ACN 121984899ACN-121984899-A

Abstract

The invention relates to an aviation Internet flow intelligent monitoring system based on multidimensional correlation analysis, and belongs to the technical field of network communication and intelligent monitoring. The system comprises a dynamic weight association unit for constructing a scene-equipment-service three-layer frame and dynamically generating a weight association table in combination with a flight phase, an abnormal flow cross-dimension tracing unit for integrating multi-dimension data to locate an abnormal root cause, a bandwidth prediction multi-factor optimizing unit for predicting requirements and optimizing bandwidth allocation through a time sequence association model, and an air-ground collaborative mining unit for constructing an altitude application suitability scoring system to generate an optimizing scheme. The method and the device realize differential control of the aviation scene flow, guarantee the priority of core business, improve the exception handling efficiency and the bandwidth utilization rate, improve the ground application high-altitude adaptation defect, and comprehensively enhance the operation stability, the accuracy and the user experience of the aviation Internet.

Inventors

  • LI HUA
  • XIONG WEI
  • CAO KAIYUAN

Assignees

  • 空地互联网络科技股份有限公司

Dates

Publication Date
20260505
Application Date
20260203

Claims (12)

  1. 1. The intelligent aviation internet flow monitoring system based on multidimensional correlation analysis is characterized by comprising a dynamic weight correlation unit, an abnormal flow cross-dimension tracing unit, a bandwidth prediction multi-factor optimization unit and an air-ground collaborative mining unit; The dynamic weight association unit acquires dynamic weight information and constructs a multi-dimensional association frame, presets a dynamic weight calculation rule based on the multi-dimensional association frame, and outputs a dynamic weight association table; The system comprises a dynamic weight association table, an abnormal flow cross-dimension tracing unit, an abnormal flow cross-dimension tracing rule system, an automatic calling device, a comparison device, a satellite link packet loss rate and time delay data matching corresponding flight phase characteristics, a multi-dimensional data comparison and verification device, a data analysis device and a data analysis device, wherein the abnormal flow cross-dimension tracing unit constructs an association tracing rule system according to the dynamic weight association table; The bandwidth prediction multi-factor optimizing unit builds a bandwidth demand prediction model through a time sequence association algorithm based on the multi-dimensional association frame and the abnormal flow root cause analysis report, integrates the association influence weights of various factors in the training process, prejudges the bandwidth demands of different airlines and different time periods, identifies the application types and key time periods of the rapid increase of the flow, and outputs a bandwidth allocation result; the space-ground collaborative mining unit combines the bandwidth allocation result, adds a high-altitude network matching attribute tag in a preset application management library, builds an application high-altitude suitability scoring system based on the application management library, and finally generates an application replacement suggestion and a protocol optimization scheme.
  2. 2. The system according to claim 1, wherein the specific process of constructing the multi-dimensional association frame is: acquiring dynamic weight information, including a flight phase state, an application emergency degree identifier and core attributes of equipment priority classification; building three layers of progressive frames according to a scene dimension, a device dimension and a service dimension, wherein the scene dimension and the device dimension are mapped through an adaptation rule, and the device dimension and the service dimension are bound according to functional requirements; And embedding the classification item index and the association interface in each layer, and automatically checking the matching degree and updating the association relation when the type is newly added.
  3. 3. The system according to claim 1, wherein the specific process of the preset dynamic weight calculation rule is: determining basic weight according to the satellite bandwidth bearing upper limit and the degree of the need of various services; Setting the adjustment coefficients of the medical emergency equipment and the related business of the flight assistance as a first gear, setting the adjustment coefficients of the office terminal and the corresponding business of the flat flight cruising stage as a second gear, setting the adjustment coefficients of the cabin entertainment terminal and the corresponding business as a third gear, wherein the first gear coefficient is larger than the second gear, and the second gear is larger than the third gear; Defining a weight adjustment triggering condition as a flight phase switching instruction pushed by a fleet system, wherein the instruction comprises a phase type and a switching time stamp; And after receiving the instruction, automatically updating the weight values of the equipment and the service dimension within a preset response time, and synchronously recording the weight data before and after adjustment, the details of the triggering instruction and the updating time.
  4. 4. The system of claim 1, wherein the specific process of obtaining the dynamic weight association table comprises mapping a flight phase state, equipment priority classification and application emergency degree identification acquired in real time with preset weights, and sorting the dynamic weight association table comprising percentage weight values under each combined scene according to the flight phase as an index and comprising specific time nodes, equipment types and application types as sub-items and labeling corresponding models and names.
  5. 5. The system of claim 1, wherein the specific process of constructing the association traceability rule system is to take the priority in the dynamic weight association table as a core, establish an association rule, set an upper limit of application quantity and a core application binding relation which can be associated with a single device according to a corresponding rule of a device running state and an application use condition, preset a corresponding relation between a critical threshold of packet loss rate and time delay and a flow abnormality degree according to an association rule of satellite link parameters and flow abnormality, and preset a normal fluctuation interval of service flow in each stage according to an adaptation rule of a flight stage and service flow fluctuation.
  6. 6. The system of claim 1, wherein the specific process of the preset threshold value is to calculate a statistical value according to historical flow data of similar equipment in a corresponding flight stage and service type, remove extreme abnormal values, preset the threshold value according to a fixed proportion of a statistical mean value based on priority in the dynamic weight association table, and perform batch adjustment according to traffic change and equipment update conditions in a quarter.
  7. 7. The system according to claim 1, wherein the specific process of comparing the same kind of application flow performance of the same-model other equipment on the same route comprises the steps of selecting the same-model other equipment on the same route in the same flight phase, extracting same-phase flow data of the same kind of application, calculating a flow average value and a fluctuation range, calculating a difference value between the same kind of application flow of the abnormal equipment and the average value, judging whether the difference value exceeds or does not exceed a preset fluctuation range, and recording specific numerical values and duration time beyond the preset fluctuation range.
  8. 8. The system according to claim 1, wherein the specific process of calculating the packet loss rate and the time delay data of the satellite link is to collect satellite link transmission data by adopting a sliding time window, count the total amount of transmitted data packets and the total amount of successfully received data packets in the window, calculate the packet loss rate by difference value, record the transmission time and the reception confirmation time of the data packets, calculate the average value of the difference value of the transmission time and the reception confirmation time as the time delay data, and dynamically adjust the window duration according to the flight phase.
  9. 9. The system according to claim 1, wherein the specific process of generating the abnormal flow root cause analysis report is that a multi-dimensional judging condition is set, when the difference value of the similar application flows exceeds a preset range, the abnormal application is judged, when the link parameter exceeds the preset standard range and the similar application of multiple devices is abnormal, the link fluctuation is judged, when and only when the single-device flow exceeds a preset threshold value, the equipment failure is judged, and the abnormal flow root cause analysis report is formed according to the judging result.
  10. 10. The system according to claim 1, wherein the specific process of constructing the bandwidth demand prediction model is: the device comprises an input layer, a factor processing layer, a training layer and an output layer; The system comprises an input layer, a factor processing layer, a training layer, an output layer, a bandwidth demand pre-judging layer, a prediction error range layer and a data processing layer, wherein the input layer is used for multi-source data access, the factor processing layer adopts an outlier detection algorithm to remove invalid data and performs standardized processing on factors with different orders; each layer transmits real-time data through a preset data interface, and retrains according to newly generated historical data periodically.
  11. 11. The system of claim 1, wherein the specific process of integrating the factor-related impact weights of each class is: Screening model adaptation characteristics, route operation data, time period flow change, user behavior habit and application heat trend as core factors; constructing a hierarchical structure of a target layer, a criterion layer and a scheme layer by adopting a hierarchical analysis method, constructing a judgment matrix, and calculating initial weights after consistency test; combining historical abnormal data in the abnormal flow root cause analysis report, calculating the contribution degree of each factor through correlation analysis, and carrying out iterative adjustment on the initial weight; And finally, forming a factor weight distribution scheme matched with the actual scene, adding factors according to the new service scene and re-distributing weights.
  12. 12. The system of claim 1, wherein the building the application high-altitude suitability scoring system comprises the specific processes of setting a core scoring dimension, distributing each dimension scoring weight according to service priority in the dynamic weight association table, formulating each dimension grading scoring standard, calculating each dimension scoring and weighted sum integrated scoring according to the standard by collecting actual operation data of the application in the high-altitude environment, and classifying scoring grades, wherein different grades correspond to different types of optimization suggestions.

Description

Intelligent aviation internet flow monitoring system based on multidimensional correlation analysis Technical Field The invention belongs to the technical field of network communication and intelligent monitoring, and particularly relates to an aviation Internet flow intelligent monitoring system based on multidimensional correlation analysis. Background With the rapid popularization of the aviation internet, satellite networks have become the core support for on-board communication, and on-board terminal types (medical devices, flight aids, entertainment terminals, etc.) and business scenarios (key data transmission, office communication, cabin entertainment, etc.) are increasingly diversified. However, aviation scenes have the particularities of scarce satellite bandwidth, sensitive flight phase (significant service priority difference between take-off/landing and flat flight phase), strong heterogeneity of air network environment (high-altitude time delay and easy beam switching), and the like, and the existing general network flow monitoring technology is difficult to adapt: The traditional monitoring mostly adopts a fixed threshold statistics and isolation analysis mode, lacks dynamic adaptation of aviation scene service priority, cannot realize differential guarantee of key service and non-key service, is characterized in that abnormal flow tracing only stays on surface data investigation, is not combined with multidimensional association analysis such as flight phase, satellite link state, equipment type and the like, is difficult to accurately position root cause, is not integrated with a multi-factor dynamic prediction mechanism due to bandwidth allocation depending on experience pre-judgment, is easy to cause bandwidth congestion in peak time, and meanwhile, the problem of adaptability of ground maturation application in a high-altitude satellite network environment is not effectively mined, and communication stability and user experience are affected. Disclosure of Invention In order to solve the problems in the prior art, the invention provides an intelligent aviation internet flow monitoring system based on multidimensional correlation analysis, The aim of the invention can be achieved by the following technical scheme: the system comprises a dynamic weight association unit, an abnormal flow cross-dimension tracing unit, a bandwidth prediction multi-factor optimization unit and an air-ground collaborative mining unit; The dynamic weight association unit acquires dynamic weight information and constructs a multi-dimensional association frame, and based on the multi-dimensional association frame, a dynamic weight calculation rule is preset to obtain a dynamic weight association table; The system comprises a dynamic weight association table, an abnormal flow cross-dimension tracing unit, an abnormal flow cross-dimension tracing rule system, an automatic calling device, a comparison device, a satellite link packet loss rate and time delay data matching corresponding flight phase characteristics, a multi-dimensional data comparison and verification device, a data analysis device and a data analysis device, wherein the abnormal flow cross-dimension tracing unit constructs an association tracing rule system according to the dynamic weight association table; the bandwidth prediction multi-factor optimizing unit builds a bandwidth demand prediction model through a time sequence association algorithm based on the multi-dimensional association frame and the abnormal flow root cause analysis report, integrates the association influence weights of various factors in the training process, prejudges the bandwidth demands of different airlines and different time periods, identifies the application types and key time periods of the rapid increase of the flow, and forms a targeted bandwidth allocation result; The space-ground collaborative mining unit is used for supplementing a high-altitude network matching attribute tag in a preset application management library by combining the bandwidth allocation result, constructing an application high-altitude suitability scoring system based on the application management library, and finally generating an application replacement suggestion and a protocol optimization scheme. The method comprises the steps of firstly obtaining three types of core data including a real-time flight stage state, an application emergency degree identifier and a device priority classification, then building a three-layer progressive multidimensional correlation framework according to scene dimensions, device dimensions and service dimensions, obtaining core attributes of each dimension, a scene-device adaptation map, association logic for binding device-service functions, embedding classification item indexes and association interfaces in each layer to ensure expansibility, determining basic weights according to the satellite bandwidth bearing upper limit and the required degree of each servi