Search

CN-122027494-A - Cascade fault defense and self-healing optimization method for space-time integrated network

CN122027494ACN 122027494 ACN122027494 ACN 122027494ACN-122027494-A

Abstract

The invention discloses a cascade fault defense and self-healing optimization method for an aerospace-earth integrated network. Aiming at the problem that cascade faults are easily caused by space-earth integrated network node heterogeneous coupling and topology dynamic change, a multi-factor coupled cascade fault propagation dynamic model is firstly established, the synergistic effect of node isomerism, link dynamics and load redistribution is quantified, then a closed loop cooperative defense system integrating early warning, isolation and self-healing is designed, intelligent decision is made by utilizing machine learning, reinforcement learning and game theory, instantaneity and energy efficiency on a resource limited edge node are ensured through a lightweight algorithm, and finally an NS-3 and MATLAB joint simulation platform is established, and the model and strategy are verified under various application scenes of the Internet of things. Experimental results show that the prediction error of the propagation model is controlled within 10%, the fault recovery rate is improved to 85% -90% by the defense strategy, the communication interruption time is reduced by more than 60%, the resource cost is reduced by 20% -25%, and theoretical support and technical guarantee are provided for the space-earth integrated network stable operation in the safety and toughness critical environment.

Inventors

  • ZHANG PEIYING
  • YU YIHONG
  • Tan Lizhuang
  • ZHANG HONGXIA

Assignees

  • 中国石油大学(华东)

Dates

Publication Date
20260512
Application Date
20260311

Claims (5)

  1. 1. The cascade fault defense and self-healing optimization method for the space-to-ground integrated network is characterized by comprising the following steps of: step 1, establishing a multi-factor coupling cascade fault propagation dynamic model of an air-space-ground integrated network; step 2, designing a closed loop cooperative defense system integrating early warning, isolation and self-healing; And 3, constructing an NS-3 and MATLAB joint simulation platform for verifying the model and the defense strategy.
  2. 2. The method for defending and self-healing optimization of cascade faults oriented to space-earth integrated networks according to claim 1, wherein the establishing of the multi-factor coupling cascade fault propagation dynamic model in the step 1 is specifically as follows: Setting node load For calculating task load Data transmission load And energy consumption load Is expressed as a weighted sum of Wherein the weight coefficient The node type and the application scene are determined; Setting node capacity To calculate the capacity Transmission capacity And energy capacity Is expressed as the minimum value of Setting the node initial fault probability model as Wherein the method comprises the steps of In order for the run-time to be run, In order to achieve a rate of consumption of the resource, For the intensity of the disturbance of the external environment, Is a scale factor; Setting fault propagation probability Representing nodes Is propagated to the node Is defined by the node Received migration load Self capacity of Link attenuation coefficient Decision, expressed as Introducing a dynamic evolution formula of the load to describe the change of the load with time, Wherein the method comprises the steps of For the purpose of adding a load to the outside, Load migrated to the neighbor node; Introducing a time and resource based recovery probability model Wherein the method comprises the steps of Is a recovery rate coefficient positively correlated to the node resource; Leading in a pre-warning threshold formula based on multi-source information fusion, if the load is predicted Or predict link quality Triggering early warning, wherein As a safety factor, the safety factor of the device, Is a link quality threshold.
  3. 3. The cascade fault defense and self-healing optimization method for an aerospace-earth integrated network according to claim 1, wherein the design in the step 2 integrates early warning-isolation-self-healing into a closed loop cooperative defense system, and the method is specifically as follows: The early warning stage combines the centrality of the medium number Center of degree And functional dependency Is a comprehensive index of (2) The importance of the nodes is evaluated, and priority resource allocation and redundancy backup are carried out according to the importance of the nodes; The isolation stage actively transfers partial loads of the risk nodes to healthy adjacent nodes after detecting the fault precursor, and dispatches the idle unmanned aerial vehicle to a fault-prone area to establish a temporary link; the self-healing stage realizes quick fault location through link quality monitoring and node heartbeat feedback, adopts dynamic topology reconstruction based on greedy optimization strategy, preferentially selects near-healthy nodes with sufficient residual capacity and reliable links for reconnection, starts redundant links for core link faults, and dynamically adjusts bandwidth according to service importance and link quality.
  4. 4. The method for cascade fault defense and self-healing optimization of space-time integrated network according to claim 3, wherein the lightweight algorithm adopted in the self-healing stage has a time complexity of For fault detection and localization, wherein For the number of network nodes to be counted, The time complexity of the dynamic topology reconstruction algorithm is that The spatial complexity of the algorithm is To meet the engineering deployment requirements of the resource-constrained nodes.
  5. 5. The cascade fault defense and self-healing optimization method for the space-time integrated network, which is disclosed by claim 1, is characterized in that in the step 3, an NS-3 and MATLAB joint simulation platform is constructed, and the method is specifically used for configuring heterogeneous parameters and motion models of satellites, air platforms and ground nodes in the NS-3, simulating link transmission delay, bandwidth and packet loss rate, and carrying out data processing, model solving and visual presentation of results in the MATLAB, so as to realize quantitative verification and performance evaluation of the cascade fault propagation model and closed loop cooperative defense system.

Description

Cascade fault defense and self-healing optimization method for space-time integrated network Technical Field The invention belongs to the technical field of network communication, and particularly relates to a network reliability and toughness enhancement method under an air-ground integrated network scene, in particular to a propagation modeling, defense and self-healing optimization method of cascading faults. Background As a core information infrastructure of sixth-generation mobile communication (6G), the air-space integrated network realizes global coverage and ubiquitous connection by cooperatively integrating space-base, air-base and foundation network resources, and provides information guarantee for key Internet of things applications such as smart cities, telemedicine, intelligent transportation, industrial automation and the like. However, due to the characteristics of space-air-ground integrated network node heterogeneous coupling, topology dynamic change and the like, cascading faults, namely failure of single or few nodes, can cause chain reaction due to the mutual dependency relationship among the nodes, cause paralysis of a large-scale network and seriously threaten the continuity of key business. The existing research on cascade faults faces four key limitations: (1) The traditional model is mostly based on static and isomorphic topological assumptions, and cannot capture the influence of differences of empty, space and ground nodes on physical properties, resource capacity and fault threshold values on fault propagation paths. (2) Insufficient modeling of dynamic topology evolution, satellite motion, air platform repositioning and ground user mobility, result in network frequency heavy traffic, and traditional static frameworks have difficulty analyzing the spread of faults in time-varying networks. (3) The system lacks a cooperative defense mechanism, the current strategy aims at single faults or adopts isolated countermeasures, a comprehensive framework integrating early warning, self-adaptive isolation and self-healing is lacking, and system-level toughness is difficult to form. (4) Neglecting satellite-borne resource constraint, namely strictly limiting satellite and air platform resources, most intelligent defense methods assume that the resources are unlimited, so that algorithm complexity is too high, and practical deployment is difficult. Therefore, how to accurately describe the propagation rule of cascading failures in the environment of the space-earth integrated network with limited resources, high dynamic performance and isomerism, and design efficient and cooperative defense and self-healing strategies becomes a technical problem to be solved currently. Disclosure of Invention The invention aims to overcome the defects of the prior art and provide an optimization method capable of accurately modeling cascade fault propagation in an air-space-earth integrated network and realizing efficient, cooperative defense and self-healing. The solution for realizing the purpose of the invention is that a cascade fault defense and self-healing optimization method for an air-space integrated network comprises the following steps: step 1, establishing a multi-factor coupling cascade fault propagation dynamic model of an air-space-ground integrated network; step 2, designing a closed loop cooperative defense system integrating early warning, isolation and self-healing; And 3, constructing an NS-3 and MATLAB joint simulation platform for verifying the model and the defense strategy. Further, in the step 1, a multi-factor coupling cascade fault propagation dynamic model is built, which is specifically as follows: First, nodes and links of an air-space integrated network are modeled. The space-ground integrated network architecture is divided into a space base layer (comprising low-orbit, medium-orbit and high-orbit satellites), a space base layer (comprising unmanned aerial vehicles and high-altitude airships) and a ground base layer (comprising 5G base stations, core network equipment and end user equipment). The nodes have high heterogeneity, and the parameters of bandwidth, time delay, packet loss rate and the like of links (inter-satellite links, satellite-ground links, air-ground links and ground links) have significant differences. To accurately characterize fault propagation, the present invention defines node loadsFor calculating task load(CPU utilization), data transfer payload(Bandwidth occupancy) and energy consumption loadWeighted sum of (inverse of the remaining energy ratio): Wherein the weight coefficient Depending on the node type and application scenario. Node capacityI.e. the maximum load that a node can handle, is defined as the minimum of calculated capacity, transmission capacity and energy capacity: the initial failure node is selected according to a node failure probability model, the probability being dependent on node runtime Resource consumption rateAnd external envir