A reliability evaluation method of trunked system based on complex network

Abstract

The invention provides a cluster system reliability evaluation method based on complex network, which mainly comprises the following steps: A, analyzing the evaluation object to construct a system network model; B, setting reliability index; C, setting a fault injection strategy for reliability evaluation; D, carrying out reliability evaluation and monitoring and recording performance indexes of that system; E: summarizing the change trend of performance index and analyzing the reliability performance of cluster system; according to the characteristics of the cluster system, the proposed reliability assessment method combines the concept of reliability with complex network, the index of sub-Dalian sub-cluster in complex network is used as the criterion to judge the collapse of cluster system, by randomly selecting the node injection fault of cluster system, setting the intensity and proportion of fault injection, calculating the size of sub-Dalian sub-cluster, when the sub-Dalian sub-cluster reaches the maximum, the system collapse is confirmed, and the reliability of cluster system is obtained by analyzing the intensity and proportion of fault injection when the critical collapseoccurs.

Description

technical field [0001] The invention proposes a method for evaluating the reliability of a cluster system based on a complex network, which relates to a method for evaluating the reliability of a cluster system based on a complex network, and belongs to the fields of complexity science and reliability. Background technique [0002] The cluster system is a complex system whose structure has the characteristics of wide distribution, specific structure, good scalability, and high autonomy in two dimensions of time and space, and has good adaptability to complex and changeable environments. Its advantage is that by setting simple rules for simple low-intelligence individuals, it can reflect the emergence of system-level capabilities at the macro level, and it also has better performance in terms of state depth perception and real-time control of the network. Due to its complex structure , the resistance to short-term shocks and disturbances is better than ordinary simple systems...

AI Technical Summary

Problems solved by technology

For a complex cluster system, although its complex structure makes it have a high anti-interference ability, it also brings difficulties to the existing system reliability evaluation:

[0004] 1. Complex faults: the structure of the cluster system is a hierarchical network structure, and the fault mechanisms and effects in each layer are different;

[0005] 2. Complex structure: Due to the coupling relationship between the cluster system and the real physical world, there are connections between software nodes and hardware nodes, and the topology structure between nodes is complex. Traditional reliability evaluation methods are mainly aimed at single hardware or single software. The evaluation did not take into account the coupling relationship between the two;

[0006] 3. Complex propagation: faults are propagated between nodes through network connections, etc. Traditional reliability analysis and evaluation such as failure mode, impact and hazard analysis (FMECA) and fault tree analysis (FTA) are mostly used in static Under the condition of single fault and single fault, less consideration is given to the situation of fault propagation, and the mutual influence between network faults cannot be ignored

Images