Unstable network robustness evaluation and detection method based on statistical model testing

Abstract

The invention provides an evaluation and detection system of the robustness of a network system, and an operating method of the evaluation and detection system. According to the evaluation and detection system, firstly, a timed automaton of a protocol of the network system is constructed, and an overall working process of the protocol is described; secondly, weighted migration is introduced into a model for expanding the timed automaton; and lastly, a random timed automaton of the network system is constructed for modeling of uncertain factors in the environment. On the basis of the random timed automaton, statistical model testing technology can be used for analysis and evaluation of the performance of the network system, and therefore the objects of parameter configuration, performance comparison and the like are achieved in a design stage and a selection stage of the protocol. According to the operating method of the evaluation and detection system, the performance of the network system in a specific environment is analyzed and evaluated mainly through the modeling of the protocol and the environment in the unstable network system, and therefore the robustness of the network system is evaluated and detected. The statistical model testing technology is based on a simulation and statistical method, and therefore an exhaustive search of the state of the network system is avoided, consumption in time and space is greatly lowered, and the statistical model testing technology can be used for processing a large-scale system.

Description

technical field [0001] The invention relates to the field of network model inspection and protocol verification, and in particular relates to a network system robustness evaluation system and a working method thereof. Background technique [0002] With the development of network technology, large-scale networks are used in more and more fields, such as medical, military, environmental monitoring and protection, etc. All nodes in a large-scale network work together to exchange and share data. In order to enable the high-level application in the network to be carried out correctly, efficiently and credibly, the correctness of the underlying protocol must be guaranteed. In large-scale networks, especially large-scale networks working in the wild and in mobile environments, the working environment often faces uncertain effects such as environmental interference and dynamic networks, and once the network is deployed, it must be verified for maintenance and changes. Therefore, i...

AI Technical Summary

Problems solved by technology

However, since this method needs to traverse all the state spaces of the system, it will face the notorious state space explosion problem, resulting in a severely limited verifiable system scale.

In addition, traditional model checking techniques can only verify the system correctness of the protocol in an ideal environment, that is, there are no uncertain factors such as information loss and node failure.

In a general network, the system has a large scale and works in a dynamic environment, that is, due to the existence of interference factors in the environment, it will face uncertain factors such as system loss and node failure.

Images