Security situation prediction method based on niche technology with fuzzy elimination mechanism

Abstract

The invention discloses a prediction method combining an improved niche genetic algorithm (INGA) and a wavelet neural network (WNN). The method comprises the following steps: selecting the WNN with a stronger nonlinear fitting capacity and a better fault-tolerant performance to predict a security situation; carrying out optimization of traditional WNN parameters through the adaptive genetic algorithm; introducing a novel niche technology with dynamic fuzzy clustering and elimination rules, carrying out niche classification of a population through dynamic fuzzy clustering to form a plurality of niches, and adjusting adaptive values of individuals through a punishment mechanism; and calculating the adaptive value of each niche individual according to classified niches, comparing the adaptive value of each individual with the adaptive value of the optimum individual of a current generation, and eliminating the niches of which the adaptive values are much different from the adaptive value of the optimum individual of the current generation so as to achieve overall elimination of the niches. The method provided by the invention has the advantages that the optimization capacity and convergence rate of the genetic algorithm are improved, the problem of high possibility of premature convergence of the genetic algorithm is solved through a higher population diversity, and the network security situation can be more accurately predicted.

Description

technical field [0001] The invention relates to the field of network security assessment, in particular to a network security situation prediction method. Background technique [0002] With the development of the network, people's daily life is increasingly dependent on the Internet. At the same time, various new attack methods emerge in an endless stream. It is difficult to solve these security problems with a single defense method. In this severe environment, the network security situation Perception technology is widely used. As the core technology of situation awareness, situation prediction is to realize the effective prediction of future events by analyzing historical and current situation information and adding certain technical means. [0003] In order to solve a series of technical problems in situation forecasting, many scholars at home and abroad began to conduct in-depth research on forecasting methods. Olabelurin et al. used the characteristics of DDOS attacks ...

AI Technical Summary

Problems solved by technology

[0004] There are many kinds of forecasting methods. The neural network has been widely used in the field of situation forecasting because of its high adaptability and self-learning ability, and its strong approximation ability to nonlinear functions. However, with the increasing complexity of the network , the neural network parameters obtained by traditional methods may have large deviations, resulting in prediction results that cannot be used as the basis for network security administrators to make correct decisions

[0005] Applying artificial intelligence optimization algorithms to neural networks can solve the problem of parameter optimization to a large extent, but there may still be problems such as insufficient prediction accuracy, slow convergence speed, easy to fall into local optimum and premature convergence. How to solve these problems become the key for scholars to study the combination of intelligent optimization algorithm and neural network

Images