| The invulnerability of Wireless Sensor Networks(WSN)is an important measurement index for the sustainable and stable provision of reliable services in the network.However,in industrial application scenarios,due to the joint influence of internal factors such as electromagnetic interference,transmission delay,node failure,and network attacks,and external complex environmental interference,WSN is difficult to maintain long-term stable operation.How to evaluate the invulnerability of WSN has become an important issue restricting the large-scale application of WSN.Therefore,this article mainly researches the invulnerability of WSN and its evaluation issues,with specific content as follows:(1)Research on WSN fault diagnosis methods.In the analysis process of WSN data,due to the correlation between the data,the corresponding changes occur when the sensor fails.This article uses these changes and the Belief Rule Base(BRB)modeling method to diagnose WSN faults.At the same time,in order to solve the problem of large deviations between the initial BRB model and the actual situation due to the lack of expert knowledge and difficulty in setting initial parameters,a BRB fault diagnosis method with attribute computation is proposed.This method calculates the initial parameters of the BRB model by analyzing the data set.Finally,experiments prove that this method can effectively identify various wireless sensor network node faults,has a better fault type judgment accuracy rate than other methods,and provides a basis for subsequent WSN invulnerability evaluation.(2)Research on WSN invulnerability evaluation methods.The reasons for network failure in complex environments and the shortcomings of past invulnerability indicators have been analyzed.An evaluation index system of WSN invulnerability under multi-dimensions such as fault phenomena,energy supply,data transmission delay,packet loss rate,and environmental adaptability is constructed,and a WSN invulnerability evaluation model is built using the BRB model.At the same time,in order to solve the combination explosion problem of the BRB model due to too many parameters and the problem of parameter optimization destroying interpretability,an optimization method of the BRB model with activation rate is proposed.This method uses the activation rate to trim rules that have never been activated or are insufficiently activated,introduces parallel computing methods to reduce computational pressure and improve computational accuracy.Experimental results show that the BRB model with an activation rate is not only faster and more accurate than the traditional BRB but also has higher computational efficiency.(3)Analysis of WSN invulnerability evaluation results.Traditional evaluation methods mostly assume that information is completely reliable and ignore the impact of external environmental factors.However,in actual engineering,due to the influence of surrounding complex environments,the results of traditional evaluation methods may not necessarily be stable.By introducing the method of perturbation analysis,this article proposes a robustness analysis method for the BRB model,provides a robustness measure for the BRB model,establishes a perturbation threshold determination model for the BRB model,and optimizes the parameters to determine the upper limit of the perturbation intensity.The model’s robustness analysis is realized.Experimental results show that the BRB model has strong robustness,and the effectiveness of the BRB model output results is guaranteed under the optimal perturbation threshold.(4)Based on the above research content,a WSN invulnerability evaluation demonstration system has been developed. |
