Research On LTE-R System Risk Assessment Method Based On Improved Evidence Theory

With the broadband of today’s communication network,big data,the internet of things and other high technology in railway-related fields of continuous application and progress,the railroad travel interval is getting smaller and smaller.The safety standards for the railroad are getting higher and higher.The Long Term Evolution-Railway(LTE-R)technology is the new research direction of railroad mobile wireless communication,which carries the core communication service of train operation.Its working condition will affect the operation efficiency of trains and even threaten the safety of personnel in severe cases.The LTE-R system risk factors have non-independence and diversity.In order to avoid the passive situation of filling loopholes everywhere after the LTE-R system is put on the line and to intervene and control the risk more comprehensively in advance.The critical ranking of each risk factor in the basis of risk assessment makes the risk can be controlled in a lower range,which has important field application and practical significance.Taking the LTE-R communication system as the research object,the study reviews the risk assessment methods and the development process of the LTE-R technology based on the actual operation in the field.Using the LTE-R system’s network structure composition and risk assessment mechanism as the theoretical foundation,new assessment methods for problems related to the process of the LTE-R system security risk assessment are proposed using technical means such as improved evidence theory,improved structural entropy power method,and nine elements fuzzy mathematics.Based on the three-dimensional indexes of occurrence,predictability,and severity,the improved structural entropy weight method and Multi-Attribute Ideal Real Comparative Analysis(MAIRCA)method rank the importance of risk factors of the LTE-R system.The research in study provides new techniques and ideas for security risk assessment management.The main research work of this thesis is as follows:Firstly,the LTE-R system’s research background,network structure,and advantages are briefly analyzed,then the risk definition and risk assessment theory are explained.According to system engineering theory,railroad safety risk-related standards,and a large amount of literature,the LTE-R system safety risk assessment index system is constructed from four aspects: business factors,network factors,environmental factors,and management factors.Secondly,the methods of determining weights and risk level affiliation are introduced.Based on the relevant principles of risk classification and literature research,the LTE-R system security risk is classified into five levels and described for each level.The structural entropy weight method to determine the weights and the nine elements fuzzy mathematics to determine the affiliation of risk factor levels are used to provide a more accurate database for the study,respectively.The evidence theory is improved in terms of conflict measurement and evidence fusion to address the uncertainty and conflict in the actual evaluation process,and the effectiveness of the improved method is verified by comparison.The improved evidence theory is proposed for the risk assessment.The method is further validated with examples,and the confidence in the evaluation results of the proposed method reaches82.94 %.Compared with the results of other methods,the confidence is increased by 31.43 %,which can more accurately evaluate the security risk of the LTE-R system and obtain evaluation results with higher confidence.Finally,to further improve the efficiency of the comprehensive application of risk assessment in LTE-R system practice,this thesis proposes a ranking method for the importance of risk factors based on the MAIRCA method.The risk factors are measured from indicators of three dimensions to address the problem of relatively single risk indicators in the ranking process.The three-dimensional indicator weights are determined by the structural entropy weighting method.The MAIRCA method is introduced for the calculation to determine the importance ranking results of risk factors.In the example analysis section,the deviation values of risk factors of the LTE-R system are calculated,and the factors are ranked in importance based on the deviation values.The other six sorting methods are compared with the MAIRCA sorting results.The average correlation coefficient of this method reaches0.9831,which is the highest value among the seven sorting methods.The validity of the MAIRCA method is verified.The method proposed in this thesis can correctly and effectively obtain the importance result of risk factors and provide reference for the LTE-R system operation and maintenance.