Study Of HeNB Self-healing Issue Based On CBR Method

In the process of LTE standardization, the concept and techniques of Home eNodeB wereproposed as a new method to solve the indoor coverage which was concerned by various operatorsand manufacturers. HeNB, also known as femtocell (femto cell), is mainly used to supplement themacro base station signal coverage, improve system capacity of the network. Since there may belarge quantities of these devices with different kinds, standard organizations proposed SON(self-organizing network) in order to reduce manual intervention, reduce operation and maintenancecosts and improve user experience.Self-healing is one of the SON functionalities, but researches onself-heling is also less.Therefore,this article proposed a solution to achieve self-healing by usingCBR.The main idea of this solution is to solve current fault by using similar proble m’s solution.This article first analyzes the demand for HeNB self-healing, after that HeNB self-healingarchitecture is proposed.The solution of Fault information upload,fault diagnosis and fault repair isalso proposed. Among them, the fault diagnosis method using CBR, is the focus of this study.Thenwe add a classification mechanisms into fault diagnosis,combined HeNB features,experiments showthat this mechanism can significantly improve the performance of fault diagnosis system.Finally, westudied the three key techniques of failure diagnosis:fault case description,fault case retrieve andretrieve rule amendments. Corresponding strategies and algorithms were proposed.Experimentalresults show that the proposed algorithm’s performance is greater than the traditional one. It canmeet the requirements of case retrivel and provide exact solutions for current fault.Retrieval ruleamendment algorithm can adjust the rules based on retrieval results or expert opinion,it can increasethe accuracy of case retrieval.