电磁频谱数据的关联规则挖掘

为更合理利用频谱资源以及更好地评估各类电磁环境，本文提出一种基于关联规则挖掘的频谱数据挖掘方案.该方案首先基于一般挖掘流程获取频谱数据中的有用信息，包括异常信息、底噪信息、占用度信息和预定时间功率信息等；再将频谱信息作为关联分析对象，通过构建关联库，构建模糊集，基于模糊关联规则挖掘算法对频谱信息进行系统性的分析.本文对传统的算子选择策略加以改进，使用大尺度参数改进模糊隶属函数.通过实测数据集的验证分析，实验结果表明，频谱信息的强关联规则能反映各种信息之间隐含的关联性以及各种信息出现的频次；基于频谱信息的关联规则挖掘能有效地简化频谱挖掘工作，通过各种信息的关联性可以通过分析一部分频谱信息而得到另外的频谱信息.频谱信息的关联规则可以用于进行电磁无线电环境的评估，选择合适的频谱信息该方案可以应用于各类电磁环境的评估.

Mining association rules for electromagnetic spectrum data

In order to make more rational use of spectrum resources and better evaluate various electromagnetic environments, this paper proposes a spectrum data mining scheme based on association rule mining. Firstly, based on the general mining process, the scheme obtains the useful information in spectrum data, including anomaly information, bottom noise information, occupancy information, predetermined time power information and so on. Then, the spectrum information is taken as the association analysis object, the association database and fuzzy set are constructed, and the spectrum information is systematically analyzed based on the fuzzy association rule mining algorithm. In this paper, the traditional operator selection strategy is improved, and the large-scale parameters are used to improve the fuzzy membership function. Through the verification and analysis of the measured data set, the experimental results show that the strong association rules of spectrum information can reflect the implicit association between various information and the frequency of various information. Association rule mining based on spectrum information can effectively simplify spectrum mining. Through the correlation of various information, another spectrum information can be obtained by analyzing a part of spectrum information. At the same time, the association rules of spectrum information can be used to evaluate the electromagnetic radio environment. By selecting the appropriate spectrum information, the scheme can be applied to the evaluation of various electromagnetic environments.