基于双向联想记忆网络的航空雷达在线入侵诊断方法研究

Aviation Radar Online Invasive Diagnosis Method Based on BAM Network 

摘要:

针对传统的航空雷达网络面临的入侵威胁,以及雷达网络存在的入侵诊断检测效率较低,数据匹配速度较慢等问题,提出了一种基于双向联想记忆网络的航空雷达在线入侵诊断方法,构建航空雷达在线入侵诊断模型,对航空雷达网络中的外部数据进行预处理,并获取数据特征以及数据特征的可辨识属性矩阵和决策辨识函数,计算测试参数集的所有特征向量,从而使入侵检测算子的匹配量减少,以此提升数据匹配效率,实现对外部入侵数据的过滤检测,从而对雷达数据网络进行在线监控,有效抵御外部异常数据的入侵,保证了航空雷达网络的安全性；仿真结果表明文章方法有效提高了航空雷达网络的在线数据检测匹配速度,诊断准确率达到93.3%,且对航空雷达的入侵诊断检测效率、误报率、漏报率等方面都有明显改善。

Abstract:

In view of the traditional aviation radar network invasion threat, and intrusion diagnostic test of radar network low efficiency, data matching speed is slow, this paper proposes an aviation radar online invasive diagnosis method based on BAM network, build invasion diagnosis model of aviation radar online, preprocessing the external data in aviation radar network, and obtain data features and characteristics of discernibility matrix and decision attribute recognition function, all the characteristic vector calculation test parameter set, so that reduce the amount of matching the intrusion detection operator, to enhance the efficiency of data matching, external intrusion data filtering detection, thus for on-line monitoring of the radar data network, effectively resist the invasion of external abnormal data, make sure the safety of aviation radar network. The simulation results show that the method effectively improves the matching speed aviation radar network data online detection, diagnosis accuracy rate reached 93.3%, and the invasion of the aviation radar detection diagnosis efficiency, the rate of false alarm, non-response rates etc have been improved significantly.