摘要
由于受到感知设备采集数据多样性以及各类干扰因素等影响,实验室网络交换机测试数据自动采集方法的效率低、抗干扰性能差等缺点严重桎梏了数据挖掘技术的发展,提出了基于遗传算法和熵的实验室网络交换机测试数据采集方法。方法将实验室网络交换机测试数据作为原始种群,采用遗传算法中的交叉算子和变异算子扩充原始种群,并对扩充后的种群做分组处理,形成候选种群,以候选种群为单位分别评估其染色体特性继承情况,将染色体特性继承情况最优的候选种群作为最终种群返回。采用基于熵的函数来度量种群之间的相似度,在相似度指导下采用一致性、完整性和准确性三个数据质量判断标准判断采集获得的测试数据质量,将符合标准的种群保留,完成自动采集。仿真测试结果表明,所提方法能够实现实验室网络交换机测试数据自动采集,具有较强的抗干扰性能,采集效率更高,数据质量更好。
At present, method of automatic collection of test data of laboratory network switch has many disadvantages ,such as low efficiency and poor anti-interference ability. In order to solve the problem, this research puts forward a new collection method based on genetic algorithm and entropy. Firstly, the method used the test data as original population and extended the original population via crossover operator and mutation operator in the genetic algorithm. Meanwhile, the research carried out packet processing for the extended population to form candidate population. The research to evaluated inheritance situation of chromosome feature respectively using the candidate population as unit and returned the candidate population with optimal inheritance situation as final population. Then, the research measured similarity among populations via function based on entropy. The research judged quality of the obtained test data using consistency, integrity, and accuracy under the guidance of the similarity and reserved the population meeting the standard. Finally, we completed the automatic collection. Simulation results show that the method can achieve the automation collection. It has better anti-interference performance. Collection efficiency is higher and data quality is better.
作者
王强
WANG Qiang(Guangdong normal University of Technology,Guangdong Guangzhou 510000,Chian)
出处
《计算机仿真》
北大核心
2019年第7期371-375,共5页
Computer Simulation
关键词
实验室
网络交换机
测试数据
自动采集
Laboratory
Network switch
Test Data
Automatic collection
