大差异数据冲击下的网络路由分类器的设计与实现

摘要

In view of great error in the path selection of router under impact of large differences data,a network routing classifier under the impact of big difference data is designed. It is mainly composed of FPGA development board,modulus col-lector and digital signal processor. The network routing classification function of the classifier is designed and then written into the FPGA development board. When there is a big difference data impact to network channel,the FPGA development board will call the network routing classification function for pre-classification of large difference data to obtain preliminary classification re-sults and send them to the modulus collector. The modulus collector transforms classification results into low-voltage difference signal. The digital signal processor executes feature extraction and parameter correction of the low-voltage differential signal,so as to obtain the best classification results. The experimental results show that the designed classifier has the characteristics of good channel gain effect,high classification rate and low channel failure rate.%针对大差异数据冲击下路由器存在路径选择误差大的问题,设计大差异数据冲击下的网络路由分类器,其主要由FPGA开发板、模/数采集器和数字信号处理器组成.分类器先对网络路由分类函数进行设计,并将其写入FPGA开发板.当网络信道存在大差异数据冲击时,FPGA开发板将调用网络路由分类函数对大差异数据进行预分类,获取预分类结果并将其传送到模/数采集器.模/数采集器将预分类结果转换成低压差分信号,数字信号处理器对该低压差分信号进行特征提取和参数修正,进而获取最佳分类结果.实验结果表明,所设计的分类器具有信道增益效果好、分类速率高和信道失效率低的特点.