中文摘要：

拒绝服务（Denial of Service,Do S）是企图使其预期用户的一台主机或其他网络资源不可用,如临时或无限期地中断或暂停连接到因特网主机的服务.为了有效地阻止Do S攻击,首先需要提高Do S攻击检测的准确性,提出一种基于改进Kohonen网络的Do S攻击检测算法.该方法通过对Do S攻击原始数据的预处理,为后续数据处理的方便和保证程序运行时加快收敛奠定必要的基础,采用检测结果的正确率作为该算法的评价指标,采用SOM学习算法是把高维空间的输入数据映射到低维神经网络上,并且保持原来的拓扑次序,然后建立S-Kohonen（Supervised-Kohonen）神经网络检测模型.实验结果表明,与传统的Kohonen方法相比,S-Kohonen网络具有更好的检测性能.

英文摘要：

Denial of service attack is an attempt to make a computer or other network resource unavailable to its intended users, such as to temporarily or indefinitely interrupt or suspend services of a host connected to the Internet. In order to effectively prevent DoS attacks, we first need to improve the accuracy of DoS attack detection. So a detection algorithm is proposed for DoS attacks, which based on improvements supervised Kohonen network,in this paper. The method reduce the complexity of the original data of DoS attacks by the normalized processing for convenient of the data processing. Detection accuracy is used for the evaluation of the algorithm. SOM learning algorithm strengthen the mapping of existing model and weaken the mapping of previous model. Then S-Kohonen neural net-work detection model is established. Experimental results show that compared with the traditional method of Kohonen, S-Kohonen net- work has more better detection performance.