中文摘要：

安全存储区域网系统需要保护海量数据，使用现有安全技术存在安全开销过大和存储空间浪费较多等问题，使得现有安全存储区域网系统I／O性能较低，无法使用。人工免疫算法能高效地识别异常数据，为了使安全存储区域网系统能保持较高的I／O性能，本文引入人工免疫算法，研究安全存储区域网系统。给出安全存储区域网中安全结点和基于博弈安全优化模块的结构；再针对安全存储区域网系统的特点，设计结点安全模块中的主要算法，并引入博弈论设计安全存储区域网系统优化算法。实现了结点安全模块的原型系统，测试安全模块的检测效率，验证了结点安全模块在具有较好安全性能的同时，具有安全开销较小且稳定的特点。最后在Lustre上实现了安全存储区域网原型系统，使用通用工具测试的系统I／O性能，结果表明所实现的安全存储区域网原型系统具有能保持较高I／O性能的特性。

英文摘要：

In order to protect enormous data stored in storage area network system, large I/O performance and space are lost using current secure technology. Artificial immune algorithm is used to check out abnormal data efficiently. In order to keep high I/O perform- ance of secure storage area network system, we research artificial immune based secure storage area network system by artificial im- mune algorithm. The structure of artificial immune based secure node and game theory based optimization model are presented. By analyzing the feature of storage area network system, this paper presents main algorithms to ensure security of node and optimization algorithm for storage area network system. Realizing prototype of secure model for node in storage area network system and testing the efficiency of checking out abnormal access request, the result shows that the secure model of node is efficient with stable time and space overhead. Realizing prototype of secure storage area network system based on Lustre and evaluating its I/O performance. The result shows that it can keep high I/O performance.