在模拟免疫系统的主组织相容复杂性的基础上，结合模糊逻辑与扩展阴性选择算法提出了一个基于免疫系统主组织相容复杂性的模糊逻辑综合决策算法，并用该算法构建了一个实际的基于网络的入侵检测系统。该算法应用高效的扩展阴性选择算法作出第一次网络流量检测，当网络数据异常特征明显时，能直接检测出入侵。若其不能准确地识别待分析数据，则利用具有检测结果准确优势的规则匹配算法作出二次检测，最后结合两次检测结论用模糊逻辑决策模型做出综合决策。A novel algorithm based on the simulation of the major histocompatibility complex feature of human immune system is proposed. The extended negative selection algorithm makes the first decision to decide whether the network traffic is intrusion. Another algorithm, rule match-based alglrithm, makes the second decision while the first decision failed to identify the intrusion. The proposed algorithm, called major histocompatibility complex feature-based synthetic fuzzy decision-making algorithm, composes the first decision and the second decision to make a synthetic fuzzy decision by taking advantage of fuzzy logic.