【摘要】 针对现有入侵响应系统难以兼顾安全性与服务质量的矛盾，提出一种基于阈值约束与均衡分配的优化方法。该方法构建融合节点重要性与风险因子的安全度量模型，并将其与时延、能耗和资金开销相结合，通过阈值约束将多目标优化转化为单目标问题；基于大学招生匹配模型，设计了攻击与防御两种均衡分配算法，采用延迟接受策略在安全与性能之间动态平衡。大规模仿真实验结果表明，新方法在降低安全风险的同时，有效控制多维服务质量开销，与现有基准相比综合效用和自适应能力显著增强，其研究成果可为网络运维提供理论依据和工程参考。更多还原

【Abstract】 Existing intrusion-response systems often face conflicting demands between robust security and service-quality metrics. To address these issues, an optimization method based on threshold constraints and balanced allocation was proposed. A security model integrating node importance and risk factors was constructed and combined with latency, energy consumption, and financial cost. Through threshold constraints, the multi-objective optimization problem was then transformed into a single-objective optimization problem. Inspired by the college admission problem, two equilibrium allocation algorithms—AttackMatcher and DefenseMatcher—were developed with a deferred-acceptance strategy to dynamically trade off security and performance. Largescale simulation results demonstrate that the new method effectively reduces security risks while controlling multi-dimensional quality of service costs. Compared to existing baselines, it shows significant enhancements in comprehensive utility and adaptive capability. Experimental results verify its theoretical and engineering applicability to real-world network operations.更多还原