摘要： 针对网络安全态势量化评估中存在信息不完全、各专家评价不一致及常规 AHP-模糊综合评价法(Analytic Hierarchy Process-Fuzzy Comprehensive Evaluation，AHP-FCE) 中各指标权值相同导致评价精度低等问题，提出一种基于不完全信息非合作博弈的改进型 AHP-FCE 网络安全态势评估模型。首先，基于网络层、主机层和服务层建立了共15个具体指标的网络安全态势指标体系，该指标体系比较全面地反映出了网络异常情况、资产自身脆弱性及外界威胁攻击等方面的风险因子；其次，将各专家对于指标赋权的行为建模成一个非合作博弈行为，采用余弦相似系数来衡量决策者打分结果的协调一致性，构造具有惩罚项的效益函数并利用混合策略纳什均衡的迭代算法得到最优的综合权重。经不同模型对比验证，发现所提模型较传统 AHP-FCE 类方法、目前常用的评价模型和其他博弈论相结合的复合赋权法，在精确性和稳定性上都取得了更好的效果。

中图分类号：TP309文献标志码：ADOI:10.19358/j.issn.2097-1788.2026.08.002中文引用格式：孙雪岩. 不完全信息博弈驱动下的GAHP-FCE网络安全态势评估模型［J］.网络安全与数据治理，2026，45（8）：14-21.
英文引用格式：Sun Xueyan. A GAHP-FCE network security situation assessment model driven by incomplete information game［J］.Cyber Security and Data Governance，2026，45（8）：14-21.

A GAHP-FCE network security situation assessment model driven by incomplete information game

Abstract： To address the problems existing in the quantitative evaluation of network security situation, such as incomplete information, inconsistent expert evaluations, and low evaluation accuracy resulting from equal indicator weights in the conventional Analytic Hierarchy Process-Fuzzy Comprehensive Evaluation (AHP-FCE) method, this paper proposes an improved AHP-FCE network security situation evaluation model based on incomplete-information non-cooperative game. Firstly, a network security situation indicator (NSSI) system with 15 specific indicators is constructed from the network layer, host layer and service layer. This indicator system can relatively comprehensively reflect risk factors concerning network anomalies, asset vulnerabilities and external threat attacks.Secondly, the indicator weighting behavior of experts is modeled as a non-cooperative game. The cosine similarity coefficient is adopted to measure the coordination and consistency of decision-makers′ scoring results. A profit function with penalty terms is constructed, and an iterative algorithm based on mixed strategy Nash equilibrium is utilized to obtain the optimal comprehensive weights.Finally, comparative validations with different models demonstrate that, compared to traditional AHP-FCE methods, commonly used evaluation models, and other composite weighting methods combined with game theory, the proposed model achieves superior performance in both accuracy and stability.

Key words : network security situation quantitative assessment; incomplete information game; Nash equilibrium; AHP-FCE; indicator weighting