To address the problem that relying solely on the quantity of strategies is insufficient to fully characterize decision advantage in decision-centric warfare (DCW), the combined influence mechanism of strategy quantity and strategy differentiation on asymmetric zero-sum game payoffs is investigated. A game-theoretic quantitative model is constructed, in which combatants are mapped to players and kill-chain combinations are mapped to strategy sets. The “circle-box” method is used to derive the induction mechanism of strategy characteristics on mixed strategy equilibrium. Furthermore, the marginal effects of strategy quantity and differentiation on Nash equilibrium payoffs are analyzed based on the geometric properties of linear programming feasible regions. The model is validated through ordinary least squares (OLS) regression using 4 320 game samples generated via Monte Carlo simulation. Results indicate that game payoffs exhibit a significant positive correlation with the logarithms of both the quantity and differentiation of one’s own strategies, while demonstrating a significant negative correlation with the logarithm of the opponent’s strategy quantity. The mechanism of offsetting an opponent’s scale advantage through maximizing strategy heterogeneity in asymmetric situations is effectively quantified, providing a computable analytical framework for the study of DCW theory.
强对抗战场环境下,传统作战体系中心节点与链路脆弱、应变迟滞的短板日益凸显。为适配高动态对抗,美国国防高级研究计划局(Defense Advanced Research Projects Agency, DARPA)于2017年提出马赛克战概念,美国战略与预算评估中心(Center for Strategic and Budgetary Assessments, CSBA)于2019年正式提出决策中心战(Decision-Centric Warfare, DCW)[1-3],并视马赛克战为其核心路径。决策中心战是聚焦于决策对抗领域、以夺取决策优势为目的的新型作战概念。区别于传统消耗战,决策中心战聚焦决策环节,通过人工智能与自主系统进行可组合兵力设计,在杀伤网中动态构建杀伤链,向对手持续施加复杂性,使其无法预判己方作战意图与行动路径,最终陷入决策困境、丧失战场主动权。
基于收集的仿真数据集,采用普通最小二乘法(Ordinary Least Squares, OLS)对综合收益模型进行回归拟合。回归分析结果如表1所示。模型的决定系数,统计量为5 631(),残差标准误(Residual Standard Error, RSE)为0.072。由于在随机博弈仿真中,收益不仅取决于策略数量和差异性等策略整体结构情况,还受到具体收益数值随机性的强影响。因此,在含有大量随机噪声的博弈系统中,该对数线性模型能够有效解释博弈收益约79.7%的决定系数,表明模型决定系数良好。
CLARKB, PATTD, SCHRAMMH. Mosaic warfare: exploiting artificial intelligence and autonomous systems to implement decision-centric operations[R]. Washington, USA: Center for Strategic and Budgetary Assessments, 2020:1-66.
[2]
CLARKB, PATTD, WALTONT A. Implementing decision-centric warfare: elevating command and control to gain an optionality advantage[R]. Washington, USA: Hudson Institute, Center for Defense Concepts and Technology, 2021:1-49.
[3]
VALENCIAJ T. Engineering decision-Centric avition operations for expeditionary advanced base operations[D]. Monterey, USA: Naval Postgraduate School, 2024:1-107.
LIX L, LIB, SUL. Dynamics of a Cournot game with bounded rational firms and various scale effects[J]. PLoS One, 2024,19(5):No.e0297275.
[7]
AZMIM B, KUMARIS, AQUILS, et al. KIFC1 inhibition: exploring the potential of propolis-derived small molecules for targeting cancer progression through in silico analysis[J]. PLoS One, 2025,20(6):No.e0324678.
KHLOPONINP. Vector space proximity based document retrieval for document embeddings built by Transformers[D]. Montreal, USA: Concordia University, 2022:1-88.
[10]
HO E, RAJAGOPALANA, SKVORTSOVA, et al. Game theory in defence applications: a review[J]. Sensors, 2022,22(3):No.1032.