基于博弈论的欺骗性防御策略优化研究综述

打开文本图片集

中图分类号：TP399 文献标志码：A 文章编号：1001-3695（2026）01-001-0001-10

doi：10.19734/j. issn.1001-3695.2025.06.0199

Survey on optimization of deceptive defense strategies based on game theory

Yuan Ruomeng1，2.3，Fan Jing1，2.3t，Peng Yu1，2，3，Xu Xing1，2，3 （1.ColegeofricldIfatooognUesityi;nKeybU manedAutooseugOhnoatofadatoriti forUniversities，Kunming 65050o，China）

Abstract：Bygrowingsophisticationofcybersecuritythreats，traditionaldeceptivedefensetechnologiesstrugletocounter APTsandcomplexnetworkscenariosduetotheirinsuficientdynamicadaptabilityinstrategies.Gametheoryprovidestheoretical supportfordefenseoptimizationbymodeling the interactive dynamicsof atack and defense strategies，yetitfaceschalen gessuch as state space explosion and inadequate modeling of incomplete information.The advancementof machine learning technology offrs newapproachestoaddressthese issues.This papersystematicallysummarizedtheresearch progressin strategyselection methods fornetwork deceptivedefense，reviewedthecurrentresearch statusof deceptive defenseand itsgametheoreticmethodologies，and proposedanintegratedframework thatcombinedgametheoryandmachinelearningfordeceptive defensestrategydesign whileintroducing typicalapplicationscenarios.Aditionall，itanalyzedand forecastedtheintegrated methodsofgametheoryand machinelearning，andconcluded withacomprehensivesummaryof theentire study.Thisesearch aimstoprovideareferenceforconstructingadefensesystemthatbalancestheoreticalinterpretabilityandpracticalalicability.

Keywords：cybersecurity；deceptive defense;game theory；machine learning

0 引言

随着网络安全威胁越来越复杂，传统的欺骗性防御技术因其策略固定、动态适应性不足，难以应对高级持续性威胁和攻击者不断变化的攻击手段[1]。（剩余30214字）