正在加载图片...
第3期 丁俐夫，等：多智能体系统安全性问题及防御机制综述 ·431· I)分布式人工智能仍面临诸多风险，新的 information[Cl//Proceedings of the 2017 IEEE 20th Inter MAS安全威胁可能涌现。由于MAS的工作机 national Conference on Intelligent Transportation Systems. 制，智能体需要在开放环境中协作，一方面，智能 Yokohama,Japan,2018:1-6. 体的自治性给系统安全带来了不确定性，另一方 [6]姚尧，卢淑娟，徐德民.IMMAS算法在复杂环境下路径 面，MAS在很大程度上依赖于通信，并且访问控 规划中的应用[J.计算机仿真，2007,24(12)：148-151. 制易受影响。因此，潜在的安全威胁可能造成非 YAO Yao,LU Shujuan,XU Demin.IMMAS for robot 法资源占用、窃取隐私数据、恶意控制甚至停机 path planning in complex environment[J].Computer simu- lation,2007,2412):148-151 等现象。 [7]IQBAL S,ALTAF W,ASLAM M,et al.Application of in- 2)MAS安全评估依旧是难点所在。由于 telligent agents in health-care:review[J].Artificial intelli- MAS的系统复杂性，MAS设计人员将需要重视 gence review,2016,46(1):83-112. 其系统安全性能评估。通用的安全评估标准，需 [8]BIJANI S,ROBERTSON D.A review of attacks and se 要进行统一的量化，以便MAS防御体系的应用 curity approaches in open multi-agent systems[J].Artifi- 推广。 cial intelligence review,2014,42(4):607-636. 3)在安全的MAS设计过程中，应合理平衡算 [9]HUYNH T D.JENNINGS N R,SHADBOLT N R.An in- 法的安全性和实用性，在尽可能少地牺牲MAS tegrated trust and reputation model for open multi-agent 系统性能的情况下建立安全机制，以更好地满足 systems[J].Autonomous agents and multi-agent systems 实际应用需求。 2006,13(2):119-154. 4)为MAS建立合理的信任机制是如今MAS [10]WANG S,HU J,LIU A,et al.Security frame and evalu- 研究的热点之一。利用智能体的社交性，建立智 ation in mobile agent system[C].International Confer- 能体社区信任评价体系，在保证系统性能的同 ence on Mobile Technology Applications and Systems 时，减小安全隐患。 Guangzhou,China,2005:1-6 5)充分考虑分布式方法在MAS安全性问题 [11]JUNG Y,KIM M,MASOUMZADEH A,et al.A survey of security issue in multi-agent systems[J].Artificial intel- 中的应用，通过建立分布式安全模型，增强MAS ligence review,2012,37(3):239-260. 安全性与自我管理功能的同时，实现采用MAS [12]Foundation for Intelligent Physical Agents(FIPA).FIPA 方法保护分布式系统的目标。 Abstract Architecture Specification 2002 [EB/OL].(2002- 参考文献： 02-25)[2018-12-01]http://www.fipa.org/specs/fipaoo001/ SC00001L.pdf. [1]GREENBERG MS,BYINGTON JC.HARPER D G.Mo- [13]Foundation for Intelligent Physical Agents(FIPA).FIPA bile agents and security[J].IEEE communications ACL Message Structure Specification [EB/OL].(2000- magazine,1998,36(7):76-85. 01-08)[2018-12-01]http//www.fipa.org/specs/fipa00061/ [2]CENTENO R,FAGUNDES M,BILLHARDT H,et al. SC00061G.html. Supporting medical emergencies by MAS[C]//Proceedings [14]Foundation for Intelligent Physical Agents(FIPA).FIPA of the 3rd Kes International Symposium on Agent and agent management specification [EB/OL].(2004-03-18) Multi-Agent Systems:Technologies and Applications. [2018-12-01]http://www.fipa.org/specs/fipa00023/ Uppsala,Sweden,2009:823-833. SCO0023K.pdf. [3]DIOUBATE M,TAN Guanzheng,MOHAMED L T.An [15]MOURATIDIS H,GIORGINI P.MANSON G.Model- artificial immune system based multi-agent model and its ling secure multiagent systems[Cl/Proceedings of the 2nd application to robot cooperation problem[C]//Proceedings International Joint Conference on Autonomous Agents of the 2008 7th World Congress on Intelligent Control and and Multiagent Systems.Melbourne,Australia,2003 Automation.Chongqing,China,2008:3033-3039. 859-866. [4]SZYMAK P,PRACZYK T.Control systems of underwa- [16]BRESCIANI P.GIORGINI P,MOURATIDIS H,et al. ter vehicles in multi-agent system of underwater inspec- Multi-agent systems and security requirements analysis[M tion[C]//Proceedings of the 11th WSEAS International LUCENA C,GARCIA A,ROMANOVSKY A,et al. Conference on Automatic Control,Modelling and Simula- Software Engineering for Multi-Agent Systems II.Berlin, tion.Istanbul,Turkey,2009:153-156. Germany:Springer,2004:35-48. [5]HUANG Xiao,ZHANG Qi,WANG Yu.Research on [17]POSLAD S,CHARLTON P,CALISTI M.Specifying multi-agent traffic signal control system based on VANET standard security mechanisms in multi-agent systems[M1) 分布式人工智能仍面临诸多风险，新的 MAS 安全威胁可能涌现。由于 MAS 的工作机 制，智能体需要在开放环境中协作，一方面，智能 体的自治性给系统安全带来了不确定性，另一方 面，MAS 在很大程度上依赖于通信，并且访问控 制易受影响。因此，潜在的安全威胁可能造成非 法资源占用、窃取隐私数据、恶意控制甚至停机 等现象。 2) MAS 安全评估依旧是难点所在。由于 MAS 的系统复杂性，MAS 设计人员将需要重视 其系统安全性能评估。通用的安全评估标准，需 要进行统一的量化，以便 MAS 防御体系的应用 推广。 3) 在安全的 MAS 设计过程中，应合理平衡算 法的安全性和实用性，在尽可能少地牺牲 MAS 系统性能的情况下建立安全机制，以更好地满足 实际应用需求。 4) 为 MAS 建立合理的信任机制是如今 MAS 研究的热点之一。利用智能体的社交性，建立智 能体社区信任评价体系，在保证系统性能的同 时，减小安全隐患。 5) 充分考虑分布式方法在 MAS 安全性问题 中的应用，通过建立分布式安全模型，增强 MAS 安全性与自我管理功能的同时，实现采用 MAS 方法保护分布式系统的目标。 参考文献： GREENBERG M S, BYINGTON J C, HARPER D G. Mobile agents and security[J]. IEEE communications magazine, 1998, 36(7): 76–85. [1] CENTENO R, FAGUNDES M, BILLHARDT H, et al. Supporting medical emergencies by MAS[C]//Proceedings of the 3rd Kes International Symposium on Agent and Multi-Agent Systems: Technologies and Applications. Uppsala, Sweden, 2009: 823–833. [2] DIOUBATE M, TAN Guanzheng, MOHAMED L T. An artificial immune system based multi-agent model and its application to robot cooperation problem[C]//Proceedings of the 2008 7th World Congress on Intelligent Control and Automation. Chongqing, China, 2008: 3033–3039. [3] SZYMAK P, PRACZYK T. Control systems of underwater vehicles in multi-agent system of underwater inspection[C]//Proceedings of the 11th WSEAS International Conference on Automatic Control, Modelling and Simulation. Istanbul, Turkey, 2009: 153–156. [4] HUANG Xiao, ZHANG Qi, WANG Yu. Research on multi-agent traffic signal control system based on VANET [5] information[C]//Proceedings of the 2017 IEEE 20th International Conference on Intelligent Transportation Systems. Yokohama, Japan, 2018: 1–6. 姚尧, 卢淑娟, 徐德民. IMMAS 算法在复杂环境下路径 规划中的应用 [J]. 计算机仿真, 2007, 24(12): 148–151. YAO Yao, LU Shujuan, XU Demin. IMMAS for robot path planning in complex environment[J]. Computer simulation, 2007, 24(12): 148–151. [6] IQBAL S, ALTAF W, ASLAM M, et al. Application of intelligent agents in health-care: review[J]. Artificial intelligence review, 2016, 46(1): 83–112. [7] BIJANI S, ROBERTSON D. A review of attacks and security approaches in open multi-agent systems[J]. Artificial intelligence review, 2014, 42(4): 607–636. [8] HUYNH T D, JENNINGS N R, SHADBOLT N R. An integrated trust and reputation model for open multi-agent systems[J]. Autonomous agents and multi-agent systems, 2006, 13(2): 119–154. [9] WANG S, HU J, LIU A, et al. Security frame and evaluation in mobile agent system[C]. International Conference on Mobile Technology Applications and Systems. Guangzhou, China, 2005: 1–6. [10] JUNG Y, KIM M, MASOUMZADEH A, et al. A survey of security issue in multi-agent systems[J]. Artificial intelligence review, 2012, 37(3): 239–260. [11] Foundation for Intelligent Physical Agents (FIPA). FIPA Abstract Architecture Specification 2002 [EB/OL]. (2002- 02-25)[2018-12-01]http://www.fipa.org/specs/fipaoo001/ SC00001L.pdf. [12] Foundation for Intelligent Physical Agents (FIPA). FIPA ACL Message Structure Specification [EB/OL]. (2000- 01-08)[2018-12-01]http//www.fipa.org/specs/fipa00061/ SC00061G.html. [13] Foundation for Intelligent Physical Agents (FIPA). FIPA agent management specification [EB/OL]. (2004-03-18) [2018-12-01]http://www.fipa.org/specs/fipaO0023/ SCO0023K.pdf. [14] MOURATIDIS H, GIORGINI P, MANSON G. Modelling secure multiagent systems[C]//Proceedings of the 2nd International Joint Conference on Autonomous Agents and Multiagent Systems. Melbourne, Australia, 2003: 859–866. [15] BRESCIANI P, GIORGINI P, MOURATIDIS H, et al. Multi-agent systems and security requirements analysis[M]// LUCENA C, GARCIA A, ROMANOVSKY A, et al. Software Engineering for Multi-Agent Systems II. Berlin, Germany: Springer, 2004: 35–48. [16] POSLAD S, CHARLTON P, CALISTI M. Specifying standard security mechanisms in multi-agent systems[M]// [17] 第 3 期 丁俐夫，等：多智能体系统安全性问题及防御机制综述 ·431·