基于中间件分布式指控系统的设计与实现
|
摘要
在现代战争中,如何能够迅速、准确地获取瞬息万变的战场信息,已经成为决定战场胜负的重要因素。随着计算机和网络技术的迅速发展,作战指挥己从传统的人工指挥方式向以计算机信息处理为核心的指挥自动化方式的转变。随着信息技术科学的进步,特别是计算机技术和通信技术的飞速发展,由两者的融合而产生的计算机网络技术及其应用为新形式指控系统的研究和应用提供了强有力的支持。基于计算机网络环境下采用分布式的协同作业手段,在高易损性环境下具备动态重组特性的指控系统是符合未来战争实际需要的。本文以搭建符合作战领域业务需要的分布式指控系统为需求,对失效检测、负载监测和动态迁移等与指控系统密切相关的关键技术进行了研究,将研究成果引入指控系统领域,形成与业务需求相结合的分布式策略,以面向对象的思想设计了指控系统的框架,并给出了相应的实现。
分布式指控系统的设计思想主要是对系统进行横向分解,划分系统为表示层、实体模型层、业务逻辑控制层、分布事务服务层、网络通信层。各层之间通过简单接口进行通信,做到层之间的单向依赖。将各层次之间的耦合度降到最低,隔离各层的内部实现,使层次内部的算法和策略替换不会影响到其它层次。在设计时以MVC框架为基础搭建系统的上层部分,将业务实体和界面呈现部分分别作为模型和视图,以控制器作为两者之间的联系;更新模型和通知视图更新。在系统的下层支撑部分,以网络通信层作为分布事务处理服务层的底层。这样,下层进行分布事务处理,以动态重组,任务迁移等手段保证指控系统的全功能运行,而上层的MVC框架,在底层业务的支撑下可以相对独立地工作。
综上所述,论文的主要工作是研究如何搭建符合作战领域业务需要的分布式指控系统。首先对相关分布式系统理论进行了研究和分析,重点研究了分布式思想及CORBA、设计模式等几方面与指控系统密切相关的关键技术。然后根据业务需求,选择和改进了多种分布式事务处理策略,以满足分布式指控系统的功能要求。在以上工作的基础上,设计了多服务端业务协作、可动态功能重组特性的分布式指控系统框架。最后给出了符合面向对象思想,具有良好扩展性的系统部分实现。
It has become an important issue about how to obtain the information in volatile battlefield quickly and accurately. With the rapid development of computer and networks technology, battle commanding is experiencing a change from traditional fashion of which human’s command is central to a new one of which auto-commanding system based on information processing by computers is vital. Along with the progress of information technology, especially the computer technology and communication technology develop rapidly, the computer network technology which is the result of the two kinds of technology provides strong support for the research and application of new type of command and control system. Based on the introduction of distributed collaboration approach in computer network environment, the command and control system which has the feature of dynamic reconfiguration in highly-vulnerability environment satisfies the practical requirements of future wars. In this paper, consider the construction of distributed command and control system that meets the need of requirement in battle field as the direction; analyze the key technology that relates the commanding system closely such as the invalidation verification, load monitoring and dynamic transfer. Introduce the result into the commanding system and form the distributed strategy that combine with the business requirement, design the commanding system framework in object-oriented paradigm, and give the implementation.
The thought of the distributed system is to horizontally decompose the system into the presentation layer, the entity model layer, the business logic layer, the distributed transaction layer, the network layer. The communication between layers use the simple interface, do the dependency in one direction. Lower the coupling between layers extremely, separate the implementation between layers and make the algorithm and strategy in the layer not influence other layers. The MVC pattern is introduce to build the upper part of the system on design level. The business entity and the graphical interface is as the model and view, the controller is the relation between them that update the model and notify the update of view. On the lower part of the system, the network communication layer is as the distributed transaction handler.
To sum up, this paper is mainly on the research of how to build the distributed command and control system meeting the requirement of the battle field. First, research and analyze of distribution related theory, the point is the key technology that relates the command and control system such as Distributed thinking and CORBA, design pattern. Second, based on the business requirement, multiple distributed transaction strategy is improved to satisfy the function requirement of the command and control system, and on the basis of the result, multi-server business collaboration and the distributed commanding system framework are designed. Finally, the implementation of the system is originated which conforms to the Object-oriented paradigm and is easily extensible.
分布式指控系统的设计思想主要是对系统进行横向分解,划分系统为表示层、实体模型层、业务逻辑控制层、分布事务服务层、网络通信层。各层之间通过简单接口进行通信,做到层之间的单向依赖。将各层次之间的耦合度降到最低,隔离各层的内部实现,使层次内部的算法和策略替换不会影响到其它层次。在设计时以MVC框架为基础搭建系统的上层部分,将业务实体和界面呈现部分分别作为模型和视图,以控制器作为两者之间的联系;更新模型和通知视图更新。在系统的下层支撑部分,以网络通信层作为分布事务处理服务层的底层。这样,下层进行分布事务处理,以动态重组,任务迁移等手段保证指控系统的全功能运行,而上层的MVC框架,在底层业务的支撑下可以相对独立地工作。
综上所述,论文的主要工作是研究如何搭建符合作战领域业务需要的分布式指控系统。首先对相关分布式系统理论进行了研究和分析,重点研究了分布式思想及CORBA、设计模式等几方面与指控系统密切相关的关键技术。然后根据业务需求,选择和改进了多种分布式事务处理策略,以满足分布式指控系统的功能要求。在以上工作的基础上,设计了多服务端业务协作、可动态功能重组特性的分布式指控系统框架。最后给出了符合面向对象思想,具有良好扩展性的系统部分实现。
It has become an important issue about how to obtain the information in volatile battlefield quickly and accurately. With the rapid development of computer and networks technology, battle commanding is experiencing a change from traditional fashion of which human’s command is central to a new one of which auto-commanding system based on information processing by computers is vital. Along with the progress of information technology, especially the computer technology and communication technology develop rapidly, the computer network technology which is the result of the two kinds of technology provides strong support for the research and application of new type of command and control system. Based on the introduction of distributed collaboration approach in computer network environment, the command and control system which has the feature of dynamic reconfiguration in highly-vulnerability environment satisfies the practical requirements of future wars. In this paper, consider the construction of distributed command and control system that meets the need of requirement in battle field as the direction; analyze the key technology that relates the commanding system closely such as the invalidation verification, load monitoring and dynamic transfer. Introduce the result into the commanding system and form the distributed strategy that combine with the business requirement, design the commanding system framework in object-oriented paradigm, and give the implementation.
The thought of the distributed system is to horizontally decompose the system into the presentation layer, the entity model layer, the business logic layer, the distributed transaction layer, the network layer. The communication between layers use the simple interface, do the dependency in one direction. Lower the coupling between layers extremely, separate the implementation between layers and make the algorithm and strategy in the layer not influence other layers. The MVC pattern is introduce to build the upper part of the system on design level. The business entity and the graphical interface is as the model and view, the controller is the relation between them that update the model and notify the update of view. On the lower part of the system, the network communication layer is as the distributed transaction handler.
To sum up, this paper is mainly on the research of how to build the distributed command and control system meeting the requirement of the battle field. First, research and analyze of distribution related theory, the point is the key technology that relates the command and control system such as Distributed thinking and CORBA, design pattern. Second, based on the business requirement, multiple distributed transaction strategy is improved to satisfy the function requirement of the command and control system, and on the basis of the result, multi-server business collaboration and the distributed commanding system framework are designed. Finally, the implementation of the system is originated which conforms to the Object-oriented paradigm and is easily extensible.
引文
[1]苏建志等指挥自动化系统,北京:国防工业出版社,1999,520-522
[2]实现作战体系一体化,北京:解放军报,2004
[3] T.D.Chandra and S.Toueg. Unreliable failure detectors for reliable distributed system.J.ACM.1996,43(2):225-267
[4] W.Chen,S.Toueg,and M.K. Aguilera.On the quality of service of failure detectors. IEEE Transactions on Computers. 2002.51(5):561-580
[5] M.Bertier,O. Marin, P.Verissimo. Using tailored failure suspectors to support distributed cooperative applications.InProc. 7th IASTED/ISMM Intl.Conf.on Parallel and Distributed Computing and Systems. Washington,D.C. USA. 1995.10:352-356
[6] C.Fetzer,M.Raynal.F.Tronel. An adaptive failure detection protocol,In Proc. Of the 8th IEEE Pacific Rim Symp. On Dependable Computin(PRDC8).2001
[7] Albert Y. Observations on using genetic algorithms for dynamic loadbalancing.IEEE Transactions on parallel and distributed Systems,2001,12(9):899-911
[8] Kunz T. The Influence of different workload descriptors on a heuristic load balancing scheme. IEEE Transactions on sofeware engineering,1991,17(7):725-730
[9] H.G. ROTITHOR,Taxonomy of dynamic task scheduling schemes in distributed computing system. IEEE ProceedingsComputers and Digital Techniques,1994,141(1).
[10] CHOU,T.C.K. and ABRAHAM, J.A.: Load balancing in distributed systems, IEEE Trans. July 1982,SE8,(4):401-412
[11] Nancy A.Lynch.Distributed Algorithms.机械工业出版社
[12] George Coulouris,Jean Dollimore,Time Kinderg. Distributed Concepts and Design. Third Edition
[13]徐敏,同构型分布式计算机的启发式任务分配算法,计算机学报[J],1998.17(2)
[14]何炎祥,刘朝阳,邓爱林,基于工作站的分布式系统中的负载分配和平衡策略,计算机工程,1999.25(11):12-14
[15]毛国君等,分布式系统中的双向启动自适应任务分配算法,计算机学报,1996.19(7).
[16]李敏勇,张建昌,新指挥控制原理,情报指挥控制系统与仿真技术,2004.26(1).
[17]郭传福,夏惠诚,分布式作战系统集成方案构想,舰船电子工程,2000.1:17-20.
[18]杨凡等,分布式指挥控制系统协同机制研究,系统工程理论与实践,2004.11(11):128-135.
[19]闵绍荣,黎洪生,未来舰艇作战系统总体设计模式研究,舰船电子工程,2004,1(15):58.
[20] Robert Martin著:《敏捷软件开发:原则、模式与实践》,清华大学出版社,2003年9月
[21] Ralph E. Johnson and Brian Foote, Designing Reusable Classes, Journal of ObjectOriented Programming, June/july 1998,1(2), 22-35
[22] Feferman S, Transfinite Recursive Progressions of Axiomatic Theories, Journal of Symbolic
[23] Jacobson, Ivar, et al. ObjectOriented Software Engineering. Reading, MA: AddsionWeslcy, 1992
[24] Meyer, Bertrand. ObjectOriented Software Construction, UpperSaddle River, NJ:Prentice Hall, 1997
[25] EricGamma等著,李英军等译,《设计模式》(DesignPatterns),机械工业出版社,2000年9月.
[26] Eric Grama, Richard Helm, Ralph Johnson, John Vlissides, Design Patterns: Elements of Reusable ObjectOriented Software
[27]侯捷,《深入浅出MFC》,武汉:华中师范大学出版社,2002年9月。
[28]谷和启,中间件技术及其应用.http://resource.51cmm.com/NewTech/NoO85.httn
[29]施晓军,解读COM与CORBA.http://info.tlm.cn/4520.htm
[30] CORBA中间件Inforbus. http://www.inforbus.com/product_info.jsp?info_id=13
[2]实现作战体系一体化,北京:解放军报,2004
[3] T.D.Chandra and S.Toueg. Unreliable failure detectors for reliable distributed system.J.ACM.1996,43(2):225-267
[4] W.Chen,S.Toueg,and M.K. Aguilera.On the quality of service of failure detectors. IEEE Transactions on Computers. 2002.51(5):561-580
[5] M.Bertier,O. Marin, P.Verissimo. Using tailored failure suspectors to support distributed cooperative applications.InProc. 7th IASTED/ISMM Intl.Conf.on Parallel and Distributed Computing and Systems. Washington,D.C. USA. 1995.10:352-356
[6] C.Fetzer,M.Raynal.F.Tronel. An adaptive failure detection protocol,In Proc. Of the 8th IEEE Pacific Rim Symp. On Dependable Computin(PRDC8).2001
[7] Albert Y. Observations on using genetic algorithms for dynamic loadbalancing.IEEE Transactions on parallel and distributed Systems,2001,12(9):899-911
[8] Kunz T. The Influence of different workload descriptors on a heuristic load balancing scheme. IEEE Transactions on sofeware engineering,1991,17(7):725-730
[9] H.G. ROTITHOR,Taxonomy of dynamic task scheduling schemes in distributed computing system. IEEE ProceedingsComputers and Digital Techniques,1994,141(1).
[10] CHOU,T.C.K. and ABRAHAM, J.A.: Load balancing in distributed systems, IEEE Trans. July 1982,SE8,(4):401-412
[11] Nancy A.Lynch.Distributed Algorithms.机械工业出版社
[12] George Coulouris,Jean Dollimore,Time Kinderg. Distributed Concepts and Design. Third Edition
[13]徐敏,同构型分布式计算机的启发式任务分配算法,计算机学报[J],1998.17(2)
[14]何炎祥,刘朝阳,邓爱林,基于工作站的分布式系统中的负载分配和平衡策略,计算机工程,1999.25(11):12-14
[15]毛国君等,分布式系统中的双向启动自适应任务分配算法,计算机学报,1996.19(7).
[16]李敏勇,张建昌,新指挥控制原理,情报指挥控制系统与仿真技术,2004.26(1).
[17]郭传福,夏惠诚,分布式作战系统集成方案构想,舰船电子工程,2000.1:17-20.
[18]杨凡等,分布式指挥控制系统协同机制研究,系统工程理论与实践,2004.11(11):128-135.
[19]闵绍荣,黎洪生,未来舰艇作战系统总体设计模式研究,舰船电子工程,2004,1(15):58.
[20] Robert Martin著:《敏捷软件开发:原则、模式与实践》,清华大学出版社,2003年9月
[21] Ralph E. Johnson and Brian Foote, Designing Reusable Classes, Journal of ObjectOriented Programming, June/july 1998,1(2), 22-35
[22] Feferman S, Transfinite Recursive Progressions of Axiomatic Theories, Journal of Symbolic
[23] Jacobson, Ivar, et al. ObjectOriented Software Engineering. Reading, MA: AddsionWeslcy, 1992
[24] Meyer, Bertrand. ObjectOriented Software Construction, UpperSaddle River, NJ:Prentice Hall, 1997
[25] EricGamma等著,李英军等译,《设计模式》(DesignPatterns),机械工业出版社,2000年9月.
[26] Eric Grama, Richard Helm, Ralph Johnson, John Vlissides, Design Patterns: Elements of Reusable ObjectOriented Software
[27]侯捷,《深入浅出MFC》,武汉:华中师范大学出版社,2002年9月。
[28]谷和启,中间件技术及其应用.http://resource.51cmm.com/NewTech/NoO85.httn
[29]施晓军,解读COM与CORBA.http://info.tlm.cn/4520.htm
[30] CORBA中间件Inforbus. http://www.inforbus.com/product_info.jsp?info_id=13