Simulation Research On A Command Information System Based On Complex Network Theory

With the rapid development of information technology in twenty-first Century,the war form and operational theory also changed radically.Complexity and networking become the unavoidable problem of command information system.Network centric warfare is the trend of future war.It has become an indisputable fact that the command information system network(i.e.command information network)is facing unprecedented challenges.The rapid development of network science provides an alternative path to solve the problems faced by the command information network.Therefore,it is necessary to carry out the research of the command information system based on the complex network theory to adapt to the future information war.Based on the full investigation of the complex network theory applying in command information system,this disertation summarizes the shortcomings of the existing research and the problems to be solved.This dissertation studied thoroughly and systematically the modeling,analysis,optimization and link prediction,the main results and contributions of this dissertation are as follows.1)Modeling and analysis of command information system based on complex networkOn the basis of OODA(Observe,Orient,Decide,Act)combat ring theory proposed by John Boyd,summarized the existing problems of the command and control process model.A cooperative model based on the theory of combat loop is proposed to describe the whole process of operational action.A modeling method for complex network of command information system is proposed as well.Finally,the formal description framework and its evaluation indicators are applied to quantitatively analyze and verify the effectiveness of centralized,decentralized and distributed command information system.The simulation results show that the centralized structural model is the most reasonable choice when the system scale is small.The distributed network structure model is the best choice When the system scale is large.The most reasonable way to optimize the traditional tree structure model,which is the separated model,is to realize networking instead of directly eliminating it,and redeploy distributed system.2)Structure efficiency analysis of command information system based on structural ring meta-modelBased on TOPCFSR model,which overcomes the shortcomings of the combat model,a command information system is decomposed into six structural meta models,namely,cooperative detection ring,intelligence sharing ring,cooperative decision ring,accusation ring,synergetic strike ring and comprehensive ring.Then,a comprehensive evaluation index based on structural loop element model and its efficiency is proposed,and the six meta-models are evaluated respectively.In order to examine the relationship between network topology and effectiveness of the operational system and the basic principles for the construction of operational systems,the quantitative evaluation is made on traditional tree hierarchy and their evolutionary model by the six meta-models,providing effective tools for the evaluation of combat systems for Network Centric Warfare.3)Invulnerability analysis of command information system based on structural ring meta-modelFirst of all,problems of existing measure of invulnerability in complex network topologies are summarized,and the above analysis is verified by simulation.The structural ring meta-model represents the specific function of the combat system,so invulnerability of the command information system can be transformed into the damage degree of structure ring meta-model after being attacked.Then,this disertation proposed a metrics of netwotk survivability based on structural ring meta-model and its efficiency.Based on the existing literatures,this disertation proposed an incomplete information attack model which includes three elements of attack information,attack mode and attack strength.Finally,different attack strategies are used to evaluate and verify the combat system’s destruction resistance.4)Invulnerability optimization of command information system based on improved artificial bee colony algorithmFirst of all,combinatorial optimization of command information system is proposed based on structural ring meta-model.Then,an improved artificial bee colony algorithm based on low difference sequence and inheritance mechanism is proposed,which is used to optimize the destruction resistance of the command information system.A combinatorial optimization model for invulnerability of command information system is established,in which the structural ring meta-model is the objective function and the number of edges is the constraint condition,and the effectiveness of the above optimization method is verified by simulation experiments.5)Link prediction of command information system based on weighted naive Bayesian methodIn view of the shortcomings of the current link prediction algorithm,a link prediction algorithm using the historical state information is proposed under the naive Bayesian framework.Firstly,aiming at the finiteness and unilateralism of instantaneous network snapshot,it is difficult to predict the dynamic evolution of severe network links,and make full use of historical state information to improve the prediction accuracy.Secondly,we take the density of network as the correlation coefficient to measure the relationship among the common neighbor nodes,and measure the different contributions of different nodes with the rate of change,so as to improve the prediction resolution.Finally,the link prediction model based on the weighted naive Bayes is proposed,taking the network density and the degree of change as the weight and combining the information of the historical state.