Research On The Key Techniques Of Unmanned Operation Network Based On Connectivity

Unmanned Operation Network(UON)is a comprehensive network including sensor network,Command and Control(C2)network,shooter network and communication network.UON contains the characteristics of traditional wireless sensor network and classical Ad Hoc network,and responses to the demand such as combat reconnaissance,search,collaborative detection and processing,task allocation,cooperative attacking and so on.As the necessary unit in modern multidimensional wars crossing land,sea and sky,UAV is also the key node in UON.Each UAV in the network can be taken as the terminal node and cluster or routing node,which can detect the targets information and environments information as a sensor,or can store and forward data packets from its neighbors as a communicator,or finish the task allocation and routing selection as a decider.With the high performances of UAVs,such as quickly networking,rich roles and good reliability,more and more UAVs have applied in current combat actions and other disaster recovery and geological exploration.Following the traditional rules in information detection fusion and routing services,the traditional sensor network will encounter some problems when they directly apply in UON.The research of this dissertation focuses on the network model of UON,the cooperative information fusion of sensor network,dynamic cluster,collaborative allocation,opportunistic routing in UON,and the main contents summarize as the following respects.1.Based on the famous Jeffery network circle in Network Centric Warfare(NCW),we firstly set the node SDI combining sensor,C2,shooter and communicator in UON.Secondly,we propose the basic unit of UON,which is named as UON Jeffery circle.Thirdly,we put forward the basic structure,composite structure and hierarchical architecture of UON driven by tasks.Finally,based on graph theory,we define UON and its vertexes,edges and correlation functions mathematically.2.We analyze the cooperative information fusion of Sensor Network in UON.Firstly,we present the general definition and model of Fusion System of Sensor Network(FS-SN).Secondly,we analyze FS-SN from two points of view.One is based on the distributed parallel detection fusion and the other is based on the centraliazed fusion with connected probatility of links.From the former aspect,we analyze the fusion rules with local optimal and global optimal hard decision.Then we propose a Binary Direct Fusion Rule(BDFR).This rule can be applied with unknown prior probabilityan and unknown distribution function of noise.The simulation results show that BDFR can enhance the fusion performance and improve the detection probability with lower fusion cost.From the latter aspect,we have given Capacity Fusion Rule(C-FR)and Maximal-ratio Combining Fusion Rule(MRC-FR)and further discuss their performances in outage probability and outage capacity.3.We discuss the cooperative tasks processing and tasks allocation of UON.Firstly,we theoretically analyze the basic models of cluster and dynamic cluster.Secondly we propose an improved model of cooperative dynamic cluster and a method of task allocation based on genetic algorithm for UON.At the same time,the network connectivity and task priorities are considered and discussed as the main elements of the proposed model.Simulation results demonstrate the efficiency of our proposed cooperative model.Through collaboration,the advantage of each node in the network attacking each target can be obtained.And the optimal task allocation scheme is also obtained with higher profit by basic genetic algorithm.4.We study the opportunistic routing of UON.We propose an Improved Forwarding Opportunistic Routing(IFOR).The selection process of forwarding candidates set(a candidate set of forwarding nodes)and the establishment of forwarding node in node forwarding mechanism are discussed in IFOR.Experiments and analysis show that,compared with those opportunistic routing protocols which are based on the traditional node forwarding mechanism,IFOR can reduce protocol computing overhead and raise end-to-end packet delivery ratio in data transmissions.For UON,the simulation illustrates the relationship between node speed and the average end-to-end delay,average reachable nodes and average node degree.Otherwise,the simulation discusses the survivability of IFOR when it is attacked by malicious nodes.