【作者】 周逊；

【导师】 曾华燊；

【作者基本信息】 西南交通大学 ， 交通信息工程与控制， 2007， 博士

【摘要】 近代战争的经验表明：信息对赢得局部战场的胜利或整个战争至关重要，因此现代战争被军事学家视为“以信息为中心的战争”。由于网络在信息交换、处理与存储上的重要性，现代战争也被视为“以网络为中心的战争”(简称“网络中心战”)。在过去的20多年中，无人驾驶飞行系统(UAS)由于用途广、制造成本低、无驾驶员及人员伤亡，受到军方的高度重视，在未来的战争中将扮演了越来越重要的角色。这就是本论文选择“与无人驾驶飞机作战网络相关技术”作为研究对象的动因。本论文的创新性贡献可以总结为两个方面：①提出了3个与无人机群作战网络相关的三个不同层次的概念设计，即GDSN、UAS-COM和DRNA：②提出并初步验证了适合未来无人机作战网络应用环境的三个有特色的路径算法。根据未来“网络中心战”的动态特性，本文将全局作战网络抽象为“全局可动态重组战略网”(GDSN-Global Dynamically-reconfigurable Strategic Network)。与飞行器相关的GDSN由总部系统(HS)、通信中继系统(CRS)以及成群的有人与无人驾驶系统战术网络(MAS／UAS-TN)动态地构建而成，以满足战略和战术上的需求。在GDSN框架的基础上，本文进一步探讨了在无人机网络模式下机间或网间的协同作战模式(Coordinated Operation Mode)并简要地介绍了笔者运用该模式在微波协同干扰系统中的具体应用。根据无人机小型化发展趋势、高速移动特点、无线通信条件和高安全性的需求，本文基于未来无人机群战术作战网络的新模式提出了无人机“动态可重定义网络体系结构”(DRNA-Dynamically Re-defi nable Network Architecture)的初步框架。第3章到第6章是本论文的主体，其重点是无人机作战网络路由协议与算法。无人机作战网络的通信条件和工作特点决定了该网络是一种特殊的无线移动自组网络(MANET-Mobile Ad-hoc NET)，因此，第3章较全面地分析和介绍了MANET典型路由算法，分析了无人机群作战网络的特殊需求，为后续章节提出新的路由算法做准备。第4章提出了一种“基于源路由和多路径OLSR协议”(SRMP-OLSR)。仿真结果表明SRMP-OLSR特别在重负载的环境下吞吐量得到提高，丢包率和传输延迟可以减少。第5章提出“负载度。”的概念并进一步改进了多路径策略。即根据低层的近期传输特性自适应性地为各条路径设置相应的“负载度”，并以此来平衡网络的负载和改善其传输能力。该算法被称为“基于负载度的自适应多路径选径”(LAMP-DSR)。第6章介绍了基于移动代理的混合多路径路由算法(MAH-MPR)。该算法结合移动代理的概念，综合了蚁群优化算法(ACO)、先验路由算法和按需路由算法的优点。仿真表明MAH-MPR协议是比AODV(Ad-hoc On-demand Distance Vector)协议更好的路由策略。第7章在对全文进行总结的基础上，对今后有待研究工作的将重点进行了分析与展望。更多还原

【Abstract】 Past experiences in wars have proved that information is vital for winning abattle or a war in modern times; therefore, modern war has been referred to as"information-centric". For the importance of networks in supporting informationexchange and processing, modern war is also referred to as beingnetwork-centric. In last two decades, Unmanned Aerial Vehicles/Systems(UAV/UAS) have drawn much attention from military circle and played moreand more important roles in the battle fields for their versatile functionality andlow-costs (in manufacture, pilot training and life-saving). This is exactly theincentive of the research work presented in this thesis.The main innovative contributions of the thesis can be summarized as: a)introduced three new concepts relevant to UAS networks, including GDSN,COM, and DRNA, which cover from the military global network, through UAStactic networks to cooperated operation mode; and b) introduced three newrouting algorithms suitable for future UAS operation.According to dynamic nature of future "network-centric war", this thesismodels the military network as a Global Dynamically-reconfigurable StrategicNetwork (GDSN). To meet strategic and tactic requirements, a GDSN relevantto aerial systems consists of a Headquarter System (HS), a CommunicationRelay System (CRS), and multiple Manned/Unmanned Aerial System TacticNetworks (MAS/UAS-TN). To further take the advantage of GDSN in jointaction among a group of UAS, a Cooperated Operation Mode (UAS-COM) isdiscussed and backed up by an application to the project called "MicrowaveInterference Network". In view that the UAS is getting smaller and smaller andmoves faster than other vehicles, and works in an unsatisfactory and hostilewireless communication environment, this dissertation introduced a frameworkof a novel network architecture called DRNA (Dynamically Re-definableNetwork Architecture).Chapters 3 through 6 are the main parts of the thesis, which address torouting algorithms at the network layer in UAS-TN although other layertechniques are more or less involved. Chapter 3 presents a comprehensive study and analysis to existing routing algorithms in MANET, which provides abackground for the research work and leads to focusing on multi-path routing.Chapter 4 combines the most popular OLSR (Optimal Link State Routing)protocol in MANET with source routing and leads to a new routing protocolcalled SRMP-OLSR (Source Routing based Multi-Path OLSR). Simulationresults of SRMP-OLSR have shown that with this protocol, throughput can beimproved especially in heavily loaded conditions, and data loss rate and meantransit delay can be reduced.Chapter 5 further refines multi-path strategy by introduction a new conceptcalled Load Degree. By adaptively assigning Load Degree to individual paths inMulti-path routing according to transmission performance at lower layers,load-balance can be maintained and overall utility of transmission capability canbe improved. This algorithm is called Load-based Adaptive Multi-PathDynamical Source Routing (LAMP-DSR).Chapter 6 introduces a new routing protocol called MAH-MPR (MobileAgent based Hybrid Multi-Path Routing), which combines the Mobile Agent(MA) concept with ACO (Ant Colony Optimization) algorithm and takesadvantages of pre-active routing and on-demand routing. Simulation resultshave shown that the performance of MAH-MPR is superior over that of AODV(Ad-hoc On-demand Distance Vector) routing strategy.Chapter 7 concludes the thesis and suggests potential research works in thefuture.更多还原