基于复杂网络理论的车载自组织网络研究
发布时间：2018-09-10 12:22
【摘要】：随着城市化进程的加快,道路交通状况越来越复杂,车载自组织网络作为智能交通系统的基础应用将是缓解交通拥堵、快速处理交通事故、提高交通安全、优化城市汽车流量及实现无人驾驶的重要方式。车载自组织网络的通讯方式借助于比较成熟的通信网络,属于多跳无线通信移动自组织网络范畴,网络部署灵活,具有较好的发展前景。传统研究专注于通信协议及其拓扑,未从宏观角度揭示其网络本质。本文基于复杂网络理论提出车载自组织网络模型,以新视角研究车载自组织网络的拓扑结构、网络本质和控制方式,较好地克服了从单一角度进行特性研究的不足。 本文首先介绍了车载自组织网络体系架构和模型算法,借用复杂网络理论思想建立了V2V子网模型、V2I子网模型及整体车载自组织网络模型,采用平均场理论进行瞬时度分布和稳态度分布分析；根据车载自组织网络节点特性及假设条件,考虑到实际的车载自组织网络并不是无限增长的,研究了具有寿命周期的网络模型和模块模型构造。其次讨论了模型的特殊情况,判定度分布指数范围,以度分布满足幂律分布规律验证车载自组织网络的无标度特征,以基尼系数验证网络的异质性；从微观角度研究了车载自组织网络信息流量分布特性和考虑容量限制的信息流量分布特性。最后以网络状态方程为基础,从改善已有网络性能和提高网络效率、稳定性的角度出发,导出被牵制控制到平衡点的网络状态方程,计算矩阵最大特征值和网络特征值,提出车载自组织网络牵制控制策略,探讨了网络的拓扑结构参数对网络同步动力学的影响。阐述了网络能控性的基本概念,从度相关性角度对能控性的影响进行了分析,通过数学实验和仿真实验分析了特定牵制控制、随机牵制控制和混合牵制控制策略的能控性。 研究结果表明车载自组织网络符合幂律分布规律,满足无标度网络特征,度分布指数2γ≤3,度值大的节点对网络影响较大,传播信息更快,信息流量与度值之间满足幂律分布规律；在考虑节点寿命周期的情况下,网络度分布不符合幂律分布规律。度分布指数越小,同步越难实现；择优连接网络比重连网络的同步更强,动态网络对随机攻击表现出更强的鲁棒性。对比不同的控制策略,特定牵制控制策略效率最高。在同配网络中,特定牵制控制节点比例的增加对整个网络的同步作用影响越来越小；在车载自组织异配网络中,特定牵制度值大的节点其控制效果易于传播到整个网络,使得网络易于控制。
[Abstract]:With the acceleration of urbanization, the road traffic situation is becoming more and more complex. As the basic application of Intelligent Transportation system (its), the vehicle-borne self-organizing network will alleviate traffic congestion, deal with traffic accidents quickly and improve traffic safety. Optimization of urban vehicle flow and realization of driverless important ways. With the help of more mature communication networks, vehicle-mounted ad hoc networks belong to the category of multi-hop wireless mobile ad hoc networks. The network deployment is flexible and has a good development prospect. Traditional research focuses on communication protocols and their topologies, and fails to reveal their network nature from a macro point of view. Based on the complex network theory, this paper presents a vehicle Ad Hoc network model, and studies the topology, nature and control mode of the vehicle ad hoc network from a new perspective, which overcomes the shortcomings of the research on the characteristics from a single point of view. In this paper, the architecture and algorithm of vehicle Ad Hoc network are introduced, and the V2V subnet model and the whole vehicle Ad hoc network model are established by using the complex network theory. The instantaneous degree distribution and steady attitude distribution are analyzed by using the mean field theory. According to the node characteristics and assumptions of the vehicle-mounted ad hoc network, considering that the actual vehicle-mounted ad hoc network is not infinitely growing, The construction of network model and module model with life cycle is studied. Secondly, the special case of the model is discussed, the exponential range of degree distribution is determined, the scale-free characteristic of the vehicle self-organized network is verified by the degree distribution satisfying the law of power-law distribution, and the heterogeneity of the network is verified by the Gini coefficient. In this paper, the characteristics of information traffic distribution in vehicle-mounted ad hoc networks and the information traffic distribution characteristics considering capacity constraints are studied from the microscopic point of view. Finally, based on the network state equation, from the point of view of improving the network performance, improving the network efficiency and stability, the network state equation which is controlled to the equilibrium point is derived, and the maximum eigenvalue and the network eigenvalue of the matrix are calculated. A containment control strategy for vehicle-borne ad hoc networks is proposed, and the influence of network topology parameters on network synchronization dynamics is discussed. This paper expounds the basic concept of network controllability, analyzes the influence of controllability from the angle of degree correlation, and analyzes the controllability of specific containment control, random control and hybrid control strategy through mathematical experiments and simulation experiments. The results show that the vehicle-mounted self-organizing network conforms to the law of power-law distribution and satisfies the characteristics of scale-free network. The nodes with large degree value and the index of degree distribution 2 纬 鈮，
本文编号：2234431
[Abstract]:With the acceleration of urbanization, the road traffic situation is becoming more and more complex. As the basic application of Intelligent Transportation system (its), the vehicle-borne self-organizing network will alleviate traffic congestion, deal with traffic accidents quickly and improve traffic safety. Optimization of urban vehicle flow and realization of driverless important ways. With the help of more mature communication networks, vehicle-mounted ad hoc networks belong to the category of multi-hop wireless mobile ad hoc networks. The network deployment is flexible and has a good development prospect. Traditional research focuses on communication protocols and their topologies, and fails to reveal their network nature from a macro point of view. Based on the complex network theory, this paper presents a vehicle Ad Hoc network model, and studies the topology, nature and control mode of the vehicle ad hoc network from a new perspective, which overcomes the shortcomings of the research on the characteristics from a single point of view. In this paper, the architecture and algorithm of vehicle Ad Hoc network are introduced, and the V2V subnet model and the whole vehicle Ad hoc network model are established by using the complex network theory. The instantaneous degree distribution and steady attitude distribution are analyzed by using the mean field theory. According to the node characteristics and assumptions of the vehicle-mounted ad hoc network, considering that the actual vehicle-mounted ad hoc network is not infinitely growing, The construction of network model and module model with life cycle is studied. Secondly, the special case of the model is discussed, the exponential range of degree distribution is determined, the scale-free characteristic of the vehicle self-organized network is verified by the degree distribution satisfying the law of power-law distribution, and the heterogeneity of the network is verified by the Gini coefficient. In this paper, the characteristics of information traffic distribution in vehicle-mounted ad hoc networks and the information traffic distribution characteristics considering capacity constraints are studied from the microscopic point of view. Finally, based on the network state equation, from the point of view of improving the network performance, improving the network efficiency and stability, the network state equation which is controlled to the equilibrium point is derived, and the maximum eigenvalue and the network eigenvalue of the matrix are calculated. A containment control strategy for vehicle-borne ad hoc networks is proposed, and the influence of network topology parameters on network synchronization dynamics is discussed. This paper expounds the basic concept of network controllability, analyzes the influence of controllability from the angle of degree correlation, and analyzes the controllability of specific containment control, random control and hybrid control strategy through mathematical experiments and simulation experiments. The results show that the vehicle-mounted self-organizing network conforms to the law of power-law distribution and satisfies the characteristics of scale-free network. The nodes with large degree value and the index of degree distribution 2 纬 鈮，
本文编号：2234431
本文链接：https://www.wllwen.com/kejilunwen/daoluqiaoliang/2234431.html
