基于蚁群粒子群算法的车辆自组织网络实际性移动模型构建

As the research foundation for protocol design and performance improvement in VANET (Vehicular Ad Hoc Networks), mobility model has absorbed many interests from the world now and past. However, the realistic characteristics of mobility model now have been overlooked by a great number of research works in which severe deviations have been detected between results from simulation and practice during the process of protocol design and performance optimization. Our project, to guarantee the truthfulness, will firstly introduce into our model the macro/micro mobility characteristics of VANET and swarm mobility properties based on a typical swarm intelligence algorithm,i.e. Particle Swarm Optimization(PSO) algorithm; to ensure the reasonability, will secondly combine our model with map information and guide vehicles by properly setting pheromones on map,which is inspired by Ant Colony Optimization(ACO) algorithm; to increase scalability, will finally add sensing and learning ability to our model. Based on our proposed realistic mobility model in VANET, the correctness of up-layer protocol design and engineering plan could be enhanced. Meanwhile, the outputs of performance evaluation depending on our research results could also be convincible. In general, by our works, the intelligence of transportation systems could be strengthened and the transforming progress from theoretical research to practical application of ITS (Intelligent Transportation System) will be greatly shortened.

移动模型是车辆自组织网络协议设计与性能优化的基础，然而，目前大部分研究采用的移动模型都忽视了模型与现实的相符程度，即实际性，导致协议设计或结果分析出现严重偏差。本申请，首先以群智能中粒子群算法为基础构建实际性移动模型基本框架，引入车辆自组织网络的宏微观移动特征和生物群体移动特征，以保证模型的真实性；随后将模型与地图信息相结合，并借鉴蚁群算法信息素思想，在地图中设置生物信息素诱导车辆，以保证模型的合理性；最后将感知学习能力赋予模型，为模型增加可扩展性。基于所提出的车辆自组织网络实际性移动模型，可提高以移动模型为基础的上层协议设计与工程规划的准确性，增强性能评估结果的可信性，为实现交通系统的智能化奠定基础，加速智能交通系统从理论研究向实际应用转化的过程。

移动模型是车辆自组织网络协议设计与性能优化的基础，然而，目前大部分研究采用的移动模型都忽视了模型与现实的相符程度，导致协议设计或结果分析出现严重偏差。本项目研究了基于群智能构建实际性移动模型基本框架的方法，通过引入车辆自组织网络的宏微观移动特征和生物群体移动特征，保证了移动模型的真实性。此外，项目还将基本框架与地图信息相结合并赋予模型感知学习能力，保证了模型的合理性和可扩展性。项目在车辆自组织网络实际性移动模型研究方面，形成了较为丰硕的成果，共产生SCI检索论文9篇，EI检索论文3篇，申请专利17项，授权4项，并获得陕西省科技进步奖二等一项。项目所提出的车辆自组织网络实际性移动模型，可提高以移动模型为基础的上层协议设计与工程规划的准确性，增强性能评估结果的可信性，为实现交通系统的智能化奠定基础，加速智能交通系统从理论研究向实际应用转化的过程。