利用复杂网络理論优化车载通信网络

In recent years, the increasing number of cars in urban areas inevitably causes more traffic jams and air pollution. Vehicular Ad-hoc Network (VANET), as being a key component of the Intelligent Transportation Systems (ITS) that allows wireless communications among vehicles, is expected to play a vital role in improving road safety and efficiency, as well as enabling ubiquitous communications in the future. The most critical technical challenge for VANET design is the highly unpredictable vehicle motion that leads to dynamic changes of network topology, and thus unstable wireless communication channels. However, there is a lack of fundamental understanding of the interactions among the road network, vehicular mobility, and communication connectivity. Complex network models provide a critical mathematical and theoretical framework for identifying networks with different properties, and fill the gap between road network and communication capability. The main objective of this proposed research is to exploit complex network theory to study the fundamental relationship between the bottom road network and the communication network above in VANETs. Through the development of mathematical and simulation models, node density and connectivity knowledge can be readily computed from empirical road network statistics. Cross-layer communication algorithms and network deployment schemes can then be designed judiciously according to the captured degrees of connectivity at different locations in the network. The research outcome will also provide insights into city planning, and lay a solid foundation for realizing a giant vehicular communication platform in the future that provides ubiquitous communication connectivity to an increasing and diverse number of devices and complex user requirements.

近年来，车辆的快速增加正为城市带来越来越多的交通堵塞和空气污染。作为智能运输系统的重要组成，车载通信网络(VANET)通过车辆间的无线通信不仅可改善道路交通安全及运输效率，更有可能成为“无所不在”通信的一种实现方式。VANET面临的一大挑战是车辆的高速移动使通信网络拓朴快速改变以及其导致的通信链路的不稳定性。然而，现有研究对于道路网络、车辆移动及通信连接度之间的关系并没有完备的理论诠释，VANET也仅止于支持非实时数据业务。复杂网络理论提供了一个能有效识别不同网络特性的数理平台，正适用于填补道路网络与通信性能之间的空白。本项目拟利用复杂网络的理论分析VANET中底层道路网络与上层通信网络之间的根本关系，建立数学及仿真模型以描述不同道路网络给通信连接度和VANET性能带来的变化，并对通信协议和网络部署进行跨协议层优化以最终提高VANET支持的数据速率和丰富其所能支持的业务种类。

数学图论中的图, 包括一组由边连接的节点, 是分析现实世界复杂系统中最简单却非常有效的数学系统之一。通过将复杂系统中的组件表示为节点, 并将它们的关联性表示为边, 可以协助我们研究复杂系统拓扑结构中的固有特性。在本项目中, 我们利用图论的概念对两种公共交通网络进行拓扑分析, 包括公交和地铁网络。我们首先对公交网络的拓扑作非交互式的单层分析。然后, 通过考虑多种公共交通网络之间的相互作用, 将公交和地铁网络拓扑结构作多层网络分析。拓扑分析揭示了在单层和多层分析下不同的网络行为。最后, 我们将图论与车联网的概念结合起来, 从宏观的角度探讨交通网络拓扑与车联网连接性的关系, 并提出了一些提高车联网通讯效率的协议和算法。总括而言, 把图论的概念应用于公共交通网络的分析, 能更有效地从区域与广域的角度了解交通网络的运作, 从而促进在未来的智能城市中部署各种操作设计和优化计划。此外, 从拓扑结构对车联网连接动态的影响表明, 网络工程师应考虑各种拓扑参数及其分布模式, 以开发可靠和现实的数据传播方法。