基于Web的物联网体系结构和感知域关键技术研究

发布时间：2018-01-21 23:32

本文关键词： 物联网感知域拓扑路由　出处：《北京邮电大学》2014年博士论文　论文类型：学位论文

【摘要】：物联网(The Internet of Things, IoT)是新一代信息技术的重要组成部分,物联网可以看作是物物相连的互联网,通过感知、识别技术,与普适计算等技术融合应用,形成了继计算机、互联网之后又一次信息产业发展的浪潮。我们在分析物联网系统各部分功能与特点的基础上,从基于Web的物联网业务环境的基本原则出发,将物联网系统架构分为感知域和业务域。无线传感器网络作为感知域中的重要内容,具有特别重要的地位。我们针对基于Web的物联网体系结构和感知域中的关键技术进行研究,取得的主要的研究成果及创新点如下： 1.当前国内外在物联网领域已经取得了大量理论研究成果和部分应用示范,但问题仍较为突出,比如封闭的内部尝试,缺乏开放性、示范性与可复制性；不能互联互通,存在严重的地区和行业壁垒,大量示范工程重复建设；产品、解决方案互不兼容,缺乏统一的概念,导致大量碎片化的框架和应用等。针对这些问题,在分析物联网系统各部分功能与特点的基础上,从基于Web的物联网业务环境的基本原则出发,可以将物联网系统架构分为感知域和业务域,提出了基于Web的物联网体系结构,将物联网Web化。构建基于Web的物联网系统服务平台,汇聚产业链上的设备和平台,引进国内外先进的技术和理念,形成物联网应用设备商店,为用户提供全方位的体验与服务,最终形成物联网应用服务云,构建物联网生态系统。 2.在无线传感器网络中,分簇是一种有效的拓扑控制方法,层次结构清晰,兼顾负载均衡,能增强网络可靠性,具有较强的扩展性与鲁棒性。目前很多分簇协议存在簇头分布不均匀,簇头节点的选取未充分考虑剩余能量与通信距离的影响,以及分簇情况不随网络状态的变化而变化等问题。针对这些问题,我们设计了一种基于虚拟坐标的动态自适应分簇(a Virtual Coordinates Dynamic Adaptive Clustering, VCDAC)协议,该协议按照虚拟分区的思想将网络分簇,并根据网络的状态动态决定分簇数量,从而降低了能耗。仿真结果显示该协议能明显延长网络的生存时间,均衡网络负载,发送更多的数据至汇聚节点,具有优良的性能。 3.无线传感器网络的拓扑控制对网络性能影响很大,优良的拓扑结构能提高路由协议和MAC协议的效率,为目标定位、时间同步与数据融合提供基础。分簇作为拓扑控制中一种层次清晰的方法,能延长网络的生存时间,经常被转化为优化问题来求最优解。传统分析优化方法需要大量的计算,计算量随问题规模成指数增长,对于微型传感器来说,少的计算量与资源需求,并能产生可接受结果的优化方法是适宜的,粒子群算法就是一种计算效率高的仿生智能优化方法。本文首先提出了一种无参数小生境粒子群优化算法,并验证其有效性,然后利用该算法,设计了一种基于小生境粒子群优化的自适应分簇(an Adaptive Clustering Protocol using Niching Particle Swarm Optimization, ACP-NPSO)协议,该协议形成的分簇分布均匀,均衡了系统的负载。仿真实验结果表明,该协议具有良好的效率,能明显延长网络的生存周期,提高系统的吞吐量。 4.无线传感器网络具有很强的应用相关性,不同应用中的路由协议差别很大,没有一个通用的路由协议,是一种与应用密切相关的网络。路由协议根据逻辑结构的不同,分为分簇型路由协议和平面型路由协议,与平面型路由协议相比较,分簇型路由协议结构清晰,能高效利用节点资源。本文设计了一种基于粒子群蚁群优化的动态分簇路由协议(a Dynamic Clustering Routing Protocol using Particle Swarm Optimization and Ant Colony Algorithm, DCRP-PSOACA),我们给出了DCRP-PSOACA的详细设计,仿真实验结果表明,该协议能有效减少节点能量消耗,提供稳健的路由路径,延长网络生存时间,实现高效通信。
[Abstract]:The Internet of things (The Internet of Things, IoT) is an important part of a new generation of information technology, the Internet of things can be regarded as material objects connected to the Internet, through perception, recognition, and ubiquitous computing integration application technology, following the formation of the computer, the Internet after another wave of development of information industry. We based on the analysis of the function of each part and the features of the IOT system, from Web based on the basic principles of networking business environment of the networking system architecture is divided into perception domain and business domain. The wireless sensor network as an important content of perception in the domain, has a very important status. According to our research on the key technology of network system the structure and sensing domain of Web based on the main research results and innovations are as follows:
1. things in the field at home and abroad has made a lot of theoretical research and application demonstration, but the problem is still more prominent, such as the closure of the internal attempt, lack of openness, demonstration and replicable; not interoperability, existing industries and regions serious barriers, a large number of repeated demonstration project construction; products, solutions programs are not compatible, the lack of a unified concept, resulting in a large number of fragments of the framework and application. To solve these problems, based on the analysis of the function of each part and the features of the IOT system, starting from the basic principle of Web based networking business environment, networking system architecture can be divided into perception and business domain domain structure was put forward based on the Web system of the Internet of things, things will be Web. The construction of Web based networking system service platform, gathering equipment and platform of the industry chain, the introduction of domestic and foreign advanced technology And idea, form the Internet of things application equipment store, provide users with a full range of experience and services, eventually form the Internet of things application service cloud, and build the Internet of things ecosystem.
2. in wireless sensor networks, clustering is an effective topology control method, a clear hierarchical structure, taking into account the load balancing, can enhance network reliability, scalability and robustness. Many existing clustering protocol for cluster head distribution is not uniform, the cluster head node selection did not fully consider the influence of residual energy with the communication distance, and the change of clustering is not with the network state changes and other issues. To solve these problems, we design a dynamic adaptive clustering based on virtual coordinate (a Virtual Coordinates Dynamic Adaptive Clustering VCDAC) protocol, the protocol of network clustering according to the idea of virtual partition, and according to the state dynamic network decision clustering number, thereby reducing the energy consumption. The simulation results show that this protocol can significantly prolong the survival time of the network, network load balancing, sending more data to the sink node Point, with excellent performance.
3. topology control in wireless sensor networks have great influence on the network performance, excellent topology can improve the efficiency of routing and MAC protocols for target localization, time synchronization and data fusion based clustering method. As a hierarchical topology control clear, can prolong the survival time of the network, are often transformed into to find the optimal solution of optimization problem. Traditional optimization analysis method needs a large amount of calculation, calculation with the exponential growth of the scale of the problem, for the micro sensor, computation and less resource requirement, and can produce acceptable results of the optimization method is suitable for the intelligent optimization method of bionic particle swarm algorithm is a kind of high calculation efficiency. This paper proposes a non parametric Niching Particle Swarm Optimization algorithm, and verify its effectiveness, and then using the algorithm, a niche particle swarm optimization based on the design. Adaptive clustering (an Adaptive Clustering Protocol using Niching Particle Swarm Optimization ACP-NPSO) protocol, the protocol of cluster distribution uniformity, balance the load of the system. The simulation results show that the protocol has good efficiency and can prolong the network life cycle, improve the throughput of the system.
4. application of wireless sensor network has a strong correlation, different routing protocols in the application are very different, not a general routing protocol, is a closely related to the application of network routing protocol. According to the logical structure of the different divided into cluster based routing protocol and planar routing protocol, routing protocol and plane in comparison, cluster based routing protocol structure clear, efficient use of node resources. This paper designs a dynamic particle swarm clustering routing protocol based on Ant Colony Optimization (a Dynamic Clustering Routing Protocol using Particle Swarm Optimization and Ant Colony Algorithm, DCRP-PSOACA), we give a detailed design of DCRP-PSOACA, the simulation results show that, this protocol can effectively reduce the energy consumption of the nodes, the routing path provides a robust, prolong the survival time of the network to achieve efficient communication.

【学位授予单位】：北京邮电大学
【学位级别】：博士
【学位授予年份】：2014
【分类号】：TP391.44;TN929.5

【参考文献】