基于节点连接能力的散发等待路由

发布时间：2018-06-04 22:56

本文选题：散发等待路由 + 节点连接强度　；参考：《安徽大学》2017年硕士论文

【摘要】：因特网的飞速发展使得全球人,不管是在生活上还是工作上,生活方式和工作方式都发生了巨大的变化。众所周知,科技不断在创新,但是在特殊的场合、特定的条件下,传统网络仍然满足不了人类的需求。诸如,某地区发生严重地震、地质灾害、火灾等情况,通信基础设施将会被严重摧毁,人们无法与外界进行通信,因此,人类要不停地探索新技术,来解决人们生活中的问题。DTN(Delay Tolerant Networks)网络应用而生,它适用于这种存在挑战性的环境之中。针对DTN的本身固有结构和特点,研究者们相继提出许多关于DTN的路由算法。诸如,蔓延路由、散发等待路由、概率路由等经典路由算法,使得网络中节点传输报文灵活,可是也会有不足之处。例如,在节点转发报文时,转发副本数有一定的盲目性,没能针对节点的性质和节点连接能力进行综合考虑。另外,网络局部或全部拥塞、资源大量被浪费等现象时常出现。一种改进型的散发等待路由算法在本文提出,综合衡量节点连接能力,动态地分发报文副本,一定程度上网络性能明显改善。在本文中,主要研究DTN网络的体系架构、DTN中几类路由算法,着重展开的是针对散发等待路由的深入研究,大致完成了如下几项工作:(1)本文先着重介绍了 DTN的研究背景状况,在此基础上分析了国内外针对DTN网络研究的现状,之后,关于DTN的由来和特点进行详细介绍,针对DTN中工作机制进行简要描述,且列举其在现实生活中的应用。(2)按照节点相遇之前的历史记录信息(如相遇频率、相遇持续时间、相遇总接触时间、平均接触时间等)可以大致预测节点未来一段时间内的连接机遇,因为节点之间相遇愈频繁表示报文交换愈频繁,相遇持续时间愈长则表示交换的报文数目愈多。基于节点历史信息,本文提出平均分离时间的概念,它既能反映接触频率又能反映出相遇接触时间。(3)在平均分离时间的基础上,提出了节点连接强度的概念,在某种程度上,有利于提高节点的性能,节点连接强度越大则表示节点性能越好;考虑到节点的连接是否稳定,基于平均分离时间又提出了节点连接稳定性的概念,利用平均分离时间的方均根值来衡量,当节点连接稳定性值越高时,意味着节点的性能也越好。(4)综合考虑节点连接强度和节点连接稳定性来衡量节点连接能力,使用The ONE仿真器,对本文提出的路由算法进行了仿真实验,并与其它传统路由算法相对比。综上所述,本文提出的基于节点连接能力的散发等待路由算法不但能提高递交率,而且在一定程度上降低了网络开销,有一定的理论意义和价值。
[Abstract]:With the rapid development of the Internet, people in the world, whether in life or work, lifestyle and working style have undergone tremendous changes. It is well known that technology is constantly innovating, but under special circumstances and conditions, the traditional network still can not meet the needs of human beings. For example, serious earthquakes, geological disasters, fires in a certain area, communications infrastructure will be severely destroyed, people can not communicate with the outside world, therefore, human beings have to constantly explore new technologies, To solve the problems in people's lives. DTNT delay Tolerant Networks) network applications, it can be used in this challenging environment. In view of the inherent structure and characteristics of DTN, many routing algorithms about DTN have been proposed one after another. Some classical routing algorithms, such as spread routing, distributed wait routing, probabilistic routing and so on, make the nodes transmit packets flexibly in the network, but there are also some shortcomings. For example, the number of forwarded replicas is blind when the nodes transmit packets, which fails to consider the nature of the nodes and the ability of the nodes to connect. In addition, local or total network congestion, a large number of waste of resources and other phenomena often occur. An improved routing algorithm for distributed waiting is proposed in this paper, which synthetically measures the ability of nodes to connect and dynamically distributes copies of packets. To a certain extent, the network performance is obviously improved. In this paper, we mainly study several routing algorithms in DTN network architecture, focusing on the in-depth study of distributed wait routing, and accomplish the following tasks: 1) this paper first introduces the background of DTN research. Based on the analysis of the current situation of DTN network research at home and abroad, the origin and characteristics of DTN are introduced in detail, and the working mechanism in DTN is briefly described. And enumerate its application in real life. 2) according to the historical record information (such as meeting frequency, meeting duration, total contact time) before the node meets, The average contact time and so on) can roughly predict the connection opportunities in the future, because the more frequent the encounter between the nodes, the more frequent the message exchange, the longer the encounter duration, the more the number of packets exchanged. Based on the historical information of nodes, this paper puts forward the concept of average separation time, which can not only reflect the contact frequency but also the encounter contact time. (3) on the basis of the average separation time, the concept of node connection strength is put forward, to a certain extent, The higher the connection strength, the better the node performance. Considering whether the connection is stable or not, the concept of node connection stability is proposed based on the average separation time. Using the square root value of the average separation time, when the node connection stability value is higher, it means that the node performance is also better. 4) considering the node connection strength and the node connection stability synthetically to measure the node connection ability. Using The ONE simulator, the proposed routing algorithm is simulated and compared with other traditional routing algorithms. To sum up, the proposed routing algorithm based on node connection ability can not only improve the delivery rate, but also reduce the network overhead to a certain extent, which has certain theoretical significance and value.
【学位授予单位】：安徽大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TP393.0

【相似文献】