无线网络中物联网业务入网建模与特性分析研究

发布时间：2018-04-27 13:57

本文选题：物联网 + 时延容忍　；参考：《西南交通大学》2017年硕士论文

【摘要】：随着现代通信技术及物联网相关应用业务的快速发展,在典型物联网业务通信传输中,其业务终端数量的迅速增长以及瞬时突发等特征表现的越来越明显,对承载网络的要求也越来越高.早期的通信网络业务模型假设系统到达过程为泊松分布,现在的互联网流量模型到达过程是重尾分布.今天网络发展已进入物联网时代,物联网中“物与物”通信的比例远高于人与人的通信,原有模型已不能满足这些物联网特征的要求。本文致力于物联网业务的建模和特性分析,基于理论分析与计算机模拟仿真相结合分析验证了典型物联网业务终端较多但每次传输数据量较小、网络流量瞬时突发这些特征给承载网络带来的各种影响,把传统的互联网业务模型和分析方法应用在物联网入网控制中。首先,本文分析了典型物联网业务的主要特征:终端数量不断增大、允许一定时延以及瞬时突发入网等。其次,对IEEE802.11无线局域网系统,分析了基于DCF协议的入网过程。对MAC层的服务时间进行了推导,研究了 MAC层服务模型以及IPP到达的业务模型,对系统服务质量指标如吞吐量、时延进行了仿真分析。接着,本文针对典型物联网业务特征,研究分析了 DCF马尔科夫链分析方法中的退避时间,根据物联网业务允许一定时延的特性,对原有算法的退避时间进行了改进。理论和仿真都证明了改进算法有利于减少入网业务终端不断增加时发生冲突的概率。然后,根据不同物联网业务的不同到达时间间隔,证明了 Beta分布能更好的刻画物联网业务的瞬时突发达到过程。基于不同情况建立了 M/M/1单一互联网业务模型、Beta/M/1单一物联网业务模型和Mix//M/1混合业务模型,实现了不同业务的同时入网仿真分析。最后,本文针对混合业务入网时产生的拥塞和抢占等问题,研究了两种有效的优化接入控制方法:一种是基于互联网业务优先的接入控制,可以保证互联网业务的实时性,而不会对典型物联网业务的数据传输有太大影响;另一种是基于入网业务不同比例的控制接入,以期达到各种业务的较理想的通信情况。
[Abstract]:With the rapid development of modern communication technology and Internet of things related application services, in the typical Internet of things business communication transmission, the rapid growth of the number of service terminals and the characteristics of instantaneous burst are becoming more and more obvious. The requirement of carrying network is higher and higher. The early traffic model of communication network assumes that the arrival process of the system is Poisson distribution, and the arrival process of the current Internet traffic model is heavy-tailed distribution. Nowadays, the development of Internet of things has entered the age of Internet of things. The proportion of "things and things" communication in the Internet of things is much higher than that of people. The original model can not meet the requirements of these characteristics of the Internet of things. This paper is devoted to modeling and characteristic analysis of Internet of things service. Based on the combination of theoretical analysis and computer simulation, it is proved that the typical Internet of things service terminal is more than others, but the amount of data transmitted each time is small. The characteristics of instantaneous burst of network traffic bring various influences to the carrying network. The traditional Internet traffic model and analysis method are applied to the control of Internet of things access network. Firstly, this paper analyzes the main characteristics of typical Internet of things services: the number of terminals is increasing, allowing a certain delay and instantaneous burst into the network. Secondly, the network access process based on DCF protocol is analyzed for IEEE802.11 WLAN system. The service time of MAC layer is deduced, the service model of MAC layer and the service model of IPP arrival are studied, and the quality of service index such as throughput and delay are simulated and analyzed. Then, aiming at the characteristic of typical Internet of things services, the Backoff time of DCF Markov chain analysis method is studied and analyzed. According to the characteristic that the Internet of things service allows a certain delay, the Backoff time of the original algorithm is improved. Both theoretical and simulation results show that the improved algorithm is helpful to reduce the probability of collision when the number of incoming service terminals increases. Then, according to the different arrival time intervals of different Internet of things services, it is proved that the Beta distribution can better describe the instantaneous burst arrival process of the Internet of things services. The M/M/1 single Internet service model Beta / M / 1 single Internet of things service model and the Mix//M/1 hybrid service model are established based on the different conditions, and the simulation analysis of different services is realized at the same time. Finally, aiming at the problems of congestion and preemption when mixed services enter the network, this paper studies two effective optimal access control methods: one is based on Internet traffic priority access control, which can ensure the real-time performance of Internet services. It does not affect the data transmission of typical Internet of things services; the other is the control access based on different proportion of network access services, in order to achieve a better communication situation of various services.
【学位授予单位】：西南交通大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TN92;TP391.44

【参考文献】