【作者】 张彬彬；

【导师】 石文孝；

【作者基本信息】 吉林大学 ， 通信与信息系统， 2019， 硕士

【摘要】 无线Mesh网络(Wireless Mesh Network,WMN)具有低成本、易维护、鲁棒性等优点,是下一代无线网络的关键组网技术之一。与单接口的WMN相比,多接口多信道(Multi-Radio Multi-Channel,MRMC)的WMN可以有效提升网络吞吐量,但MRMC WMN增加了网络中并行传输链路的数目,使链路间的干扰增大。合理的信道分配方案可以有效降低MRMC WMN链路间干扰,提升网络容量,因此,设计出合理的信道分配算法尤为重要。目前人们对MRMC WMN的信道分配算法的研究多集中在正交信道。常用的IEEE 802.11b/g可用的正交信道只有3条,频谱资源有限,同信道干扰问题严重制约了网络性能。随着通信技术的发展,部分重叠信道带来的干扰可以得到有效抑制,目前已有研究表明使用部分重叠信道可以有效提升网络性能。因此,对部分重叠信道分配算法的研究具有重要意义。本文针对单播WMN部分重叠信道干扰度量不精确的问题,通过构建信干噪比矩阵度量链路间干扰,以最小化干扰为目标提出一种负载感知POCs信道分配算法IAPOCA(Interference-Aware Partially Overlapped Channels Assignment,IAPOCA)。该算法利用信干噪比矩阵指导信道分配,综合考虑链路使用不同信道时对其它链路和全网的影响,为负载和干扰大的链路优先分配干扰最小的信道,以达到降低链路间干扰和提高网络吞吐量的目的。多播通信是点对多点的通信方式,合理的利用多播技术可以提高频谱的利用率和网络的吞吐量。本文针对多播WMN部分重叠信道的干扰度量不精确的问题,使用区域负载表示节点对全网负载的影响,提出了一种多播WMN部分重叠信道分配算法LIAM(Load and Interference-Aware Multicast,LIAM)。该算法在计算负载干扰权重时,根据多播的单向传输特性,分别计算父节点对其它节点的干扰和子节点在接收信号时受到的干扰,依据负载干扰权重为节点分配干扰最小的信道。在分配信道时,将父节点和其子节点的所有链路视为一个整体,充分利用无线广播优势。本文使用Matlab实现所提单播和多播算法的信道分配结果,使用NS-3完成算法的性能仿真。仿真结果表明,LIAM和IAPOCA算法可以有效的提升整体网络的吞吐量、降低端到端时延和丢包率。更多还原

【Abstract】 Wireless Mesh Network(WMN)with the advantages of low cost,easy maintenance and robustness.It is one of the key networking technologies for the next generation Wireless networks.Compared with single-interface WMN,multi-interface multi-channel(MRMC)WMN can effectively improve the network throughput.But MRMC WMN increases the number of parallel transmission links in the network and increases the interference between links.Reasonable channel allocation scheme can reduce interference between the links and increase network capacity.Therefore,it is particularly important to design a reasonable channel allocation algorithm.At present,most of the research on channel assignment algorithm of MRMC WMN focuses on orthogonal channel.The commonly used IEEE 802.11b/g has only three available orthogonal channels,and the spectrum resources are limited.With the development of communication technology,interference caused by partially overlapping channels can be effectively suppressed.At present,studies have shown that the use of partially overlapping channels can effectively improve network performance.Therefore,it is of great significance to study the algorithm of partially overlapped channel assignment.In this paper,to solve the problem of imprecision measurement of unicast WMN partially overlapping channel interference,a load-aware POCs channel assignment algorithm(IAPOCA)is proposed,by constructing a signal-to-noise ratio matrix to measure inter-link interference and aiming at minimizing interference.This algorithm uses the signal to interference plus noise ratio(SINR)matrix to guide channel assignment,and comprehensively considers the influence of different channels on other links and the whole network.It assigns the channel with the least interference to the links with large load and interference,so as to reduce inter-link interference and improve network throughput.Multicast communication is a point to multipoint communication method.Reasonable use of multicast technology can improve spectrum utilization and network throughput.In this paper,aiming at the problem that the interference measurement of multicast WMN partially overlapped channels is inaccurate,the regional load is used to represent the influence of nodes on the whole network load,and a multicast WMN partially overlapped channel assignment algorithm(LIAM)is proposed.When calculating the load interferenceweight,according to the one-way transmission characteristics of multicast,this algorithm calculates the interference of the parent node to other nodes and the interference of the child node when receiving the signal.It assigns the channel with the least interference to the node according to the load interference weight.In channel assignment,all links of the parent node and its children are considered as a whole to make full use of the advantages of wireless broadcasting.In this paper,Matlab is used to achieve the channel assignment results of the unicast and multicast algorithm,and NS-3 is used to complete the performance simulation of the algorithm.Simulation results show that the LIAM and IAPOCA algorithms can effectively improve the overall network throughput,reduce end-to-end delay and packet loss rate.更多还原