基于功率控制的三维无线Mesh网络拓扑控制研究

发布时间：2018-07-06 09:48

本文选题：无线Mesh网络 + 三维空间　；参考：《电子科技大学》2014年硕士论文

【摘要】：与传统网络不同,无线Mesh网络(Wireless Mesh Network,WMN)是一种动态的、自组织自配置的灵活网络。WMN中的各节点可独立完成节点参数配置,并且自动建立和维护网络连接。WMN具有很多优势,如数据可靠性强、组网成本低、网络维护简单、业务覆盖面广等。WMN作为一种新型组网技术,已经是无线通信时代的宠儿。但是,在三维场景中Mesh节点的分布更密集,存在节点干扰强、资源竞争激烈、网络性能差等问题。功率控制技术与拓扑控制技术在提高网络性能中都起到了至关重要的作用。通过功率控制技术得到合理的发射功率,合理的发射功率即可以减少节点干扰,又可以提高资源利用率。通过拓扑控制技术选择出性能较好的链路,回避性能较差的链路,有效地提高了网络性能。因此本文将两种技术相结合,先由功率控制技术得到合理的发射功率,然后由拓扑控制技术选择合理的链路。论文主要工作包括:1.目前WMN的功率控制算法大多数仅针对二维平面单优化目标的情形,而针对三维空间多优化目标的研究较少。本文研究了一种基于遗传算法的三维功率控制算法(3D-PCGA),该算法以节点干扰强度和节点连通度为优化目标,使用遗传算法求解后可以为Mesh节点配置优化的发射功率。将3D-PCGA算法与单目标优化算法做网络性能对比,在数据流数大于21条时,吞吐量提高13%,端到端时延缩短了17.5%,包提交成功率提高了14%。2.在大多数WMN拓扑控制算法中节点会采用最大发射功率进行通信,但发射功率大必定会增加网络干扰,从而导致网络吞吐量降低、时延增加、丢包率提高等问题。本文在3D-PCGA算法的基础上,研究了一种冲突链路最小的三维拓扑控制算法(3D-CSTC)。该算法首先计算每条链路的冲突链路大小,然后选择冲突链路较小的链路连接到网络,回避冲突链路较大的链路。3.网络容错性、端到端时延、吞吐量以及包提交成功率是拓扑控制算法考虑的重要指标。本文与其他三维拓扑控制算法相对比,仿真结果表明:由3D-CSTC算法建立的拓扑结构具有更强的网络容错性;当节点分布密度小于41.25 10??个/m3时,3D-CSTC算法的平均节点度可以提高25%;当数据流数大于18时,3D-CSTC算法可以将网络吞吐量提高9%,端到端时延缩短22%,包提交成功率提高了8%。
[Abstract]:Different from the traditional network, Wireless mesh Network (WMN) is a kind of dynamic, self-organized and self-configured flexible network. WMN nodes can independently complete the node parameter configuration, and automatically establish and maintain the network connection. WMN has many advantages. As a new networking technology, WMN, such as strong data reliability, low network cost, simple network maintenance and wide service coverage, has become the darling of the wireless communication era. However, the mesh nodes are more densely distributed in the 3D scene, there are some problems, such as strong interference of nodes, fierce competition of resources, poor network performance, and so on. Both power control and topology control play an important role in improving network performance. The reasonable transmission power can be obtained by power control technology, which can not only reduce the node interference, but also improve the resource utilization ratio. The network performance is improved effectively by using topology control technology to select the link with better performance and avoid the link with poor performance. Therefore, the two technologies are combined in this paper. Firstly, the reasonable transmission power is obtained by the power control technology, and then the reasonable link is selected by the topology control technology. The main work of the thesis includes: 1. At present, most of the power control algorithms of WMN are only for the case of two-dimensional single target, but there are few researches on the multi-optimization of three-dimensional space. In this paper, a 3D power control algorithm based on genetic algorithm (3D-PCGA) is proposed. The algorithm takes the intensity of interference and the connectivity of nodes as the optimization targets, and the optimal transmission power can be configured for mesh nodes by using genetic algorithm. The network performance of 3D-PCGA algorithm is compared with that of single objective optimization algorithm. When the number of data streams is more than 21, the throughput is increased by 13 percent, the end-to-end delay is shortened by 17.5and the success rate of packet submission is increased by 14.2. In most WMN topology control algorithms, the nodes use the maximum transmit power to communicate, but the high transmission power will inevitably increase the network interference, which will lead to the decrease of network throughput, the increase of delay, the increase of packet loss rate and so on. Based on the 3D-PCGA algorithm, a 3D topology control algorithm with minimal collision link (3D-CSTC) is studied in this paper. The algorithm first calculates the size of the collision link for each link, then selects the link with a smaller collision link to connect to the network, avoiding the link. 3. Network fault-tolerance, end-to-end delay, throughput and packet submission success rate are important parameters of topology control algorithm. Compared with other 3D topology control algorithms, the simulation results show that the topology structure established by 3D-CSTC algorithm is more fault-tolerant when the node distribution density is less than 41.25 10? The average node degree of 3D-CSTC algorithm can be increased by 25% when the number of data flows is more than 18:00. When the data flow number is greater than 18:00, the 3D-CSTC algorithm can increase the throughput of the network by 9 steps, shorten the end-to-end delay by 22 points, and increase the success rate of packet submission by 8%.
【学位授予单位】：电子科技大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：TN929.5

【参考文献】