基于随机几何图的无线传感器网络拓扑控制-应用数学专业论文.docxVIP

摘 要 如何合理利用能量等资源一直是无线传感器网络研究领域的关键问题，而拓 扑控制是目前有效利用能量的关键技术之一。本文给出基于随机几何图的无线传 感器网络拓扑控制方法。 首先，给出一种基于临近图所生成的拓扑来调整节点发射功率的无线传感器 网络的拓扑控制策略。该方法打破在临近图拓扑中采取平面路由的传统思想的束 缚，不再限定网络中数据收发仅在所生成的临近图拓扑边上进行，而是以临近图 的连通特性为依据，引入三个通信范围调整的控制系数，合理地扩大节点的通信 范围，使得各个节点有合适数量的邻居节点，有效减少通信干扰发生的同时，提 高数据收发的成功率，保证网络的通信质量。 其次，给出了基于轮盘赌的概率路由控制机制。该方法将轮盘赌的思想与概 率路由相结合，利用定义的路径成本，按照轮盘赌的方式以一定概率选择出下一 跳节点。该方法考虑了节点的剩余能量信息，通过随机的方式均衡节点间的通信 负荷，避免特定节点被过度消耗。在临近图拓扑控制策略下运行该机制，能够更 有效地利用节点能量，延长网络生存时间。 最后，以仿真实验检验了本文所给拓扑控制策略的节点度和数据收发成功率 方面的性能，并对给出的控制系数的范围作了相关实验分析。同时，与轮盘赌概 率路由机制相结合，考察了二者在能量消耗方面的优越性，从整体上验证了本文 方法的合理性与有效性。 关键词：无线传感器网络 拓扑控制 临近图 概率路由 轮盘赌 Abstract In wireless sensor networks, each node’s energy is limited. This makes it a hot issue that how to utilize energy efficiently. As one of the key technologies of wireless sensor networks, topology control has been extensively researched, which aims at enhancing the efficiency of energy utilization and extending the lifetime of the whole network. First of all, this paper analyzes existing geometric random graph based topology control techniques and points out some drawbacks of them, proximity graph based methods in particular. On that basis, combined with three control parameters, a novel topology control strategy based on proximity graph is presented. It breaks the rule to use the planarity of the generated proximity graphs, i.e. no longer restricts data communication on the edges of proximity graphs. The greatest advantage lies in the fact that it can enlarge node degrees and increase data delivery rates of existing localized routing methods. Meanwhile, to reduce the whole network’s energy consumption, roulette thought is introduced to the process of probabilistic routing. In consideration of the defined path cost, which is determined by energy consumption of sending data and node’s residual energy, the next hop relay node is selected through roulette-based probabilistic routing mechanism with certain probability. To some extent, this mechanism avoids the overconsumption of