【作者】 汤强；

【导师】 汪秉文；

【作者基本信息】 华中科技大学 ， 控制理论与控制工程， 2010， 博士

【摘要】 拓扑控制是无线传感器网络中一项关键技术,它不仅可以改善网络拓扑结构,同时也可为其他网络层次的算法提供基础平台。无线传感器网络是一个能量受限的网络,拓扑控制算法可针对该特点构造专门的网络拓扑结构,从而达到提高网络寿命以及均衡节点之间能量的目的。因此,本文在基于能量受限的无线传感器网络的基础上设计几种拓扑控制算法,以达到延长网络生命时间和提高节点之间的能量均衡性的目的。拓扑控制是指在网络满足一定连通度和覆盖度的情况下,通过调整节点的传输功率来选择合适的骨干节点并将网络中多余的通信链路删除,以达到优化网络数据传输方式的一种网络管理方法。层次型拓扑控制是拓扑控制策略中非常重要的一种。层次型拓扑控制是指通过选择合适的骨干节点,将网络分割成多个层次的数据传输子网。层次型结构可使非骨干节点在没有进行通信的时候进入到休眠状态,同时在骨干节点上对接收的数据包进行融合处理,从而通过减少发送的数据量来达到节能的目的。在层次型拓扑控制中,节点之间的路径不需要进行定期维护。网络运行是基于轮的方式。在每轮中,所有节点都要参与网络的重构。在重构过程中,簇头的选择非常重要,如果簇头选择不当将造成网络中节点剩余能量的不均衡,从而造成网络生命时间的下降。因此,论文通过设计合理的簇头预测算法对网络簇头进行选择。当某轮结束的时候,簇头节点对接下来几轮的簇头进行预测,预测依据是节点的剩余能量。在簇头预测机制的基础上,文章提出了层次型拓扑控制算法CP-DCRP。如果基站的计算能力允许,则可以采用基站计算然后广播的方式,并且在簇头之间采用多跳传输机制,避免因簇头直接和基站进行通信而造成能量浪费的情况。多跳传输机制有多种,本文采用基于能量距离的局部最短路径算法。根据上述构想,论文提出了一个多方面综合改进的拓扑控制算法SCPEBMT。当节点传输数据时,如果选择多跳传输方式,则下一跳节点的选择至关重要。邻居集合的创建是网络节点选择下一跳节点时非常关键的一环。如果邻居集合中的节点选择比较合理,同时数量比较充足,则会使网络节点在发送数据到下一跳节点时不仅能够节省很多能量,同时也会使邻居节点集合中节点的剩余能量比较均衡。因此,论文提出了基于邻居集合的两个拓扑控制算法RNS-MCRP和NS-MCRP。在分簇拓扑中,节点之间的数据传输方式可以是多跳传输也可以是单跳传输。但是,对于某个具体的拓扑控制算法的某个具体的簇来说,可能由于多跳机制的不合理性,导致多跳传输方式的能耗大于单跳传输方式的能耗。因此,可通过寻找一个能耗临界点来设计基于混合传输方式的拓扑控制算法。基于上述思想,文章提出了基于混合传输的拓扑控制算法MS-LEACH。在层次型拓扑控制中,基于骨干节点的拓扑控制算法是对基于分簇结构的拓扑控制算法的一种扩展。基于骨干节点的拓扑控制算法一般采用启发式算法实现。所生成的骨干节点组成了数据转发与收集的骨干网,并且该骨干网覆盖了网络中所有的非骨干节点。因此,基于上述思想,论文提出了基于骨干节点的拓扑控制算法VBT-EBHRP,同时也提出了构造近似最小数目的骨干网构造算法MCDSA。在无线传感器网络中,构造近似最小数目节点的骨干网对于减少传输的数据包、降低干扰、降低延时都具有重要作用。论文最后对全文进行总结,指出文章的创新点以及提出了今后可继续研究的工作。更多还原

【Abstract】 Topology Control is a key technology of Wireless Sensor Network (WSN). The topology control algorithms can not only improve the network structure but also provide the basic platform for the algorithms of the other network layers. WSN is an energy constrained network, and the topology control algorithm can construct a special network structure to improve the network lifetime as well as the energy balance of the network. Thus, some topology control algorithms are designed based on the energy constrained WSN to prolong the network lifetime as well as improve the energy balance.If the connectivity as well as the coverage is satisfied, the management of the nodes is optimized by a topology control algorithm, which adjusts the transmission power and selects the appropriate backbone nodes and deletes the redundant links. The hierarchy topology control is a very important kind of the topology control strategy, which selects the backbone nodes and divides the network into different sub-networks with different levels. The hierarchy topology makes the non-backbone nodes sleep when the nodes are not communicating. Meanwhile the backbone nodes fuse the collected data for saving energy by decreasing the number of the packet data sent to the next hops.In the clustering topology control, the maintenance of the link between any pair nodes is not needed and the network’s running is based on the rounds. In every round, all nodes participate into the network reconstruction. In the process of reconstruction, the selection of the cluster head is very important, because selecting a good cluster head can be good for the nodes residual energy balance. So, a cluster head selection mechanism is designed, which is when some round is ended the cluster heads of this round predict some cluster heads of the next several rounds. The prediction is based on the node’s residual energy. So, based on the prediction mechanism a topology control algorithm called CP-DCRP is proposed in the dissertation.If the powerful computation ability is permitted, the base station (BS) computes and broadcasts the network information, and the multi-hop transmission can be applied among the cluster heads, which saves a lot of energy compared with the single-hop transmission. There are many kinds of multi-hop transmissions, and the local shortest path algorithm based on the energy distance metric is adopted. Based on the ideas described above, a comprehensive improvement topology control algorithm called SCPEBMT is proposed.If the network nodes transmit the data and the multi-hop transmission is adopted, selecting the next hop node is very important. The creation of the neighbor set is a critical step of selecting the next hop. If the nodes in the neighbor set are selected well and the number of the neighbors is adequate, a lot of energy can be saved and the energy balance of the nodes in the neighbor set is very good when a node transmit data to the next hop. So, based on the neighbor set, two topology control algorithms called RNS-MCRP and NS-MCRP respectively are proposed in the dissertation.In the clustering topology, the transmission between nodes can be multi-hop transmission or single-hop transmission. But, for the right specific cluster of a specific topology control algorithm, the energy consumptions of the multi-hop transmission and the single-hop transmission can be equal to each other because of the bad multi-hop transmission. So, a topology control algorithm based on the hybrid transmission can be designed, which is based on the critical point of the energy consumption, and thus a topology control algorithm called MS-LEACH is proposed.In the hierarchy topology control, the topology control algorithm based on the backbone is an expansion of the clustering topology control. The heuristic algorithm is applied in the backbone topology control algorithm design. All the backbone nodes construct the data collection and relay backbone network, which covers all the non-backbone nodes in the network. Based on the ideas described above, a topology control algorithm called VBT-EBHRP is proposed. At the same time, MCDSA is proposed for constructing a sub-network with the approximate minimum backbone node number. In the WSN, the backbone network with the approximate minimum node number is good for decreasing the number of the packet data, the interference and the delay.Finally, a summary is written for the research work, and the innovations as well as the future work are introduced.更多还原