能量感知的无线传感器网络路由算法研究

发布时间：2019-02-20 20:53

【摘要】：无线传感器网络采用自组织的方式将数量众多的造价低廉、运行能耗较低的微型传感器节点构建成网络,实现对目标检测区域各种信息的实时感知、监测和采集,并对采集到的数据进行处理分析。无线传感器节点具有使用方便灵活、稳定性强、扩展方便、经济性好、安全可靠等特点,在军事监控、系统控制、医疗卫生和环境监测以及其他行业都得到了广泛的推广,受到各行各业研究开发人员的高度重视。在实际开发应用中还存在着许多影响网络进行大规模推广的障碍。例如:在网络的使用过程中,需要进行部署的传感器节点数量众多,且位置随机、监测环境复杂,这使得部分节点在运行过程中能量消耗剧烈,会造成能量很快消耗殆尽,节点停止工作;网络进行大范围监测时,要使用大量的节点进行部署,造成整体成本较高,影响网络的推广与普及;在实际部署中,传感器节点往往使用电池为节点进行供电,节点的微型化发展使得电池容量受到限制,而且节点数量的巨大会造成对节点进行电池更换十分不便。通过采用能量模型优化策略对网络进行研究,从网络路由拓扑出发,为建立高效、节能的网络提供新的思路。针对网络能耗不均衡问题,根据不同类型的应用特点,综合节点能耗模型和多路径路由技术,对节点剩余时间预测、网络拓扑、时间同步进行研究。(1)能量模型方面,将组成CC2530传感器节点的各个模块能量消耗特点和节点电池放电规律进行统合研究,构建能量模型实现对节点运行过程中各模块消耗情况的统计,以及对节点工作时间和剩余能量的统计,并根据节点工作能耗情况对剩余工作时间进行预测。针对TinyOS2.x的仿真工具无法实现节点能量检测的不足,设计计算节点剩余能量的TinyOS2.x能量组件,实现能量监测,有利于对网络能量进行研究。(2)路由协议方面,由于节点使用的电池容量受限,运行过程中网络拓扑会不断发生变化,路由要随应用的变化而变化。采用非均匀分簇方法进行簇首选择;通过节点竞争力和节点能量管理机制相互协作决定最优簇首节点;根据传输能耗、剩余能量和传输距离等影响因素采用簇间连通算法进行簇间多跳路径选择,实现负载的均衡分配。(3)时间同步方面,针对传感器节点数据时间不一致,提出时间同步算法。使用网络分层策略减少同步通信开销;采用同步误差补偿机制降低同步误差的影响,利用时钟补偿机制减少运行的累积误差。在保证时间精度前提下,降低同步次数减少同步开销。
[Abstract]:Wireless sensor networks (WSNs) use self-organization to construct a network of small sensor nodes with low cost and low energy consumption, which can realize real-time sensing, monitoring and collecting of various information in target detection area. The collected data are processed and analyzed. Wireless sensor nodes have been widely used in military monitoring, system control, health care, environmental monitoring and other industries, with the advantages of convenient use, strong stability, easy expansion, good economy, safety and reliability, etc. By a variety of industry research and development personnel attached great importance. There are still many obstacles in the development and application of network. For example, in the process of using the network, the number of sensor nodes that need to be deployed is large, the location is random, and the monitoring environment is complex. The node stops working; When the network is monitored on a large scale, a large number of nodes should be deployed, which results in high overall cost and affects the popularization and popularization of the network. In actual deployment, the sensor nodes often use batteries to power the nodes, the development of the node miniaturization limits the battery capacity, and the huge number of nodes will cause the node battery replacement is very inconvenient. By using the energy model optimization strategy to study the network, starting from the network routing topology, this paper provides a new way of thinking for the establishment of efficient and energy-saving network. Aiming at the imbalance of network energy consumption, according to the characteristics of different types of applications, combined with node energy consumption model and multipath routing technology, the residual time prediction, network topology and time synchronization of nodes are studied. (1) Energy model. The characteristics of energy consumption of each module of CC2530 sensor node and the discharge law of node battery are studied, and the energy model is constructed to realize the statistics of the consumption of each module during the operation of the node. The working time and residual energy of the node are counted, and the remaining working time is predicted according to the working energy consumption of the node. Because the simulation tool of TinyOS2.x can not realize the energy detection of nodes, the design of TinyOS2.x energy component to calculate the residual energy of nodes is beneficial to the research of network energy. (2) in the aspect of routing protocol, the design of TinyOS2.x energy component to calculate the residual energy of nodes is beneficial to the research of network energy. Due to the limited battery capacity used by the nodes, the network topology will change continuously during operation, and the routing will change with the application. The non-uniform clustering method is used to select the cluster heads, and the optimal cluster heads are determined by the cooperation of the node competitiveness and the energy management mechanism of the nodes. According to the influence factors such as transmission energy consumption, residual energy and transmission distance, inter-cluster connectivity algorithm is used to select multi-hop paths between clusters to achieve load balancing. (3) in the aspect of time synchronization, the data time of sensor nodes is inconsistent. A time synchronization algorithm is proposed. The network delamination strategy is used to reduce the synchronous communication overhead, the synchronization error compensation mechanism is adopted to reduce the influence of synchronization error, and the clock compensation mechanism is used to reduce the cumulative error. On the premise of ensuring time precision, the synchronization times are reduced and the synchronization overhead is reduced.
【学位授予单位】：兰州交通大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TP212.9;TN929.5

【参考文献】