【作者】 直敏；

【导师】 原萍；

【作者基本信息】 东北大学 ， 计算机应用技术， 2008， 硕士

【摘要】 无线传感器网络(Wireless Sensor Network, WSN)是由部署在监测区域内大量的微型传感器节点,通过无线通信方式形成的一个多跳的自组织的网络系统。其目的是协作地感知、采集和处理网络覆盖区域中感知对象的信息,并发送给观察者。无线传感器网络是计算机科学技术的一个新的研究领域,具有非常广阔的应用前景,已经引起了学术界和工业界的高度重视。路由技术是无线传感器网络的关键技术,也是影响网络整体性能的重要因素之一。由于无线传感器网络和其它的通信网路,包括无线Ad hoc网络相比,有着截然不同的特点,这就使得无线传感器网络的路由研究极具挑战性。近年来提出了不少专门针对无线传感器网络的路由协议,如泛洪法(Flooding)、信息协商传感器协议(SPIN)、定向扩散协议(DD)等。本文详细的研究了这几种路由协议,并在此基础上提出了基于蚁群优化算法的无线传感器网络路由算法。该算法是由Sink节点在整个网络中定期的广播名为IMP(Interest Message Packet)的报文,向整个网络传送Sink节点所感兴趣的数据特征,并且建立和更新梯度场,形成和更新各条路径上的信息素。信息素的形成不仅要考虑节点之间的梯度,还要考虑节点的剩余能量。当某个节点发现符合要求的数据时,该节点成为源节点,并向Sink节点发送类似蚂蚁的数据包。节点在选择路由的时候,根据信息素的浓度来计算各相邻节点被选择作为下一跳的概率。也就是说,梯度大而且剩余能量多的相邻节点更有可能成为下一跳,因此数据包的发送不会总是沿着一条路径进行。这样,在尽量沿着梯度最大的路径发送数据的同时,尽可能平均的消耗各个节点的能量。同时,为了防止一些节点由于作为中间节点转发数据包过早死亡而导致这些节点所在的区域失去监控,在算法中还设置了节点的能量阈值,当节点的能量小于能量阈值时,就尽可能不要再承担转发数据的任务。通过对该路由算法的仿真,验证了该算法有效性,同时,仿真表明算法具有较小的平均延迟和平均能量消耗。更多还原

【Abstract】 Wireless Sensor Networks(WSN)are composed of a large quantity of micro-sensors deployed in monitoring fields, they form the multi-hop and self-organizing network systems via wireless communicating. These sensors are used in cooperation with others to feel, collect, and process the information transmitted to the observer in monitoring fields. As a new researching area of computer science and technology, WSN, which will be used widely in the future, has aroused high attention in both academia and industrial circles.Routing technique is not only the key technique in WSN, but also one of the most important factors affecting the performance of the network. Investigations about routing in WSN is very challenging due to several characteristics that distinguish it from contemporary communication and Ad hoc networks. In recent years, some specific routing protocols for WSN have emerged, such as Flooding, SPIN, DD, and so on. This paper makes detail investigations on those routing protocols. On the basis of assimilating the excellent ideas among them, an Ant-Colony-Based Routing Algorithms in WSN is proposed.Ant-colony-based routing algorithm is that: Sink node regularly broadcasts IMP(Interest Message Packet) through the network in order to setup and update gradient field, at the same time, it can also form and update pheromone on each route, then transmit data which is interested by Sink node to the whole network. The gradient and remaining energy of the neighborhood are integrated into the computation of pheromone concentration. If a node has the right data to send, it becomes a source node which will transmit ant-like data packet to Sink. When choosing the route, the node will choose its next hop with probability, which is computed on the basis of pheromone concentration. And the data packet will not always be transmitted on the same route. Meanwhile, some nodes just live for a little while because of forwarding too many data packets from sources, and it leads to the loss of monitoring for some particular area. So an energy threshold for each node is set in the algorithm to prevent it. When the energy of some node is less than the energy threshold, it should avoid undertaking data-transmitting task. The simulation of Ant-Colony-Based Routing Algorithm proves itself to be effective, and simulation results also show that such algorithm is of less average delay and average dissipated energy更多还原