【作者】 王娟；

【导师】 冯海林；

【作者基本信息】 西安电子科技大学 ， 应用数学， 2009， 硕士

【摘要】 在无线传感器网络中,如何合理使用网络中的节点以延长网络寿命是传感器网络规划中的一个关键问题。节点调度算法是解决上述问题的一种可行方案,它以节点轮流工作的方式来节省能源,延长网络寿命。目前很多调度算法需要知道节点的地理位置信息,需要借助GPS等定位系统,不利于构建低成本的传感器网络;已有的一些应用较好的随机调度算法都是基于布尔感知模型建立的,这不符合实际的情形。基于以上问题,本文以随机调度算法为基础,主要作了以下一些工作:首先概述了随机调度算法并分析了其优缺点,随机调度算法是一种基于网络局部覆盖的方法,无需知道节点的位置信息、容错性好,且局部覆盖控制策略可以简化网络的协议设计,使网络配置更加灵活。但这些算法使用的布尔感知模型并不符合实际的感知情形,为此本文采用更符合实际的概率感知模型,在此感知模型下,分析了随机调度算法中网络覆盖度与节点数之间的关系,给出了解析表达式,利用该表达式可确定出需要布撒的节点数,解决了随机调度算法的节点配置问题。其次,针对随机调度算法存在的另一个问题,即节点是利用产生随机数的方法加入到不同的工作子集,从而导致初始子集中节点分布不均,本文在概率感知模型下提出了一种基于节点平均度的随机调度算法。该算法通过邻居节点之间信息的传输,利用节点平均度约束使每个节点的邻居节点尽量均匀分散在不同的子集中。仿真结果表明,经新算法处理之后,每个子集中的节点呈现均匀分布;分析表明,新算法具有极低的时间复杂度和消息复杂度,具有很好的实用性。更多还原

【Abstract】 In Wireless Sensor Networks (WSN), it is a crucial problem that how to prolong the network lifetime using the nodes reasonably in programming sensor networks. Node scheduling algorithm is a feasible method to resolve above-mentioned question, which refers to make the nodes work in turns to save energy. At present, many scheduling algorithms need geographic information with the help of GPS positioning system, which is not conducive to build a low-cost sensor networks; at the same time, all the existing random scheduling algorithms which have better applications are on a basis of Boolean sensing model, and it doesn’t accord with the actual case. In view of the above- mentioned problems, based on the existing random scheduling algorithm , we have accomplished the following work.First of all, the existing random scheduling algorithm is described briefly, and the strengths and weaknesses are analyzed. The random scheduling algorithm is on local coverage of network and without the help of geographic information, and has a good fault tolerance. In addition, using local coverage control strategy can not only simplify the protocol design of network, but also configure nodes more flexibly. But the Boolean sensing model used doesn’t accord with the actual sensing case, so based on probabilistic sensing model more close to the practical applications, we have analyzed the relationship between coverage intensity and the number of nodes and given the expression by which we can get the nodes needed to deploy. So configuring nodes in random scheduling algorithm has been resolved.Secondly, to resolve another question in random scheduling algorithm, that is, how to make the nodes evenly distributed in every subset composed of some sensor nodes, a new method - a random scheduling algorithm based on the average degree of nodes is presented, which is on the basis of probabilistic sensing model. Taking advantage of information transmission between neighbor nodes and average degree of nodes, it makes every node’s neighbor nodes distributed in different subsets evenly. The simulation results show that the nodes in every subset brought by the new algorithm are uniformly distributed, and analysis shows that the new algorithm can work with a very low time complexity and message complexity, so it has a great practicability.更多还原