无线传感器网络能量有效性的研究

【作者】 王珺；

【导师】 谢俊元；

【作者基本信息】 南京大学 ， 计算机科学与技术， 2012， 博士

【摘要】 近年来,无线传感器网络(WSN)被广泛应用在军事国防、生物医疗、环境监测、抢险救灾、危险区域远程控制等许多重要的领域,由于传感器网络中的传感器节点能量严重受限,同时节点在部署之后很难回收,因而针对无线传感器网络的能量效率的研究一直是无线传感器网络研究的热点和难点问题,目前提高网络生命期主要的研究方法分为几大类：一是减少网络中总通信量,主要可以通过数据融合,数据压缩和数据预测技术,节点轮流休眠等机制实现;第二种方法是通过设计出高效的基于能量优化的网络协议来提高网络的生命期,第三种方法是在网络中引入移动节点的方法,这种方法中又分为引入移动Agent来采集数据以较少通信的距离,以及在网络中引入移动sink节点使得网络的负载更加均衡以减少能量空洞现象。本论文针对无线传感器网络的能量效率问题展开深入的研究,提出多种方法来提高网络的能量效率,延长网络的生命期,主要创新性工作体现在以下几个方面：(1)提出了一种综合考虑sink位置、能量及链路状态的跨层路由协议LEACH-SC本文结合跨层设计的思想提出了一个新的基于sink位置和能量优化的分簇路由协议LEACH-SC(LEACH-selective cluster)协议。新协议首先针对LEACH协议簇头选举可能导致的分布不均衡等特点,设计了新的综合考虑网络密度、节点剩余能量以及链路状态等因素的簇头选举算法;同时改变成员节点加入簇的方法,让节点选择离sink与自己的“中心点”最近的簇头作为自己的通信簇头,这样使得网络的通信距离最小,降低全网的耗能；同时针对目前分簇算法中固定簇头数量的情况,提出了自适应计算簇头数量的算法。最后通过理论分析与仿真结果验证了LEACH-SC协议与LEACH协议、HEED协议相比,可降低网络的总能耗,平衡节点间的能量消耗,明显提高了传感器网络的生命期,而与中心控制的LEACH-C协议相比性能相当,但是LEACH-SC协议的开销远远小于LEACH-C。(2)提出了一种适用于传感器网络的基于退避策略的高效多跳分簇算法由于能量的有效性可以通过多跳(multi-hop)的通信来完成,并且多跳分簇可以有效降低“被迫簇头”的数量,因此本文提出将多跳技术与一种基于退避的分簇算法结合来组织无线传感器网络中的节点。该退避算法充分考虑了节点的剩余能量,并且这种自适应的退避机制可以很好实现节点间的负载均衡以及簇头分布的均匀性,显著的提高系统的生存时间,同时这个新的协议只带来很小的处理开销和通信开销。之后通过理论推导和分析证明了算法通过合理设置参数,簇头能够很均匀的分布在网络中,最后网络仿真验证了算法的能量有效性。此外为了适应大规模无线传感器网络的场景,我们又将该算法进一步扩展到分层多跳簇头(hierarchy of cluster heads)的情况下,通过理论分析可获得最优参数,并通过仿真发现通过分层的分簇可以实现在网络规模大且sink巨离监测区域较远时网络生命期的极大改善。(3)采用模拟退火算法对数据融合相关的移动代理路由问题给出了一种近似最优解。网络中引入数据融合相关的移动代理可以明显减少网络中传输的数据量,可有效的提高网络的生命期,而移动代理停留在各个点的顺序会决定整个网络的能量消耗,因而需要找到一条移动代理的最佳移动路线,使移动代理沿着整个路线收集数据所消耗能量最小。我们将多跳网络环境下的移动代理路由问题建模成一个顶点加权的游客问题,因而这个路由优化问题实际上是一个NP完全问题,考虑到无线传感器网络有限的计算能力和严格的能耗要求,最终使用模拟退火算法对这个问题给出一个近似的最优解,并通过仿真验证了算法的优越性。(4)提出了一种分区域多sink节点智能移动算法RSMA和相应的路由协议,能较好地解决网络的能量空洞问题。在传感器网络中引入移动sink节点可实现全网节点之间的能量均衡,提高网络的生命期。本文提出了一种分区域的多sink移动算法RSMA (Restricted Self-determination Mobility Algorithm)。RSMA算法通过对各个sink的移动范围进行有效控制以减少节点的平均路由跳数,选择目标位置时变为综合考虑节点的剩余能量、节点的密度等因素,这样得到的目标位置更加合理,并增加了紧急救助策略；同时为了进一步整体优化网络性能,我们又提出了一种适合于移动sink的能量均衡有效的路由协议EBER,该协议在建立路由树的时候考虑树的各个分支均衡分布,同时限制每个节点最大的转发次数,因而路由树上各个节点能耗较为均衡,通过这样的方式可以大大提高传感器网络的生命周期。最后通过搭建仿真平台对算法性能进行评估,仿真结果表明RSMA算法配合EBER路由协议可以在全网中实现能耗均衡,有效延长网络生命期。本课题的研究得到国家自然科学基金(60503021,60872018)、江苏省科技计划(BG2006027,BG2006039)以及新一代宽带无线移动通信网国家科技重大专项(2011ZX03005-004-03,2011ZX03005-005)等的资助。更多还原

【Abstract】 In recent years, wireless sensor networks(WSNs) are widely used in the military area, medical care, environment monitoring、remote control in the hazardous areas and so on. Since the sensor nodes are energy constrained, and after the nodes are deployed, they cannot be recharged. So the research work on how to increase the energy efficiency of wireless sensor networks is always a hot research area. In general, there are many ways to increase the network lifetime. One way is to decrease the total communication traffic by using the data aggregation, data compression, data prediction and duty cycling techniques. The second solution is to design effective network protocols to increase the network lifetime. The last way is to utilize mobility to improve network lifetime. We can also divide this method into two categories:one is to use mobile agent to collect data in order to conserve energy consumption, another way is to balance the energy consumption by introducing the mobile sink.In this paper we made thorough research work on how to increase the energy efficiency of wireless sensor networks, and we’ve proposed many method to increase the network lifetime, the main contributions of my paper are:(1) We have proposed a cross-layer clustering protocol-LEACH-SC which considers the sink’s position and energy.In this paper, we have proposed a novel cross-layer clustering protocol-LEACH-SC(LEACH-selective cluster)) protocol which considers the sink’s position and energy. LEACH-SC first solve the problem of uneven distribution of cluster-head of LEACH protocol. We design a new cluster-head election algorithm which considers the network density, energy and link quality etc. When cluster heads are elected, LEACH-SC, it changed the principle of choosing a communicating cluster head. The nodes choose the cluster head which is the most closest to center point between the node itself and the sink. Therefore the total distance will be shorter, and the total energy consumption will be conserved too. By analysis and through simulation, we found that compared with LEACH, LEACH-SC protocol can greatly reduce the overall network energy consumption, balancing the energy consumption among all the nodes in the network, extending the lifetime of the network. It significantly improved the network’s performance, with the good scalability and robustness.(2)An Energy Efficient Backoff Hierarchical Clustering Algorithms for Multi-hop Wireless Sensor Networks was proposed.As we all known, energy efficiency can be obtained by multi-hop communication., and multi-hop Clustering can decrease the number of "forced cluster-head". In this paper, we integrate the multi-hop technique with a backoff-based clustering algorithm to organize the sensors in wireless sensor networks. The adaptive backoff clustering algorithm not only considers the residual energy, but also realizes load balance among sensor node and achieves fairly uniform cluster head distribution across the network. So the new proposed algorithm is able to significantly prolong system life. At the same time the protocol incurs low overhead in terms of processing cycles and messages exchanged. Based on the energy model in a clustered sensor network, we proved that by choosing appropriate parameters, the volunteer cluster-heads can be evenly distributed in the working region. In addition, simulation results also demonstrate our algorithm is more energy-efficient than the classical ones. In order to meet the requirement of large-scale sensor networks, our algorithm is also extended to generate a hierarchy of cluster heads to obtain better network management and energy-efficiency. By analysis, we can get optimal parameters. And by simulations, we found out that the hierarchical networks have better performance than single-level clustering networks in terms of system life when the network scale is very big and the sink is located far away from the monitoring region.(3) We used the simulated annealing algorithm to solve the routing problem of mobile Agent which is responsible for data fusion.When mobile Agents which are responsible for data fusion are introduced into the wireless sensor networks, the system lifetime can be greatly increased because the communication distance is greatly decreased. But the order and number of nodes on the route traversed by a mobile agent will determine the energy consumption and hence it is desired to choose an optimal route to minimize the total energy consumption so that the system lifetime can be extended to the maximum. The routing problem of the mobile agent in multi-hop networks can be modeled as a traveling salesman problem. So this problem is also NP-complete problem. Taking account of the limited computation ability and limited energy of wireless sensor networks, we present an approximation algorithm which is based on simulated annealing to solve this NP-Complete problem. Simulation experiments demonstrate that the proposed heuristic is more energy-efficient.(4) A new mobile sink strategy called restricted self-determination mobility algorithm (restricted self-determination mobility algorithm, RSMA) is proposed to solve the network hole problem.By introducing the mobile sinks into the wireless sensor networks can balance the energy consumption and prolong the lifetime of the networks. In this paper, we come up with a new mobile sink strategy called restricted self-determination mobility algorithm(RSMA). The new algorithm RSMA control the moving range of all sinks, so the average routing hops can be reduced. When sink nodes decided the moving location, in order to find the more reasonable target location, the RSMA considers the energy, nodes’density etc,which can prolong the lifetime of the networks. Meanwhile, the emergency rescue method, which has been added into our new strategy is able to solve the energy hole problem. In order to further improve the performance of the networks, we propose a new effective routing protocol for wireless sensor networks with multiple mobile sinks which is called Energy balanced and efficient Routing(EBER) protocol. The protocol balance the energy consumption between every branch’s nodes when setting up the routing trees. At the same time, it constrained the maximum forwarding number of every nodes so that it can balanced the energy consumption among the whole tree. Therefore, the system lifetime of wireless sensor networks can be extended greatly. By setting up the simulation platform, we make some comparisons between the UPDA, RSMA and RSMA with the new routing protocols. Simulation results show that RSMA with the new routing protocols can reduce the overall network energy consumption, balancing the energy consumption among all the nodes in the network, prolong the lifetime of the network effectively.This paper’s work are supported by the National Natural Science Foundation of China(60503021,60872018), Jiangsu Province Science and Technology Project(BG2006027, BG2006039) and National S&T Dedicated Mega-Project2011ZX03005-004-03and2011ZX03005-005.更多还原