节点文献
WSN中流量自适应的跨层MAC协议研究
Research on Traffic Adaptive Cross-layer MAC Protocol in Wireless Sensor Networks
【作者】 刘方;
【导师】 王珂;
【作者基本信息】 中国矿业大学 , 计算机应用技术, 2014, 硕士
【摘要】 无线传感器网络研究中,MAC层处于网络协议的底层部分,控制信道分配,对网络性能有很大影响。针对无线传感器网络特点,MAC层的设计需要尽可能降低网络能耗,同时兼顾时延和吞吐量。另外,传统的严格分层设计方法不能很好的适应无线通信的特点,不能充分利用网络资源实现最优性能。近年来无线传感器网络的跨层优化设计成为新的研究方向,很多研究者提出利用层间信息交互设计通信协议。本文从无线传感器网络MAC层以及跨层优化角度出发进行了研究。首先概述了现有无线传感器网络路由协议和MAC协议,然后分析MAC层中相关技术并总结跨层设计的方法。针对SMAC协议缺点加以改进,提出了EA-MAC协议。该协议在原有算法基础上,添加节点的相关性分析算法和流量自适应的占空比机制,并在已有的流量自适应退避机制上引入剩余能量概念。节点相关性算法中,通过计算节点采集信息相关性,采用“分簇”机制,对节点划分区域,并在区域中选择代表节点传输本区域的数据。算法有效减少了竞争信道并传输重复数据的冗余节点。流量自适应占空比机制中,根据模型预测的流量值与设定阈值大小进行比较,动态改变占空比,减少节点空闲侦听。退避机制中,根据节点的剩余能量和预测流量值调整竞争窗口的大小以及退避时间的长短,从而在网络流量大时,非均匀分布的长时间退避可以避免数据碰撞的产生;剩余能量多的节点退避时间短,能够优先接入信道,均衡了全网的能量消耗,延长了网络的有效运作时间。仿真证明在流量动态变化的网络中,EA-MAC协议比SMAC协议有更好的吞吐量和时延。在EA-MAC协议的基础上,结合跨层优化设计,本文提出了基于能量感知的EACL协议。在路径建立时,综合考虑节点到汇点的通信代价期望值以及节点剩余能量建立路由表。节点发送数据时,网络层通过接口函数调用物理层参数并计算最优发送功率值。MAC层共享网络层的路由信息,通过在CTS帧添加下一跳节点字段,使得无关节点提前进入睡眠,避免空闲侦听能耗,减少数据冲突。仿真证明,网络流量变化时,EACL协议有更好的时延性能,在网络流量固定时,接收相同数目的数据包,EACL协议相对于EA-MAC协议需要较小的网络总能耗。
【Abstract】 In wireless sensor networks, MAC(Media Access Control)protocol, which isnear the underlying parts of network protocols and responsible for channel resourceallocation, is quite important for network performance. For the special feature ofwireless sensor networks, the designed MAC protocol should reduce the energyconsumption as much as possible. Meanwhile, delay and throughput should also betaken into consideration. On the other hand, traditional layered protocol can’t satisfythe demands of wireless communication, since it couldn’t fully utilize networkresource to achieve the best performance. Recently, the cross-layer optimizationdesign of WSN is becoming a hot research topic. Some researchers have proposed todesign network communication protocols through sharing interlayer information. Theresearch in this thesis is processed from the perspective of wireless sensor networksMAC protocol and cross-layer optimization.Firstly, existing wireless sensor network routing protocol and MAC protocol arepresented, the related technology in MAC protocol is analyzed and the cross-layerdesign approaches are summarized. EA-MAC protocol is proposed on the basis ofSMAC protocol to remedy the shortcomings of SMAC, in which nodes correlationanalysis algorithm and traffic adaptive duty cycle mechanism are added. The residualenergy is also introduced into the existing traffic adaptive bakeoff mechanism. In thenodes correlation algorithm, all network nodes are divided into several zones bycomputing nodes correlation of collected information and using “clustering”mechanism, and every zone chooses a node as a representative node to transmit data.This method can effectively reduce redundant nodes transmitting duplicate data. Intraffic adaptive duty cycle mechanism, duty cycle is regulated dynamically todecrease idle listening by comparing the threshold set before with the flow valueobtained from predict flow model. In backoff mechanisms, by adjusting the value ofcontention window and backoff time, the non-uniform distribution and long backofftime of nodes can avoid data collisions when network traffic is heavy. In addition,nodes with more remaining energy have priority to access the channel and shorterbackoff time, which can keep balance of the whole network energy consumption andlengthen network lifetime. Simulation results show that the EA-MAC protocol hasbetter throughput and shorter delay than SMAC protocol in network with dynamictraffic. Combined with cross-layer design, EACL protocol, which is based onenergy-aware, is proposed on the basis of EA-MAC protocol. When path isestablished, routing tables are built according to the expectation value ofcommunication cost between nodes and sink node and nodes residual energy. Whennodes start transmitting data, network layer calculates optimum transmit power valueby using interface function and calling the parameters of physical layer. With sharingthe routing information from network layer, MAC layer adds next-hop node field inCTS frame, therefore, insignificant nodes would go to sleep early to avoid idlelistening consumption and reduce data conclusion. Simulation results certify thatEACL has better delay performance when network traffic change dynamically, andhas less network consumption with the same numbers of received data packets whennetwork traffic keep fixed.
【Key words】 Wireless sensor networks; EA-MAC protocol; nodes remainingenergy; traffic adaptive mechanism; EACL algorithm; Cross-layer optimization;