文档介绍：摘 要
融合了传感器、嵌入式计算、网络和无线通信四大技术的无线传感器网络(WSN)是一种全新的信息获取和处理技术,时间同步技术是 WSN 的一项关键支撑技术,WSN 的很多应用如 TDMA 调度、协同休眠、定位、数据融合等都需要网络中的节点时钟保持同步。由于传感器节点是能量受限的,降低能耗是设计 WSN 协议要考虑的最重要的目标之一。传输质量和网络寿命是衡量能量受限的多跳无线传感器网络性能的重要指标,合适的 MAC 协议可以降低能耗,延长网络寿命。
首先,本文深入的研究了传感器网络时间同步基本概念和现有经典时间同步机制,并分析和总结了经典同步机制的优缺点。经典时间同步机制可分为三类:基于接收者—接收者的同步机制如 RBS;基于发送者—接收者的同步机制如 TPSN、LTS;基于发送者的同步机制如 DMTS、FTSP。
其次,在现有同步机制的基础上提出一种新的基于分簇的低功耗多跳
传感器网络同步机制。该机制结合了单向广播同步机制和双向成对同步机制,在不引起同步滞后的前提下显著的减小了同步报文的开销。在分级网络结构基础上,提出一种分簇算法将 LCTS 扩展到多跳网络中,并对时钟漂移进行补偿,保证了一定的精度。对本机制进行了误差和能耗分析,在 Simsync 平台上进行了仿真实验,并与已有的典型算法进行了比较。
最后,基于跨层设计的思想,联合考虑拥塞控制、冲突控制和能耗控
制,用以求解传输层源端速率、MAC层链路访问概率和网络层节点寿命的最优值,来达到最大化网络效用和网络寿命的折中。通过变量代换将非凸和不可分离的原问题转化为基于对偶分解的凸优化问题。通过垂直和水平对偶分解,提出基于子梯度算法的完全分布式策略。通过仿真验证了算法的收敛性和有效性。
关键词无线传感器网络;时钟同步;低功耗;网络优化;拥塞控制;媒介访问控制;跨层设计
Abstract
Wireless works (WSN), which is integrated by sensor, puting and munication technologies, is a novel technology for acquiring and processing information. Time synchronization is a critical supporting technique of WSN, a work essential for TDMA scheduling, coordination dormant, localization, data fusion and many other applications of WSN. Transmission quality work lifetime is of great importance for the performance of energy-constrained multi-hop wireless works, in which transmission quality is measured work utility function work lifetime is constrained by the energy of the sensor node. Sensor nodes are energy constrained, reducing energy consumption is one of the most important conditions in design of WSN protocols. Proper MAC protocol can reduce energy consumption and work lifetime.
First, we carry out an extensive and thorough investigation on the basic time synchronization theory and the existing synchronization mechanisms for WSN. Advantages and disadvantages of the typical synchronization mechanisms are analyzed and summarized. Traditional time synchronization mechanisms are classified into three types: the receiver-receiver based mechanisms such as RBS; the sender-receiver based mecha