【作者】 刘斌；

【导师】 梁凤梅；

【作者基本信息】 太原理工大学 ， 电路与系统， 2011， 硕士

【摘要】 跨层设计是最近几年通信网络领域研究的热点之一。跨层设计是相对于传统的分层设计思想而言的。传统的分层设计思想在有线网络设计中的应用非常成功。但是在无线网络中,传统的分层设计思想有一定的局限性。例如网络层的多跳路由算法较少地考虑到物理层和链路层的影响,基于分层设计的协议栈无法对有限的无线资源和能量资源充分利用。在无线传感器网络中,系统分层相对简单,若能采用跨层设计思想,将系统看成一个整体来设计,会提升传感器网络的性能。跨层设计可将底层的信息通过有效途径反馈到上层,上层可以根据底层的信息做出相应决策,进而优化上层通讯协议。跨层设计并非完全否定传统的分层设计理念,而是将网络各层的信息统一调度和共享,以更高效的方式分配和利用有限的资源。基于跨层设计思想的优化技术,在原有分层协议栈的基础上添加跨层模块,使相互独立的网络各层之间的信息有效沟通。美国芯科实验室在2010年正式推出EZMAC/EZHOP协议栈,这种协议栈主要规定了系统在媒体接入控制层的功能。本文对EZMAC协议栈进行了跨层优化。主要完成了以下改进工作：竞争与时分复用机制的融合、单一系统双波特率、自动选频机制、包头滤波器和信道错误计数器。这些优化改进可使系统底层信息有效反馈至上层协议栈,为上层决策做出参考。通过实验验证,跨层优化技术对系统性能的提升作用。时钟同步是传感器网络的重要支撑技术之一。本文分析了造成传感器网络时钟不同步的原因,提出一种基于跨层优化技术的时钟同步算法。本算法考虑到物理层不可避免的因素——晶体振荡偏差,并且利用软件对此偏差进行了补偿修正。然后在实验中使用仪器测量两层节点的实际时间偏差,验证了本算法的实用性本文最后将跨层优化技术应用到无线粮情测控系统的设计中。设计了以无线微控制器Si1001为核心的传感器节点。根据现场应用要求,将跨层优化的EZMAC协议栈移植到网络中不同类型的节点。在协议栈移植的过程中,通过软硬件联合调试解决了一系列实际问题。基于跨层优化技术的无线粮情测控系统最终应用于中央储备粮淮安直属库中。本文研究了跨层优化技术并将其应用于无线传感器网络的实际项目中。利用跨层优化思想改进了EZMAC办议栈,提出了基于跨层优化技术的时钟同步算法及其实现。通过对无线粮情测控系统的硬件节点设计和软件协议栈的移植,验证了跨层优化技术在传感器网络中的优势。更多还原

【Abstract】 The cross layer design is one of the hot research issues in the field of communication network for the past few years. The traditional separate layer design method did really a good job in cable networks. However in wireless communication it has a great deal of limitations. For instance the multi hop route algorithm in network layer has little considerations of the impact from the physical and media access control layer; adequate utilization to the limited wireless and power resources are not able to be implemented by the separate layer stack.There is a simple network structure of wireless sensor network. If the whole system can be seen as an integral structure with the cross layer method, the performance of sensor network can be improved. Cross layer design builds an effective tunnel which helps the info exchange between low layer and high one. The upper layer will take actions to optimize the system performance by what it receives from the low layer. The separate layer design certainly will not be denied by the cross layer method but improved in the way of info management and sharing offered by cross layer method. The limited resources can be allocated in a more effective way in order to upgrade the network performance.The cross layer optimization technology is originated from cross layer design. This kind of tech adds a cross layer module into the traditional separate layer stack, which largely improves the communication ability between different network levels. The EZMAC/EZHOP stack has been brought out by Silicon Laboratory in 2010 which is simply a MAC protocol. Cross layer optimization to EZMAC is presented in this paper such as the integration of competition and TDMA, one system two baud rate, auto frequency selection, packet filter and channel error counter. This optimization helps a lot to smooth info transferring tunnel and provide reference to upper layer stack. Experiments have been made to demonstrate the priority of cross layer consideration.Time synchronization is one of the vital support technologies in sensor network. This paper describes a typical time synchronization algorithm based on cross layer consideration and presents the cause, crystal deviation, to the asynchronous phenomenon in sensor network. The crystal deviation has been corrected by the cross-layer MAC protocol. The experiment which is implemented by the oscilloscope demonstrates the practicability of this algorithm.The application to cross layer optimization is carried out in the wireless grain monitoring system. The hardware platform with Si1001 wireless MCU is designed and the cross-layer EZMAC protocol is explanted into all kinds of nodes in the network. Lots of trouble issues have been solved during the development process. Finally the wireless grain monitoring system has been deployed in a grain depot which belongs to the central government.In this paper, research efforts have been devoted to cross layer optimization technology, and a real WSN project based on this tech has been implemented. EZMAC protocol and time synchronization algorithm which are improved by this technology are introduced too. The implementation of hardware and software in wireless grain monitoring system illustrates the priority of cross layer optimization in wireless sensor network.更多还原