基于可塑性干扰的无线传感器网络时间同步机制研究

In order to meet the application demand of large-scale and high density wireless sensor network, this paper is designed to exploit a time synchronization with constructive interference based flooding mechanism in wireless networks and meanwhile exploit constructive interference for fast network flooding and implicit time synchronization in wireless sensor network. According to the diversity of underlying protocol of wireless sensor network, a molding mechanism about constructive interference should be mainly discussed and be applied to a number of protocols. The fact that existing network flooding protocol can only be applied to solve the problem of sensor network within small scale makes it necessary for us to put forward a network flooding protocol based on the virtual grid skeleton and constructive interference, which can be widely used in a large-scale and high density wireless sensor network and increases the success rate of flooding packet transmission. In order to solve the problem of low coupling, high consumption and accuracy difference between network flooding and time synchronization, we propose a time synchronization with constructive interference based flooding protocol and through the protocol, the convergence speed and accuracy of time synchronization will be significantly improved. Moreover, the power dissipation will be reduced at the same time. Generally speaking, the project will lend a great help to the development of large scale and high density wireless sensor networks and to the improvement in the core competitiveness of the telecommunication industry in the world.

根据大规模、高密度无线传感器网络应用需求，研究一种基于可塑性干扰的洪泛时间同步机制，并发完成无线传感网络中的快速洪泛与时间同步。根据无线传感器网络底层协议的多样性，研究可塑性干扰的成型机制，构建面向多协议的可塑性干扰模型；考虑到现有网络洪泛协议只适用于小规模传感器网络的问题，研究一种基于虚拟网格骨架及可塑性干扰的网络洪泛协议，能够适用于大规模，高密度部署的无线传感器网络，提高洪泛数据包接收率；针对实际无线传感网应用中网络洪泛与时间同步耦合度低、耗能高、精度差的问题，研究一种基于可塑性干扰的洪泛时间同步协议，提高时间同步精度与收敛速度，降低功耗。本项目的研究成果有利于大规模、高密度无线传感器网络的应用发展，提高我国在国际无线传感器网络领域的核心竞争力。

本项目依据研究目标总结了无线传感器网络时间同步的国内外研究现状和面临的挑战。根据无线传感器网络实现全网时间同步时传感器节点组织方式的不同，分别对结构化时间同步和非结构化时间同步两类时间同步协议进行了研究，并结合统计信号处理技术研究了时钟偏差的估计和补偿方法。由于复杂的传感器网络环境导致网络延迟的不确定性问题，研究了任意网络延迟分布模型下的时钟参数估计方法，提出一种基于变分贝叶斯的复合粒子滤波方法和一种基于序贯置信传播的分布式时间同步方法，提高时钟参数的估计精度；针对因无线信道不稳定引起的数据包丢失的问题，提出一种健壮的时间同步方法，提高时间同步方法的健壮性；针对洪泛时间同步方法可扩展性差的问题，提出一种基于适应值跟踪的稳定时间同步方法，提高洪泛时间同步方法的可扩展性，避免基于可塑性干扰的洪泛时间同步需要高配置硬件的需求。