基于蒙特卡罗方法的水下传感网移动节点定位方法研究

In underwater acoustic sensor networks, unknown nodes and object localization technology is an important support technology. Node location information is the basis of node sensing and data transmission and other network functions. Due to the different characteristics of the underwater sensor networks, existing terrestrial sensor network localization algorithm can not be directly applied to underwater. The project mainly studies the localization of mobile sensor networks based on the Monte Carlo method, the concrete content is as follows: First, starting from the sensing localization information processing, wavelet transform is used to process the underwater ambient noise filter. Through the integration of multiple dimensions of the sensing localization information, a data optimization pool is built to provide better support for underwater sensor network positioning. Second, for underwater objects have to move with the flow characteristics, the project intends to use gray prediction method to propose the mobile sensor network transmission model and the mobile node localization algorithm based on motion prediction. The basic theory of gray forecast in locating a mobile underwater acoustic sensor networks is also considered in this research. Third, in the environment of the anchor nodes are sparse underwater deployment, the virtual anchor nodes and invisible anchor nodes are used to establish location model for the sparse anchor node deployment scenarios and solve the positioning problem of blind area. The research of this project will provide new ideas and new algorithms for underwater acoustic sensor network node localization.

在水下传感器网络中，未知节点的定位是一个重要的支撑技术，节点位置信息是节点感知和数据传递等网络功能的基础。由于水下传感器网的不同特点，现有陆地传感器网络定位算法不能直接应用于水下。本项目主要研究基于蒙特卡罗的水下传感器网移动节点定位算法，内容如下：一、从传感定位信息处理入手，利用小波变换对节点采集的位置信息所遭受的水下环境噪声做滤除处理，通过融合多维度定位数据信息，构建一个数据优化池，为水下传感器网定位提供更好的支撑。二、针对水下物体随水流具有移动的特性，拟采用灰度预测的方法，建立基于运动预测的水下移动传感网传输模型，解决因节点移动影响水下定位精度的问题，同时研究灰度预测在移动水声传感网定位中的基础理论。三、在水下环境锚节点部署较为稀疏时，项目利用虚拟锚节点和隐形锚节点等手段，建立适应于锚节点稀疏部署场景的定位模型，解决定位盲区问题。本项目的研究将为水下传感器网节点定位提供新思路和新算法。

在水下传感器网络中，未知节点的定位是一个重要的支撑技术，节点位置信息是节点感知和数据传递等网络功能的基础。本项目研究了水声定位系统中，声速非沿直线传播和声速非固定不变对定位精度的影响，水下节点处于运动状态如何能够实现高精度定位问题；水下三维入侵检测场景中三维k-栅栏覆盖构建的问题；水下环境引起的多径效应和多普勒频移导致误码率高的问题；复杂环境下，移动节点遇到障碍物、信号不稳定等问题时，提出了高精度定位模型。针对视频资源的视频描述技术，考虑视频序列的时序信息、运动信息等，提出了多模态视频描述模型和分级的LSTM架构。本项目的研究将为水下传感器网节点定位提供新思路和新算法。