路图无关的异构移动网络导航理论与关键技术研究

With the rapid development of wireless communication technology and mobile computing technology, heterogeneous mobile networks, composed of cellular networks, wireless local-area networks, wireless metropolitan-area networks, wireless personal area networks, have drawn people's great attention in the past several years. Due to the close interaction between the physical world and the digital space, the network navigation technology has been widely used in a variety of crucial fields, covering military, transportation, logistic, health care, people's livelihood, etc. Existing navigation techniques, including satellite-based navigation, network-based navigation, and context-aware navigation, rely on a prerequisite that the road map information is always available to guarantee the success of the navigation. This proposed project, however, aims to reform the traditional work mode of extracting the navigation trajectory (routing path) necessarily from a series of points (localization) and surfaces (road map) and designs a set of new solutions. We first theoretically explore the necessary conditions and constraints for navigability and reachability without the road map information in the network. Based on our theoretical study, we will further design efficient navigation protocols. In particular, we will target at three challenging issues: 1) localization without the positions of access points, 2) matching of the logical space and physical space without the floor plane (indoor) / the traffic plane (outdoor), and 3) dynamic target discovery with the multi-user cooperation. This project will take full advantage of the mobility nature of off-the-shelf mobile terminals (e.g., mobile phones, PDAs, and laptops) and the low-power sensors associated with those mobile devices. In addition, this project will also completely utilize the correlation among the spatial, temporal, and social domains of mobile users to eventually provide a pervasive, low-power, reliable navigation service with the existing deployment of heterogeneous mobile networks.

随着无线通信技术和移动计算技术的快速发展，由蜂窝网络、无线局域网、无线城域网、无线个域网构成的异构移动网络被广泛关注。物理世界与信息世界的不断融合，使得网络导航技术在军事、交通、物流、医疗、救援、民生等方面具有广泛需求。现有导航技术(包括卫星导航、网络导航、感知导航)均将路图信息作为导航成功的前提条件，本项目拟突破传统导航系统设计中由点（定位）、面（路图）获取线（路径）的固有模式，在理论上探索路图无关的网络导航可导性与可达性判定及约束条件，在技术上从接入点位置未知定位、无需建筑物平面图(室内)/道路交通图(室外)的逻辑空间与物理空间匹配、多用户协同的动态目标发现与导引三方面入手，充分利用常规移动终端（移动电话、PDA、平板电脑）和其配备的低功耗传感器，充分挖掘多用户移动在空间、时间、社交等方面的相关性，在现有广泛部署的异构移动网络条件下，提供比现有导航技术更为普适、低耗、可靠的导航服务。

广泛部署的无线网络为室内定位和导航提供了基础设施。现有工作从测距和非测距两个方面研究室内定位和导航的基础理论和关键技术。围绕三个方面展开科学研究：弱约束的异构网络导航和定位方法、基于射频信号的室内平面图自动生成和导航和基于射频信号特征分析的移动个体及其行为识别。提出提出无勘测室内定位方法--WILL，该方法无需传统基于指纹室内定位方法的离线指纹数据库生成，通过在线的指纹数据库和室内平面图数据库的构造，形成虚拟房间定位和移动个体的位置估计。提出基于环境动态性感知的无线传感器网络导航原理和技术--OPEN，该方法通过传感器网络感知环境动态性定期更新导航路径队列，以期最小化用户在局部区域移动时的振荡行为。提出利用物理层信息的无设备人员监测的方法。无设备被动(DfP)的人体检测在基于位置服务像智能空间、人机交互以及财产安全中扮演着关键角色。在典型的室内场景中评估了提出方案，实验表明基于指纹的人体检测模型在4个方向的结果为平均8％的假阳率和平均7％的假阴率，基于阈值的人体检测模型平均假阳率和假阴率均在10%左右。