Multi-sensors Systems for Ubiquitous Personal Navigation

导航技术正在以惊人的速度发展，特别是在民用社区，车辆和个人导航器等设备在我们的日常生活中变得司空见惯。虽然车载和机载导航是一项成熟的服务，但通过智能手机的个人导航系统(PNS)才刚刚开始获得势头，这是由于应用市场的激增、硬件的发展和消费者的采用。PNS支持基于移动位置的服务(LBS)，这些服务变得无处不在，并嵌入到社交网络、资源跟踪和人员定位等许多应用中。基于全球定位系统(GPS)的导航设备是这些应用的主要解决方案。然而，GPS信号并不总是可用的，这对密集的城市峡谷、树冠下和室内深处等GPS信号不可用的环境提出了巨大的挑战。微型机电系统(MEMS)技术的最新进展与电子产品的小型化相结合，使得在大众消费市场的新一代导航设备中部署多个传感器成为可能。例如，在许多智能手机中，嵌入了廉价的基于MEMS的惯性导航传感器、磁力计和气压计，以支持游戏和娱乐应用。然而，在成本和规模方面的收益是以导航精度为代价的。MEMS传感器的误差特性比其更高级别的传感器大得多，这意味着集成策略对这些传感器的成功和精度至关重要。此外，人类的移动使个人导航不同于其他导航平台(如汽车、火车、轮船等)，具有挑战性：个人导航意味着速度和方向的频繁变化，位置不能像汽车和火车那样被限制在预先定义的轨道上。根据定义，移动设备处于难以预测的位置；各种应用模式根据用户的活动(如行走、驾驶、跑步等)频繁变化。本研究的目标是应对和解决上述挑战，并为当前存在于下一代智能手机中的泛在、连续、可靠和无缝的使用传感器的个人导航技术提供一种多传感器辅助的解决方案。

Navigation technologies are expanding at a phenomenal rate, especially within the civil community, with devices such as vehicle and personal navigators becoming commonplace within our everyday lives. While vehicle and airborne navigation is a well-established service, personal navigation system (PNS) through smart phones has just started to gain momentum fuelled by the application marketplace proliferation, hardware evolution, and consumer adoption. PNS enables mobile location based services (LBS) which become pervasive and are embedded in many applications such as social networking, resource tracking and positioning people . Global Positioning System (GPS) based navigation devices are the dominant solution for these applications. However, GPS signals are not always available and this poses a great deal of challenge for GPS signal-denied environments such as dense urban canyons, under tree canopy and deep indoors. Recent advances in Micro-Electro-Mechanical Systems (MEMS) technology combined with the miniaturization of electronics have empowered the deployment of multiple sensors into a new generation of navigation devices in the mass consumer market. In many smart phones, for example, inexpensive MEMS-based inertial navigation sensors, magnetometers, and barometers are embedded to enable gaming and entertainment applications. Nevertheless, the gains in cost and size come with a price in terms of navigation accuracy. MEMS sensors have much larger error characteristics than their higher grade counterparts meaning that the integration strategy is paramount to the success and accuracy of these sensors. Furthermore, human movement makes personal navigation different and challenging from other navigation platforms (such as for cars, trains, boats, etc.): Personal navigation implies frequent changes of speed and orientation and position cannot be constraint to predefined tracks as for cars and trains. Mobile devices, by definition, are handheld in a position that is difficult to be predicted; various application modes alter frequently depending on user's activities such as walking, driving, running etc. The goal of this research is to address and solve the above challenges, and provide a multi-sensor assisted solution for ubiquitous, continuous, reliable, and seamless techniques for Personal Navigation using sensors currently exist in next generation smart phones.