Dead Reckoning And RSSI Based Indoor Positioning System

With the explosion of information, processing massive data andproviding accurate information become critical. Associating information withposition tag is one of the most common ways. And with the popularity ofmobile devices and personal devices, the positioning system is reloaded as ahot topic. A positioning system enables device to determine its position, andmakes the position of the device available for position-based services such asnavigating, tracking or monitoring, etc. GPS-based positioning system isdedicated to provide out-door civil positioning. However, line of sight (LOS)transmission is required between receiver and satellite, which leads toinability to use these systems in indoor environment.Indoor Positioning System (IPS) has been designed to provide locationinformation of persons and devices. Accuracy and convergence time areconsidered as the particularly important factors for positioning techniques.Many technologies exist for indoor positioning. Navigation using inertialsensing has also been important. Dead Reckoning (DR), as one of the earliesttechniques used by sailors, can estimate one’s position by advancing a knownposition using course, speed, time and distance to be traveled. Given adetected step, its length and its direction, a person’s position can bedetermined by dead-reckoning. While DR is efficient for short-termpositioning, this method results in a positioning error accumulation over timeand distance. This paper introduces a Wireless Body and Environmental SensorTracking platform, for personal indoor positioning. The system is built withportable on-body sensor nodes and assisted sensor nodes deployed in thetargeted indoor area. It takes a hybrid approach with pedestrian deadreckoning and radio-based localization. Real-time inertial measurements arecombined with RSSI-based information, and then processed with an ExtendedKalman Filter to be weighted in the location estimation according to theirreliability. It also incorporates with an adaptive Step Length Algorithm toreduce the deviation of the measurements. The experimentation results showthat our system can improve the accuracy of the positioning by66.3%compared to pure inertial solution. With the popularity of wearable deviceswith inertial sensors and wireless communication chips, we believe that thisapproach is very promising for personal indoor positioning services.