Location, Time and Context in Systems: Rover - An Example

Опубликовано 7 сентября 2016, 16:39

In order to make computing devices more useful to most of us the next challenge we face is that of deriving and integrating location, accurate time and other context information in an easy to use framework. In this talk we present several technologies that have been developed at the University of Maryland to address these issues. For indoor and outdoor location determination we have developed three technologies, PinPoint, Horus, and Nuzzer. PinPoint uses the measurement of the time of flight to estimate the distance, without requiring synchronization of the clocks or using pair wise pinging, leading to a scalable, O(1) algorithm per node to localize N nodes. We have implemented and tested this approach using standard WiFi NICs and have obtained accuracies of the order of 5 feet while using 40 MHz clocks for timing. Horus uses the signal strength measurements and takes into account detailed characteristics of the signal resulting in accuracies of less than three feet indoors. Nuzzer technology also relies on the measurement of the signal strength but does not expect the person to be carrying any communication device, detecting the presence and location of a person as he/she enters a region in which WiFi is deployed. Accurate timing requires accurate clock synchronization. We have developed techniques which permit the clocks in large networks (of Internet scale) to be synchronized to the accuracies of sub-microsecond. We are integrating all these technologies in the framework of the Rover system which we will demonstrate during the presentation.