Accurate Indoor Localization with Zero Start-up Cost

Recent years have witnessed a major interest in developing accurate RF-based localization systems and have seen the advent of systems that demonstrate tens of centimeters of accuracy. However, such systems require either deployment of new infrastructure, or extensive fingerprinting of the environment through training or crowdsourcing, impeding their wide-scale adoption. In Ubicarse, the Inventors address this problem by developing a new formulation of SAR that is suitable for hand-held devices, and emulates a large antenna array on the device. This allows devices to accurately identify the spatial direction of incoming RF signals and localize themselves relative to the locations of neighboring Wi-Fi access points without requiring fingerprinting or modified infrastructure.
Ubicarse uses a new, translation-resilient formulation of SAR that allows the user to twist his/her mobile device about its vertical axis. They use rotation estimate from the gyroscope and exploit the MIMO capability of modern wireless cards to develop a new SAR formulation that operates on relative wireless channels to estimate the direction of the access point’s signal, purely based on the device’s orientation with no information required on its exact position or translation. To achieve high-precision geotagging, Ubicarse exploits the synergy between its WiFi-based localization capability and stereo vision algorithms used for 3-D reconstruction. Specifically, it combines vision algorithms that accurately localize an object with respect to the set of camera coordinates from which the object was imaged, with SAR localization, that identifies the global positions of the device’s camera accurately. The combined method allows Ubicarse to localize an object to within tens of centimeters in global coordinates – a promising step towards a future where users can tag or search for objects in the physical world using only mobile devices, much as information is currently searched for on the web.(O:P)
Ubicarse not limited to localizing RF devices; it combines RF localization with stereo-vision algorithms to localize common objects with no RF source attached to them - Ubicarse displays localization accuracy of the mobile device within few tens of centimeters without specialized infrastructure or fingerprinting
Dina Katabi
Lab Representatives
Share to Facebook Share to Twitter Share to Google Plus Share to Linkedin