Research Associate
* Developed apps and performed comprehensive tests to assess the indoor localization capability of location services on smartphones, i.e., Apple’s CoreLocation, Google’s Fused Location Provider, etc.* Designed and implemented the web portal (perfloc.nist.gov) for NIST’s PerfLoc(Performance Evaluation of Smartphone Indoor Localization Apps) Prize Competition, which allows participants to register and log in to their accounts, download reference datasets, upload location estimates, etc. The web portal instantly calculates and returns performance metrics corresponding to each upload and updates participants’ scores and rankings on the leaderboard.* Developed a novel Wi-Fi based lightweight indoor localization system in which Viterbi algorithm is applied. Location can be calculated in real time and location estimates at all earlier times along user’s trajectory can be dynamically corrected. Achieved an average horizontal localization error of 0.7 m in office environments.* Compared ranging performance of Wi-Fi Round Trip Time (RTT) and Ultra-wideband (UWB) through empirical results. In the light of COVID-19, prototyped a Bluetooth based exposure notification system on Raspberry Pi and evaluated its proximity detection performance by extensive tests. Explored ultrasonic ranging as an alternative to Bluetooth to provide better accuracy.
