|Speaker :||Uichin Lee|
|Time:||10:30 am - 11:30 am|
|Location:||LINCS Meeting Room 40|
Indoor localization has attracted muchÂ attentionrecently dueÂ to its potential for realizing indoorÂ location-awareapplicationÂ services. This talk considers aÂ time-criticalscenario with a team ofÂ soldiers or first respondersÂ conductingemergency mission operations inÂ a large building inÂ whichinfrastructure-based localization is notÂ feasible (e.g., dueÂ tomanagement/installation costs, power outage,Â terrorist attacks). In thisÂ talk, I will present a collaborative
indoor positioning scheme called CLIPSÂ that requires no preexistingÂ indoorinfrastructure. Assuming that each userÂ has a receivedÂ signalstrength map for the area in reference, an applicationÂ canÂ compare and select a set of feasible positions,when theÂ deviceÂ receives actual signal strength values at runtime. Then,Â deadÂ reckoning is performed to remove invalidcandidateÂ coordinatesÂ eventually leaving only the correct onewhich can be shared
amongst theÂ team. The evaluation resultsfrom an Android-based testbed showÂ thatÂ CLIPS converges toan accurate set of coordinates much faster
thanÂ existing non-collaborativeschemes (more than 50% improvementÂ underÂ theconsidered scenarios).
Biography: Dr. Uichin Lee is an assistantÂ professor in the DepartmentÂ of Knowledge Service Engineering at KAIST, Korea.Â He received hisÂ Ph.D. degree in Computer Science from UCLA in 2008. BeforeÂ joiningÂ KAIST, he worked for Alcatel-Lucent Bell Labs at Holmdel as aÂ member
of technical staff. His research interests includeÂ mobile/pervasiveÂ computing and social computing systems.