Locationoid: An Accurate Energy-Efficient Location Provider for the Android Platform

Google Tech Talk
September 20, 2010

Presented by Prof. Moustafa Youssef.

ABSTRACT

In this talk, we present our recent results on using the internal sensors available on the Android platform for enhancing both the accuracy and energy-efficiency of the Android location provider. In the first part of the talk, we present CellSense, a probabilistic RSSI-based fingerprinting location determination system for GSM phones. We discuss the challenges of implementing a probabilistic fingerprinting localization technique in GSM networks and present the details of the CellSense system and how it addresses the challenges. Results for two different testbeds, representing urban and rural environments, show that CellSense provides at least 23.8% enhancement in accuracy in rural areas and at least 86.4% in urban areas compared to other RSSI-based GSM localization systems. This comes with a minimal increase in computational requirements. In the second part of the talk, introduce a low-energy calibration-free localization scheme based on the available internal sensors in many of today's phones. We start by energy profiling the different sensors that can be used for localization. Based on that, we propose GAC: a hybrid GPS/accelerometer/compass scheme that depends mainly on using the low-energy accelerometer and compass sensors and uses the GPS infrequently for synchronization. We implemented our system on Android-enabled cell phones and evaluated it in both highways and intra-city driving environments. Our results show that the proposed hybrid scheme has an exponential saving in energy, with a linear loss in accuracy compared to the GPS accuracy.

Bio: Moustafa Youssef is an Assistant Professor at the Department of Computer Science and Engineering, Egypt-Japan University of Science and Technology (E-JUST). Dr. Youssef received his Ph.D. degree in computer science from University of Maryland, USA in 2004 and a B.Sc. and M.Sc. in computer science from Alexandria University, Egypt in 1997 and 1999 respectively. His research interests include mobile and wireless networks, location determination technologies, pervasive computing, and sensor networks. Dr. Youssef is a recipient of the of the 2003 University of Maryland Invention of the Year award for his Horus location determination technology. He is also an elected member for the honor society Phi Kappa Phi and a life member for the Egyptian Society for Talented. Dr. Youssef has more than 60 publications in the top networking journals and conferences with thousands of citations on Google scholar. He is also the author of four book chapters. He is the recipient of many awards for his papers including the spotlight paper of the December 2009 issue of the IEEE Transactions on Parallel and Distributed Systems, Second place winner, ACM SIGMobile undergraduate student research competition, September 2009, Most popular challenge paper award, the Thirteenth ACM Annual International Conference on Mobile Computing and Networking (Mobicom'07), September 2007, Best paper runner-up, The 2006 IEEE GLOBECOM Conference, among others.

Category:

Science & Technology

Tags:

• google tech talk
• android
• mobile
• localization
• gps

[hebelken]

These tech talks are pretty cool. It's a shame that all of them have horrible audio.

[porscheghcje]

Energy sources without the need for fuel or energy input exist ,But the Oil companies want these technologies unknown to the masses,Get the blueprints for a free energy motor at LT-MAGNET-MOTORdotCOM ,Free yourself!

[mclaarson]

Hm and what if you are on foot? In a car you might have a gps anyways.

[senoctar]

@Arachidon99 I don't think he ever said that. He said it would be better in terms of power consumption, not accuracy. The GPS was always used as the reference point.

[Arachidon99]

Is he actually proposing a system which would be more accurate than GPS, yet he wants to use GPS to determine the location of the spots where he is collecting an RF fingerprint? Tricky.. It might be possible using statistics though. But I'm not very confident, probably takes way to long to set this up.