A Magnetic Fingerprinting-based Precise Automobile Localization using Smartphones

Precise automobile position tracking is important in terms of safety. The state-of-the-art navigation frequently fails to provide precise outdoor localization especially in urban areas due to GPS errors. To track vehicle position, conventional approaches have used particle filter algorithm containing estimated vehicle position and weight values. Our approach is to provide a precise magnetic fingerprinting-based outdoor localization. Our intuition is that many landmarks can improve accuracy of tracking algorithm, and we propose a precise magnetic fingerprint map for many landmarks. We preliminary tested magnetic fingerprinting features on the road in our university. To provide feasible and reliable magnetic fingerprinting features, we will further investigate the features in diverse conditions such as in-vehicle magnetic field, magnetic feature variation according to lateral vehicle positions, external magnetic effects (e.g., parked car, electric poles, traffic lights or iron fence), time variances and differences between different smartphones.

sh2

People:

  • Sanghoon Jeon (topjsh0331@dgist.ac.kr)
  • Haengju Lee (haengjulee@hanyang.ac.kr )
  • Sang Hyuk Son (son@dgist.ac.kr)

Reference:

Sanghoon Jeon, Haengju Lee, and Sang Hyuk Son, “WIP Abstract: A Magnetic Fingerprinting-based Precise Automobile Localization using Smartphones”, IEEE International Conference on Embedded and Real-Time Computing Systems and Applications (RTCSA), Daegu, Korea, Aug 2016.