配布型BLEタグとタンデムスキャナを用いた屋内位置推定手法

Bibliographic Information

Other Title
  • A Location Estimation Method using Mobile BLE Tags with Tandem Scanners

Search this article

Abstract

物流や人流の管理において,動線の可視化や移動の効率化,ナビゲーションの提供を行うためには屋内位置推定が必要となる.精度やコストが異なる様々な手法が提案されているが,特にイベント来場者の行動分析では,準備期間が短いため事前計測が必要な手法や,多数の機材を設置する手法の適用は難しい.本論文では,少ない事前準備で高精度な推定を実現する,配布型Bluetooth Low Energy(BLE)タグを利用した屋内位置推定手法を提案する.提案手法はBLEパケットを収集するスキャナを環境固定とし,BLEタグを来場者が持ち運ぶ形式としてタグの配布のみで推定を利用可能とした.位置推定は事前計測を必要としない3点測位とパーティクルフィルタを組み合わせたアルゴリズムを用い,変更可能な複数のパラメータを導入してグリッドサーチにより精度を向上させられるようにした.また,パケットロスの発生による精度の低下を防ぐため,複数のBluetoothアダプタを1台に備えるタンデムBLEスキャナを開発した.日本科学未来館で開催されたG空間EXPO2016においてデータ収集実験を行い,精度を検証したところ平均誤差3.6メートルを達成した.

In this paper we propose an indoor location estimation method using mobile Bluetooth Low Energy (BLE) tag, which can achieve estimation higer accuracy with less labor. To track the activities of people indoors, such as trajectory of walking or interest, location estimation technology is essential. To activate the estimation system in short preparatory phase, the methods which need fingerprinting or many devices is hard to use. To use for the survey of the activities of the visitors at a large-scale exhibition, proposed method is designed to use BLE packet scanners which are fixed to the target area and mobile BLE tags which are carried by the visitors. Estimation algorithm is based on the combination of trilateration and particle filter to avoid signal strength investigation, and has several parameters to control the algorithm's behavior. We develop tandem BLE scanner which has multiple Bluetooth adapters on one scanner to mitigate the effect of frequent packet loss. We have executed data collection experiment at G-EXPO2016 in Miraikan museum, Tokyo, Japan. The mean estimation error of the proposed method is around 3.6 meters.

Journal

Related Projects

See more

Details 詳細情報について

Report a problem

Back to top