Advances in Science, Technology and Engineering Systems Journal Vol. 7, No. 3, 129-138 (2022) www.astesj.com ASTES Journal ISSN: 2415-6698 Indoor Position and Movement Direction Estimation System Using DNN on BLE Beacon RSSI Fingerprints Kaito Echizenya 1 , Kazuhiro Kondo *,1 1 Graduate School of Science and Engineering, Yamagata University, Yonezawa, Yamagata 9928510, Japan ARTICLE INFO ABSTRACT Article history: Received: 28 March, 2022 Accepted: 05 Jun,e, 2022 Online: 24 June, 2022 Keywords: BLE beacons RSSI fingerprints indoor positioning deep learning In this paper, we propose a highly accurate indoor position and direction estimation system using a simple fully connected deep neural network (DNN) model on Bluetooth Low Energy (BLE) Received Signal Strength Indicators (RSSIs). Since the system’s ultimate goal is to function as an indoor navigation system, the system estimates the indoor position simultaneously as the direction of movement using BLE RSSI fingerprints recorded indoors. To identify the direction of movement along with the position, we decided to use multiple time instances of RSSI measurements and fed them to a fully-connected DNN. The DNN is configured to output the direction with the location simultaneously. RSSIs are known to be affected by various fluctuating factors in the environment and thus tend to vary widely. To achieve stable positioning, we examine and compare the effects of temporal interpolation and extrapolation as preprocessing of multiple RSSI sequences on the accuracy of the estimated coordinates and direction. We will also examine the number of beacons and their placement patterns required for satisfactory estimation accuracy. These experiments show that the RSSI preprocessing method optimum for practical use is interpolation and that the number and placement of beacons to be installed does affect the estimation accuracy significantly. We showed that there is a minimum number of beacons required to cover the room in which to detect the location if the estimation error is to be minimized, in terms of both location and direction of movement. We were able to achieve location estimation with an estimation error of about 0.33 m, and a movement estimation error of about 10 degrees in our experimental setup, which proves the feasibility of our proposed system. We believe this level of accuracy is one of the highest, even among methods that use RSSI fingerprints. 1 Introduction In recent years, almost everyone owns a smartphone. Some of the most often used applications on smartphones are those with location and navigation capabilities, such as Google maps. Global Posi- tioning System (GPS) signals received from satellites are used for location detection in many mapping and navigation applications. However, these applications cannot be used indoors since the recep- tion of GPS signals is generally poor due to the building structures. As an alternative to GPS, indoor positioning systems have been developed in recent years using various methods such as WiFi, In- door Messaging System (IMES), and other wireless LAN standards, as well as dedicated devices such as Light Detection and Ranging (LiDAR) [1–6]. However, there are many issues to be solved for the construction of the system, such as the availability of terminals supporting these standards, and the installation cost. Many of the indoor localization systems were based on WiFi. For example, [7, 8] use Deep Networks with WiFi RSSI fingerprints. Both use Autoencoders as the DNN. It is shown in [8] that an aver- age error of 1.21 m is possible in an apartment of 14.5 m by 4.5 m, but the number of Access Points (APs) is 59, which is considerably large for a room this size. [9] uses Channel State Information (CSI) fingerprints instead of RSSI fingerprints with an Autoencoder. CSIs provide receive levels of multiple subcarrier signals which can be collectively used to detect the receive level and direction more ac- curately than a simple RSSI. They were able to achieve an average error of 0.9425 m inside a living room of size 4 m by 7 m with just one AP. However, CSI is not available on all network interface cards. In this research, we focused on BLE beacons [10, 11]. BLE is one of the short-range wireless communication standards, and it is possible to conduct wireless communication at a lower cost and lower power consumption (often operating on battery for years without the need for replacement), and are easier to install and * Corresponding Author: Kazuhiro Kondo, 4-3-16 Jonan, Yonezawa, Yamagata 9928510, Japan, kkondo@yz.yamagata-u.ac.jp www.astesj.com https://dx.doi.org/10.25046/aj070315 129