improving indoor localization using bluetooth low energy ... · • couchbase nosql database •...

35
Improving Indoor Localization Using Bluetooth Low Energy Beacons Rakib Shahriar

Upload: others

Post on 17-Oct-2020

1 views

Category:

Documents


0 download

TRANSCRIPT

Page 1: Improving Indoor Localization Using Bluetooth Low Energy ... · • Couchbase NoSQL database • User subscription using Google accounts. Test Site. The Campus Building. Notes on

Improving Indoor Localization Using BluetoothLow Energy Beacons

Rakib Shahriar

Page 2: Improving Indoor Localization Using Bluetooth Low Energy ... · • Couchbase NoSQL database • User subscription using Google accounts. Test Site. The Campus Building. Notes on

Objectives

Describes basic principles of a radio-based indoor localization.

Implemented a distributed system for collecting radio fingerprints by mobile devices with the Android operating system.

Bluetooth Low Energy (BLE) technology is used to improve indoor localization performance.

The localization of stationary objects based on WiFi, Bluetooth Low Energy, and their combination has been evaluated using the data measured during the experiment in a building.

Page 3: Improving Indoor Localization Using Bluetooth Low Energy ... · • Couchbase NoSQL database • User subscription using Google accounts. Test Site. The Campus Building. Notes on

Authors

• Pavel Kriz• Filip Maly• Tomas Kozel

• Faculties of Informatics and Management, University of Hradec Kralove

Page 4: Improving Indoor Localization Using Bluetooth Low Energy ... · • Couchbase NoSQL database • User subscription using Google accounts. Test Site. The Campus Building. Notes on

Problem statement

Locating objects outside is usually not a problem to locate a person or a mobile device.

Localization becomes extremely complicated inside buildings in high-density urban areas because of rare line-of-sight to the tracking systems (GPS, GLONASS, Galileo).

Traditional localization techniques are based on radio networks (WiFi) and fingerprints of signal strengths of WiFi devices.

Localization accuracy is influenced by a number of circumstances, for example, by characteristics of transmitters and receivers and characteristics of the environment which influence the radio signal propagation.

How to improve accuracy?

Page 5: Improving Indoor Localization Using Bluetooth Low Energy ... · • Couchbase NoSQL database • User subscription using Google accounts. Test Site. The Campus Building. Notes on

Solution

Bluetooth Low Energy (BLE) technology is used as an alternative supplementing WiFi access points.

Combination of BLE and WiFi access points provide more accurate localization.

Because of power consumption, where WiFi access points cannot be put, BLE enabled devices are installed to strengthen the coverage.

Page 6: Improving Indoor Localization Using Bluetooth Low Energy ... · • Couchbase NoSQL database • User subscription using Google accounts. Test Site. The Campus Building. Notes on

Basic localization approaches

Page 7: Improving Indoor Localization Using Bluetooth Low Energy ... · • Couchbase NoSQL database • User subscription using Google accounts. Test Site. The Campus Building. Notes on

Triangulation

• This is a object location detection technique• Methods based on triangulation can be further divided into lateration and

angulation• Method of estimations:

• Distance from several transmitters based on signal attenuation• Time characteristics of the signal propagation (TOA: Time Of Arrival, TDOA:

Time Difference of Arrival)• Direction of the received signal (AOA: Angle of Arrival)

• Works good for open spaces

Page 8: Improving Indoor Localization Using Bluetooth Low Energy ... · • Couchbase NoSQL database • User subscription using Google accounts. Test Site. The Campus Building. Notes on

Fingerprinting

• This method localizes objects by machine learning algorithms• Has two phases:

• Learning vectors of Received Signal Strength Indicator (RSSI) • Localization itself—the device to be localized measures the RSSI values and compares

them with the data in the fingerprint database using a suitable method• Few Methods of estimation:

• K-NN (Used in this work)• SVM• Neural Network

Page 9: Improving Indoor Localization Using Bluetooth Low Energy ... · • Couchbase NoSQL database • User subscription using Google accounts. Test Site. The Campus Building. Notes on

Bluetooth-Based Localization

• Nokia invented Bluetooth Low Energy (BLE) in 2010.

• Proximity estimation based on signal strength.

• Previous Bluetooth protocol was not feasible because of high cost and high power assumption, BLE changed the possibilities.

Page 10: Improving Indoor Localization Using Bluetooth Low Energy ... · • Couchbase NoSQL database • User subscription using Google accounts. Test Site. The Campus Building. Notes on

iBeacon Technology

Page 11: Improving Indoor Localization Using Bluetooth Low Energy ... · • Couchbase NoSQL database • User subscription using Google accounts. Test Site. The Campus Building. Notes on

How it works?

• iBeacon is a protocol developed by Apple.

• It uses the Bluetooth Low Energy standard.

• It broadcasts its identifier to nearby portable electronic devices.

• Can be run by a coin battery up to two years.

Page 12: Improving Indoor Localization Using Bluetooth Low Energy ... · • Couchbase NoSQL database • User subscription using Google accounts. Test Site. The Campus Building. Notes on

How utilized in this work?

• Used beacons are made by Estimote.

• Beacons transmit its identification data to BLE enabled Android smartphones/tablets.

• Advertisement data = MAC Address + UUID + Major Number + Minor Number

Page 13: Improving Indoor Localization Using Bluetooth Low Energy ... · • Couchbase NoSQL database • User subscription using Google accounts. Test Site. The Campus Building. Notes on

Methods and Architecture

Page 14: Improving Indoor Localization Using Bluetooth Low Energy ... · • Couchbase NoSQL database • User subscription using Google accounts. Test Site. The Campus Building. Notes on

Positioning Method

• Weighted 𝑘𝑘-Nearest Neighbors (𝑘𝑘-NN) in Signal Space method

• Compare untagged fingerprints with tagged fingerprints in the database using Euclidean distance

Where,m is vector for untagged fingerprints is vector for tagged fingerprintD is the distance

Where,P is the position of measured untagged fingerprint

Page 15: Improving Indoor Localization Using Bluetooth Low Energy ... · • Couchbase NoSQL database • User subscription using Google accounts. Test Site. The Campus Building. Notes on

System ArchitectureNotes:

• Data acquisition in JSON format

• Couchbase NoSQL database

• User subscription using Google accounts

Page 16: Improving Indoor Localization Using Bluetooth Low Energy ... · • Couchbase NoSQL database • User subscription using Google accounts. Test Site. The Campus Building. Notes on

Test Site

Page 17: Improving Indoor Localization Using Bluetooth Low Energy ... · • Couchbase NoSQL database • User subscription using Google accounts. Test Site. The Campus Building. Notes on

The Campus Building

Page 18: Improving Indoor Localization Using Bluetooth Low Energy ... · • Couchbase NoSQL database • User subscription using Google accounts. Test Site. The Campus Building. Notes on

Notes on Test Site• 52m × 43m area

• Several Cisco made WiFi transmitters

• 17 Estimote made BLE beacons

• Evenly placed in corridors and classrooms

• Putting behind dropped ceilings reduced localization performance

• Performance improved by putting beacons on the bottom side of the mineral fiber ceiling tile

Page 19: Improving Indoor Localization Using Bluetooth Low Energy ... · • Couchbase NoSQL database • User subscription using Google accounts. Test Site. The Campus Building. Notes on

Evaluation & Discussion

Page 20: Improving Indoor Localization Using Bluetooth Low Energy ... · • Couchbase NoSQL database • User subscription using Google accounts. Test Site. The Campus Building. Notes on

Comparison of localization accuracy

Page 21: Improving Indoor Localization Using Bluetooth Low Energy ... · • Couchbase NoSQL database • User subscription using Google accounts. Test Site. The Campus Building. Notes on

Localization accuracy

depending on scanning duration (scanning started

at time 0)

Page 22: Improving Indoor Localization Using Bluetooth Low Energy ... · • Couchbase NoSQL database • User subscription using Google accounts. Test Site. The Campus Building. Notes on

Which technique gives faster localization?

• BLE promises faster initial localization than WiFi does.

• This effect becomes even stronger in combination with WiFi.

Page 23: Improving Indoor Localization Using Bluetooth Low Energy ... · • Couchbase NoSQL database • User subscription using Google accounts. Test Site. The Campus Building. Notes on

For example, in the 2nd

second of the scanning, the authors were unable to

localize the mobile device in 168 positions using WiFi, in 36 positions using BLE, and in 20 positions using combination of BLE and

WiFi.

Page 24: Improving Indoor Localization Using Bluetooth Low Energy ... · • Couchbase NoSQL database • User subscription using Google accounts. Test Site. The Campus Building. Notes on

Localization accuracy depending

on scanning duration

(scanning started 4 s before time 0)

Page 25: Improving Indoor Localization Using Bluetooth Low Energy ... · • Couchbase NoSQL database • User subscription using Google accounts. Test Site. The Campus Building. Notes on

BLE Beacons Density

• Group A in red

• Group B in Green

Page 26: Improving Indoor Localization Using Bluetooth Low Energy ... · • Couchbase NoSQL database • User subscription using Google accounts. Test Site. The Campus Building. Notes on

Comparison of localization

accuracy among different BLE deployment configurations

Page 27: Improving Indoor Localization Using Bluetooth Low Energy ... · • Couchbase NoSQL database • User subscription using Google accounts. Test Site. The Campus Building. Notes on

Any abnormalities during experimentation?• Yes

• BLE beacons 13-17 were put completely hidden inside the wooden tables with metal sides.

• But still compared to the ceiling setup of BLE beacons, beacons 13-17 covered wider areas.

Page 28: Improving Indoor Localization Using Bluetooth Low Energy ... · • Couchbase NoSQL database • User subscription using Google accounts. Test Site. The Campus Building. Notes on

Localization results summary

Page 29: Improving Indoor Localization Using Bluetooth Low Energy ... · • Couchbase NoSQL database • User subscription using Google accounts. Test Site. The Campus Building. Notes on

Follow up work-1

• Time: 2017• Title: Optimization of Algorithms in Relation to iBeacon• Authors: Jan Budina, Martin Zmítko, Pavel Kříž• Institution: University of Hradec Kralove, Czech Republic

• Objective: A short paper on the description of algorithms in relation to the iBeacon. Implementation of kNN algorithm was described with mathematical details.

Page 30: Improving Indoor Localization Using Bluetooth Low Energy ... · • Couchbase NoSQL database • User subscription using Google accounts. Test Site. The Campus Building. Notes on

Follow up work-2

• Time: September 2017• Conference: ICCCI 2017 (Springer Link)• Title: Different Approaches to Indoor Localization Based on Bluetooth Low Energy

Beacons and Wi-Fi• Authors: Radek Bruha, Pavel Kriz• Institution: University of Hradec Kralove, Czech Republic

• Objective: Similar to this paper but they added a comparison of K-NN algorithm and modified Particle Filter Algorithms.

Page 31: Improving Indoor Localization Using Bluetooth Low Energy ... · • Couchbase NoSQL database • User subscription using Google accounts. Test Site. The Campus Building. Notes on

A newer technology

• Li-Fi is a visible light communications system that is capable of transmitting data at high speeds over the visible light spectrum, ultraviolet and infrared radiation.

• The term was first introduced by Harald Haas during a 2011 TEDGlobal talk in Edinburgh.

Page 32: Improving Indoor Localization Using Bluetooth Low Energy ... · • Couchbase NoSQL database • User subscription using Google accounts. Test Site. The Campus Building. Notes on

My Opinion

Page 33: Improving Indoor Localization Using Bluetooth Low Energy ... · • Couchbase NoSQL database • User subscription using Google accounts. Test Site. The Campus Building. Notes on

Positives

• The authors used few state-of-the-art technologies to detect object positions in an indoor environment.

• Use of Bluetooth Low Energy (BLE) allows multiple advantages • Improvement on localization accuracy• Improvement on human smartphone/tablet interaction

• Experimentation was shrewdly performed considering practical application challenges.

Page 34: Improving Indoor Localization Using Bluetooth Low Energy ... · • Couchbase NoSQL database • User subscription using Google accounts. Test Site. The Campus Building. Notes on

Negatives

• The authors presented object localization on corridors in a indoor environment but skipped denser areas like the rooms or closets.

• Sticking to only Android compatible devices may mislead the experimentation results. iOS based devices should have been used.

• Manuscript is well written, however, a few graphs are very hard to understand.

• Cost of installation should also have included in the discussion.

Page 35: Improving Indoor Localization Using Bluetooth Low Energy ... · • Couchbase NoSQL database • User subscription using Google accounts. Test Site. The Campus Building. Notes on

Questions?