東華大學圖書館 |

Data-Driven Indoor Mobility Analyses, Modeling, and Encounter Classification for IoT Applications.

紀錄類型:	書目-電子資源 : Monograph/item
正題名/作者:	Data-Driven Indoor Mobility Analyses, Modeling, and Encounter Classification for IoT Applications./
作者:	Al Qathrady, Mimonah.
出版者:	Ann Arbor : ProQuest Dissertations & Theses, : 2020,
面頁冊數:	161 p.
附註:	Source: Dissertations Abstracts International, Volume: 82-06, Section: B.
Contained By:	Dissertations Abstracts International82-06B.
標題:	Computer engineering. -
電子資源:	http://pqdd.sinica.edu.tw/twdaoapp/servlet/advanced?query=27837884
ISBN:	9798698567158

Data-Driven Indoor Mobility Analyses, Modeling, and Encounter Classification for IoT Applications.
Al Qathrady, Mimonah.

Data-Driven Indoor Mobility Analyses, Modeling, and Encounter Classification for IoT Applications. - Ann Arbor : ProQuest Dissertations & Theses, 2020 - 161 p.

Source: Dissertations Abstracts International, Volume: 82-06, Section: B.

Thesis (Ph.D.)--University of Florida, 2020.

This item must not be sold to any third party vendors.

Data-driven indoor mobility models and encounter distance classification are essential elements in the design and evaluation of numerous Internet of things (IoT) indoor applications. One critical example of such applications is infection tracing in smart hospitals, also referred to as 'contact tracing' in epidemiology. Infection tracing needs a realistic characterization of encounters, such as proximity classification and distance estimation between nodes. Most importantly, the simulation and evaluation of the tracing systems and many other indoor IoT applications require a realistic indoor mobility model. Consequently, this dissertation comprises three main parts: 1- infection tracing in the intelligent hospital (i-hospital), 2- encounter distance estimation and classification, and 3- data-driven indoor mobility analyses and modeling for location-centric and user-centric metrics. The first part of this work describes the i-hospital project, including infection tracing approaches and challenges. The second part is about the encounter distances and proximity classification. Many existing encounter-based solutions rely on wireless data, such as access point associations or Bluetooth scanning records. However, the wireless encounter records reflect when two nodes are within sensing or communication range from each other, and not within the problem range of interest, such as infection range as in the infection tracing problems. In this part, the Bluetooth Low Energy (BLE) is targeted as a technology since it is built specifically to facilitate IoT. The BLE transmission power (TX power) is integrated with a Received Signal Strength Indicator (RSSI) into the state-of-art parametric and machine learning models to improve the distance estimation and proximity classification models. In addition, a new promising classifier-based distance estimation model architecture is proposed and evaluated extensively to show its effectiveness. Furthermore, we have built a library that contains the TX power and RSSI data of BLE encounters within defined distances of up to 22 meters in different indoor environments. Understanding and modeling nodes' mobility patterns are vital for future IoT and mobile services design and performance evaluation.Consequently, the third (and most comprehensive) part of this dissertation focuses on data-driven indoor mobility analyses and modeling. It introduces a detailed data analyses methodology framework that classifies analyses into two perspectives: user-centric and location-centric. In this work, real wireless traces at the building level have been studied from individual and pairwise dimensions from each perspective to provide guidelines for realistic indoor modeling. The location-centric individual patterns analysis part covers buildings density predictions and distributions and proposes an advanced architecture for meta density predictions, while location pairwise patterns part studies flux analysis of the buildings' users. The user-centric analysis covers pairwise encounters while individual dimension studies users visitation preferences and frequency.

ISBN: 9798698567158Subjects--Topical Terms:

621879
Computer engineering.
Subjects--Index Terms:

Analyses

Data-Driven Indoor Mobility Analyses, Modeling, and Encounter Classification for IoT Applications.
LDR:04591nmm a2200493 4500 001 2344564
005 20220531064554.5
008 241004s2020 ||||||||||||||||| ||eng d
020 $a 9798698567158
035 $a (MiAaPQ)AAI27837884
035 $a AAI27837884
040 $a MiAaPQ $c MiAaPQ
100 1 $a Al Qathrady, Mimonah. $3 3683351
245 1 0 $a Data-Driven Indoor Mobility Analyses, Modeling, and Encounter Classification for IoT Applications.
260 1 $a Ann Arbor : $b ProQuest Dissertations & Theses, $c 2020
300 $a 161 p.
500 $a Source: Dissertations Abstracts International, Volume: 82-06, Section: B.
500 $a Advisor: Helmy, Ahmed A.
502 $a Thesis (Ph.D.)--University of Florida, 2020.
506 $a This item must not be sold to any third party vendors.
520 $a Data-driven indoor mobility models and encounter distance classification are essential elements in the design and evaluation of numerous Internet of things (IoT) indoor applications. One critical example of such applications is infection tracing in smart hospitals, also referred to as 'contact tracing' in epidemiology. Infection tracing needs a realistic characterization of encounters, such as proximity classification and distance estimation between nodes. Most importantly, the simulation and evaluation of the tracing systems and many other indoor IoT applications require a realistic indoor mobility model. Consequently, this dissertation comprises three main parts: 1- infection tracing in the intelligent hospital (i-hospital), 2- encounter distance estimation and classification, and 3- data-driven indoor mobility analyses and modeling for location-centric and user-centric metrics. The first part of this work describes the i-hospital project, including infection tracing approaches and challenges. The second part is about the encounter distances and proximity classification. Many existing encounter-based solutions rely on wireless data, such as access point associations or Bluetooth scanning records. However, the wireless encounter records reflect when two nodes are within sensing or communication range from each other, and not within the problem range of interest, such as infection range as in the infection tracing problems. In this part, the Bluetooth Low Energy (BLE) is targeted as a technology since it is built specifically to facilitate IoT. The BLE transmission power (TX power) is integrated with a Received Signal Strength Indicator (RSSI) into the state-of-art parametric and machine learning models to improve the distance estimation and proximity classification models. In addition, a new promising classifier-based distance estimation model architecture is proposed and evaluated extensively to show its effectiveness. Furthermore, we have built a library that contains the TX power and RSSI data of BLE encounters within defined distances of up to 22 meters in different indoor environments. Understanding and modeling nodes' mobility patterns are vital for future IoT and mobile services design and performance evaluation.Consequently, the third (and most comprehensive) part of this dissertation focuses on data-driven indoor mobility analyses and modeling. It introduces a detailed data analyses methodology framework that classifies analyses into two perspectives: user-centric and location-centric. In this work, real wireless traces at the building level have been studied from individual and pairwise dimensions from each perspective to provide guidelines for realistic indoor modeling. The location-centric individual patterns analysis part covers buildings density predictions and distributions and proposes an advanced architecture for meta density predictions, while location pairwise patterns part studies flux analysis of the buildings' users. The user-centric analysis covers pairwise encounters while individual dimension studies users visitation preferences and frequency.
590 $a School code: 0070.
650 4 $a Computer engineering. $3 621879
650 4 $a Remote sensing. $3 535394
650 4 $a Epidemiology. $3 568544
650 4 $a Artificial intelligence. $3 516317
650 4 $a Information technology. $3 532993
650 4 $a Health care management. $3 2122906
653 $a Analyses
653 $a Indoor mobility
653 $a IoT applications
653 $a Mobility
653 $a Prediction
653 $a Proximity
653 $a Internet of Things
653 $a Smart hospitals
653 $a Wireless data
653 $a Distance estimation
653 $a Bluetooth Low Energy
653 $a Machine learning
690 $a 0464
690 $a 0489
690 $a 0800
690 $a 0799
690 $a 0766
690 $a 0769
710 2 $a University of Florida. $b Communication Sciences. $3 3683352
773 0 $t Dissertations Abstracts International $g 82-06B.
790 $a 0070
791 $a Ph.D.
792 $a 2020
793 $a English
856 4 0 $u http://pqdd.sinica.edu.tw/twdaoapp/servlet/advanced?query=27837884