Cover
Title Page
Preface
Acknowledgments
1 Body Sensor Networks
1. 1.1 Introduction
2. 1.2 Background
3. 1.3 Typical m‐Health System Architecture
4. 1.4 Hardware Architecture of a Sensor Node
5. 1.5 Communication Medium
6. 1.6 Power Consumption Considerations
7. 1.7 Communication Standards
8. 1.8 Network Topologies
9. 1.9 Commercial Sensor Node Platforms
10. 1.10 Biophysiological Signals and Sensors
11. 1.11 BSN Application Domains
12. 1.12 Summary
13. References
2 BSN Programming Frameworks
1. 2.1 Introduction
2. 2.2 Developing BSN Applications
3. 2.3 Programming Abstractions
4. 2.4 Requirements for BSN Frameworks
5. 2.5 BSN Programming Frameworks
6. 2.6 Summary
7. References
3 Signal Processing In‐Node Environment
1. 3.1 Introduction
2. 3.2 Background
3. 3.3 Motivations and Challenges
4. 3.4 The SPINE Framework
5. 3.5 Summary
6. References
4 Task‐Oriented Programming in BSNs
1. 4.1 Introduction
2. 4.2 Background
3. 4.3 Motivations and Challenges
4. 4.4 SPINE2 Overview
5. 4.5 Task‐Oriented Programming in SPINE2
6. 4.6 SPINE2 Node‐Side Middleware
7. 4.7 SPINE2 Coordinator
8. 4.8 SPINE2 Communication Protocol
9. 4.9 Developing Application in SPINE2
10. 4.10 Summary
11. References
5 Autonomic Body Sensor Networks
1. 5.1 Introduction
2. 5.2 Background
3. 5.3 Motivations and Challenges
4. 5.4 State‐of‐the‐Art
5. 5.5 SPINE‐*: Task‐Based Autonomic Architecture
6. 5.6 Autonomic Physical Activity Recognition
7. 5.7 Summary
8. References
6 Agent‐Oriented Body Sensor Networks
1. 6.1 Introduction
2. 6.2 Background
3. 6.3 Motivations and Challenges
4. 6.4 State‐of‐the‐Art: Description and Comparison
5. 6.5 Agent‐Based Modeling and Implementation of BSNs
6. 6.6 Engineering Agent‐Based BSN Applications: A Case Study
7. 6.7 Summary
8. References
7 Collaborative Body Sensor Networks
1. 7.1 Introduction
2. 7.2 Background
3. 7.3 Motivations and Challenges
4. 7.4 State‐of‐the‐Art
5. 7.5 A Reference Architecture for Collaborative BSNs
6. 7.6 C‐SPINE: A CBSN Architecture
7. 7.7 Summary
8. References
8 Integration of Body Sensor Networks and Building Networks
1. 8.1 Introduction
2. 8.2 Background
3. 8.3 Motivations and Challenges
4. 8.4 Integration Layers
5. 8.5 State‐of‐the‐Art: Description and Comparison
6. 8.6 An Agent‐Oriented Integration Gateway
7. 8.7 Application Scenarios
8. 8.8 Summary
9. References
9 Integration of Wearable and Cloud Computing
1. 9.1 Introduction
2. 9.2 Background
3. 9.3 Motivations and Challenges
4. 9.4 Reference Architecture for Cloud‐Assisted BSNs
5. 9.5 State‐of‐the‐Art: Description and Comparison
6. 9.6 BodyCloud: A Cloud‐based Platform for Community BSN Applications
7. 9.7 Engineering BodyCloud Applications
8. 9.8 Summary
9. References
10 Development Methodology for BSN Systems
1. 10.1 Introduction
2. 10.2 Background
3. 10.3 Motivations and Challenges
4. 10.4 SPINE‐Based Design Methodology
5. 10.5 Summary
6. References
11 SPINE‐Based Body Sensor Network Applications
1. 11.1 Introduction
2. 11.2 Background
3. 11.3 Physical Activity Recognition
4. 11.4 Step Counter
5. 11.5 Emotion Recognition
6. 11.6 Handshake Detection
7. 11.7 Physical Rehabilitation
8. 11.8 Summary
9. References
12 SPINE at Work
1. 12.1 Introduction
2. 12.2 SPINE 1.x
3. 12.3 SPINE2
Index
End User License Agreement

List of Tables

Chapter 01
1. Table 1.1 List of commercial sensor node platforms.
2. Table 1.2 Summary of representative BSN systems.
Chapter 02
1. Table 2.1 BSN‐application development approaches comparison.
2. Table 2.2 Common tasks of BSN applications.
3. Table 2.3 Requirements for BSN frameworks.
Chapter 03
1. Table 3.1 API exposed by SPINE at the coordinator station.
2. Table 3.2 SPINE‐tested mobile personal devices.
Chapter 04
1. Table 4.1 SPINE2 application‐level messages.
Chapter 05
1. Table 5.1 Activity recognition accuracy affected by short faults over all channels and C = 3.
2. Table 5.2 Activity recognition accuracy affected by short faults over a specific channel and C = 3.
3. Table 5.3 Accuracy improvements over all channels and C = 3.
Chapter 06
1. Table 6.1 Comparison among agent‐oriented platforms (Agilla, ActorNet, AFME, and MAPS) for WSNs.
Chapter 08
1. Table 8.1 Sensing services of the BSN system for human activity recognition
Chapter 09
1. Table 9.1 Architectures for the integration of wireless sensor networks with cloud computing: a comparison.
2. Table 9.2 Architectures for the integration of body area networks with cloud computing: a comparison.
Chapter 11
1. Table 11.1 Posture/movement recognition accuracy.
2. Table 11.2 Stress threshold for HRV parameters.

List of Illustrations

Chapter 01
1. Figure 1.1 Common wearable sensors and their location on the human body.
2. Figure 1.2 A three‐tier hierarchical BSN architecture: (1) body sensor tier, (2) personal area network tier, and (3) global network tier.
3. Figure 1.3 Typical hardware architecture of a sensor node.
4. Figure 1.4 Peer‐to‐peer topology.
5. Figure 1.5 Star topology.
6. Figure 1.6 Mesh topology.
7. Figure 1.7 Clustered topology.
8. Figure 1.8 Mica Mote.
9. Figure 1.9 TelosB Tmote Sky.
10. Figure 1.10 Different revisions of the Shimmer platform.
Chapter 02
1. Figure 2.1 Reference model of a middleware‐based programming framework.
Chapter 03
1. Figure 3.1 The SPINE middleware architecture.
2. Figure 3.2 The SPINE Node software architecture.
3. Figure 3.3 The SPINE Coordinator software architecture.
4. Figure 3.4 Java desktop implementation of the SPINE Management GUI (sensor‐node configuration dialog window).
5. Figure 3.5 Android implementation of the SPINE Management GUI (sensor and function configuration dialog windows).
6. Figure 3.6 Data processing chain supported by the SPINE High‐level Data Processing plug‐in.
7. Figure 3.7 High‐Level Data Processing layered software architecture.
Chapter 04
1. Figure 4.1 The software layering approach in the SPINE2 middleware.
2. Figure 4.2 A task‐oriented application with tasks instantiated on different nodes.
3. Figure 4.3 Software architecture of the node‐side part of the framework.
4. Figure 4.4 The two‐layer protocol stack (a) and the packet fields (b).
5. Figure 4.5 The SPINE2 components interacting with the user applications.
Chapter 05
1. Figure 5.1 The multiplane autonomic architecture of a SPINE‐* application.
2. Figure 5.2 Examples of application with self‐configuring property; (a) the reconfiguration task is driven by the output results of the Processing task; (b) the reconfiguration task is driven by the desktop application running on the BSN coordinator.
3. Figure 5.3 Example of application with self‐healing property.
4. Figure 5.4 Example of application with self‐optimization property.
5. Figure 5.5 Example of application with self‐protection property.
6. Figure 5.6 The task‐based applications on the waist node (a) and on the thigh node (b).
7. Figure 5.7 The tested activities’ sequence.
8. Figure 5.8 The autonomic application running on the waist node.
Chapter 06
1. Figure 6.1 Architecture of MAPS.
2. Figure 6.2 Agent behavior model of MAPS.
3. Figure 6.3 Agent modeling of BSNs. (a) Master/slave model, (b) Master/slave + peer‐to‐peer model, and (c) Super peer model.
4. Figure 6.4 Architecture of the agent‐based activity monitoring system.
5. Figure 6.5 Finite state machine of the sensor agents: WaistSensorAgent and ThighSensorAgent.
Chapter 07
1. Figure 7.1 BSN infrastructures based on the logical communications among individuals and base stations.
2. Figure 7.2 The reference CBSN Network Architecture.
3. Figure 7.3 Activity diagram of basic CBSN operations.
4. Figure 7.4 CBSN Functional Architecture components.
5. Figure 7.5 C‐SPINE Functional Architecture components.
6. Figure 7.6 Inter‐BSN component interaction.
Chapter 08
1. Figure 8.1 An example of the Building Network environment.
2. Figure 8.2 The overall BMF framework architecture.
3. Figure 8.3 BN/BSN integration: a scenario.
4. Figure 8.4 BN/BSN integration layers.
5. Figure 8.5 BMF‐BN/SPINE‐BSN integration based on the gateway approach.
6. Figure 8.6 Class diagram of the agent‐based gateway.
7. Figure 8.7 Interaction between the agent‐based gateway (pair <BMFAgent, SPINEAgent>) and the BMF Coordinator.
8. Figure 8.8 Architecture of the in‐building physical activity recognition system.
Chapter 09
1. Figure 9.1 The cloud computing ecosystem.
2. Figure 9.2 Reference architecture for the integration of BSN and cloud computing.
3. Figure 9.3 The BodyCloud architecture.
4. Figure 9.4 Workflow schema of the BodyCloud approach for developing community BSN applications.
5. Figure 9.5 ECGMonitoring DataFeed modality.
6. Figure 9.6 ECGMonitoring GroupAnalysis modality.
7. Figure 9.7 EcgToRR workflow.
8. Figure 9.8 GUI view. (a) ECG wave plotting and (b) beat per minute instantaneous value.
9. Figure 9.9 CDRDetection DataFeed modality.
10. Figure 9.10 SingleCDRAnalysis modality.
11. Figure 9.11 SingleCDR workflow.
12. Figure 9.12 GUI view: detection of a CDR.
13. Figure 9.13 RehabMonitoring DataFeed modality.
14. Figure 9.14 Single RehabMonitoringAnalysis modality.
15. Figure 9.15 RehabMonitoring workflow.
16. Figure 9.16 GUI view: knee rehabilitation.
17. Figure 9.17 RawAccelerationDataFeed modality.
18. Figure 9.18 Single ActivityMonitoring Analysis modality.
19. Figure 9.19 ActivityMonitoring workflow.
20. Figure 9.20 GUI view: activity statistics.
Chapter 10
1. Figure 10.1 Architecture and function platforms.
2. Figure 10.2 Mapping of function and architecture.
3. Figure 10.3 Pattern architectural schemas: (a) Sensor Data Collection for Monitoring; (b) Multisensor Data Fusion for Detection/Classification of Events.
4. Figure 10.4 SPINE‐based Platform Design process schema.
Chapter 11
1. Figure 11.1 Two screenshots of the developed activity recognition Android app.
2. Figure 11.2 The main monitoring window of the stress detection system.
3. Figure 11.3 Block diagram of the proposed adaptive QRS detection algorithm.
4. Figure 11.4 The proposed CDR detection algorithm applied to RRi series.
5. Figure 11.5 A screenshot of the developed CDR detection mobile application.
6. Figure 11.6 The E‐Shake application.
7. Figure 11.7 The E‐Shake system architecture.
8. Figure 11.8 Two screenshots of the rehabilitation digital assistant.
Chapter 12
1. Figure 12.1 The task‐oriented application defined and deployed in “SPINE2SimpleTest.java” application.

This edition first published 2018
© 2018 John Wiley & Sons, Inc.

All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means, electronic, mechanical, photocopying, recording or otherwise, except as permitted by law. Advice on how to obtain permission to reuse material from this title is available at http://www.wiley.com/go/permissions.

The right of Giancarlo Fortino, Raffaele Gravina, and Stefano Galzarano to be identified as the authors of this work has been asserted in accordance with law.

Registered Office
John Wiley & Sons, Inc., 111 River Street, Hoboken, NJ 07030, USA

Editorial Office
111 River Street, Hoboken, NJ 07030, USA

For details of our global editorial offices, customer services, and more information about Wiley products visit us at www.wiley.com.

Wiley also publishes its books in a variety of electronic formats and by print‐on‐demand. Some content that appears in standard print versions of this book may not be available in other formats.

Limit of Liability/Disclaimer of Warranty
In view of ongoing research, equipment modifications, changes in governmental regulations, and the constant flow of information relating to the use of experimental reagents, equipment, and devices, the reader is urged to review and evaluate the information provided in the package insert or instructions for each chemical, piece of equipment, reagent, or device for, among other things, any changes in the instructions or indication of usage and for added warnings and precautions. While the publisher and authors have used their best efforts in preparing this work, they make no representations or warranties with respect to the accuracy or completeness of the contents of this work and specifically disclaim all warranties, including without limitation any implied warranties of merchantability or fitness for a particular purpose. No warranty may be created or extended by sales representatives, written sales materials or promotional statements for this work. The fact that an organization, website, or product is referred to in this work as a citation and/or potential source of further information does not mean that the publisher and authors endorse the information or services the organization, website, or product may provide or recommendations it may make. This work is sold with the understanding that the publisher is not engaged in rendering professional services. The advice and strategies contained herein may not be suitable for your situation. You should consult with a specialist where appropriate. Further, readers should be aware that websites listed in this work may have changed or disappeared between when this work was written and when it is read. Neither the publisher nor authors shall be liable for any loss of profit or any other commercial damages, including but not limited to special, incidental, consequential, or other damages.

Library of Congress Cataloging‐in‐Publication Data

Names: Fortino, Giancarlo, 1971– author. | Gravina, Raffaele, 1982– author. | Galzarano, Stefano, 1984– author.
Title: Wearable computing : from modeling to implementation of wearable systems based on body sensor networks / Giancarlo Fortino, Raffaele Gravina, Stefano Galzarano.
Description: 1st edition. | Hoboken, NJ : John Wiley & Sons, 2018. | Includes bibliographical references and index. |
Identifiers: LCCN 2017053912 (print) | LCCN 2017059016 (ebook) | ISBN 9781119078821 (pdf) | ISBN 9781119078838 (epub) | ISBN 9781118864579 (cloth)
Subjects: LCSH: Wearable computers. | Sensor networks.
Classification: LCC QA76.592 (ebook) | LCC QA76.592 .F67 2018 (print) | DDC 004.167–dc23
LC record available at https://lccn.loc.gov/2017053912

Cover design by Wiley
Cover images: © nopporn/Shutterstock; © Sergey Nivens/Shutterstock

Preface

Wearable computing is a relatively new area of research and development that aims at supporting people in different application domains: health care, fitness, social interactions, video games, and smart factory. Wearable computing is based on wearable sensor devices (e.g. to measure heart rate, temperature, or blood oxygen), common life objects (e.g. watch, belt, or shoes), and personal handheld devices (e.g. smartphones or tablets). Wearable computing has been recently boosted by the introduction of body sensor networks (BSNs), i.e. networks of wireless wearable sensor nodes coordinated by more capable coordinators (smartphones, tablets, and PCs).

In particular, BSNs enable a very wide range of application scenarios in different industry sectors. We can categorize them into different domains: e‐Health, e‐Emergency, e‐Entertainment, e‐Sport, e‐Factory, and e‐Social.

e‐Health applications span from early detection or prevention of diseases, elderly assistance at home, to post‐trauma rehabilitation after surgeries. e‐Emergency applications include BSN systems to support fire fighters, response teams in large‐scale disasters due to earthquakes, landslides, terrorist attacks, etc. e‐Entertainment domain refers to human–computer interaction systems typically based on BSNs for real‐time motion and gesture recognition. e‐Sport applications are related to the e‐Health domain, although they have a nonmedical focus. Specifically, this domain includes personal e‐fitness applications for amateur and professional athletes, as well as enterprise systems for fitness clubs and sport teams offering advanced performance monitoring services for their athletes. e‐Factory is an emerging and very promising domain involving industrial process management and monitoring, and workers’ safety and collaboration support. Finally, e‐Social applications may use BSN technologies to recognize user emotions and cognitive states to enable new forms of social interactions with friends and colleagues. An interesting example is given by a system that involves the interaction between two people’s BSNs to detect handshakes and, subsequently monitor their social and emotional interactions.

Although the basic elements (sensors, protocols, and coordinators) of a BSN are available (already from a commercial point of view), developing BSN systems/applications is a complex task that requires design methods based on effective and efficient programming frameworks. In this book, we will provide programming approaches and methods to effectively develop efficient BSN systems/applications. Moreover, we also provide new techniques to integrate BSN‐based wearable systems with more general Wireless Sensor Network systems and with Cloud computing.

This book, entitled Wearable Computing: From Modeling to Implementation of Wearable Systems Based on Body Sensor Networks, is based on an intense and extensive basic and applied research activity driven by the SPINE project (http://spine.deis.unical.it), whose authors are cofounders, responsible, and main developers. Thus, the book is connected to the SPINE website to provide readers with software and tools for the development of their wearable computing systems.

This book is aimed at a large audience in the Wearable Computing domain, that is gaining considerable research interest and momentum, and is expected to be of increasing interest to academic researchers and particularly to commercial developers. Upon reading this book the audiences will perceive the following benefits:

Learn the state‐of‐the‐art in research and development on wearable computing, wireless BSNs, wearable systems integrated with mobile computing, wireless networking, and cloud computing.
Obtain a future roadmap by learning advanced technology and open research issues.
Gather the background knowledge to tackle key problems, whose solutions will enhance the evolution of next‐generation wearable systems.
Use the book as a valuable reference for a technical professional in a related industry.
Use the book as a text book in the late undergraduate or the graduate level to prepare students who intend to perform research in the field of the book or intend to be employed in a related industry.

The main topics of the book are the following:

Wearable Computing, the study or practice of inventing, designing, building, or using miniature body‐borne computational and sensory devices. Wearable computers may be worn under, over, or in clothing, or may also be themselves clothes.
Wireless Sensor Networks (WSNs), collections of tiny devices capable of sensing, computation, and wireless communication operating in a certain environment to monitor and control events of interest in a distributed manner and collectively react to critical situations. WSN applications span various domains such as environmental and building monitoring and surveillance, pollution monitoring, agriculture, health care, home‐automation, energy management, earthquake, and eruption monitoring.
Body Sensor Networks (BSNs), involving wireless wearable physiological sensors applied to the human body for medical and nonmedical purposes. In particular, they allow for the continuous measurement of body movements and physiological parameters, such as heart rate, muscular tension, skin conductivity, and breathing rate and volume, during the daily life of a user.
In‐node Signal Processing, a central computing method in advanced wireless sensor platforms through which data processing is carried out directly on the sensor node to preprocess data acquired from sensors, to fuse data coming from other sensor nodes, and, notably, to perform higher level computation such as classification and decision making.
Mobile Computing, human–computer interaction by which a computer is expected to be transported during normal usage. Mobile computing involves mobile communication, mobile hardware, and mobile software. Communication issues include ad‐hoc and infrastructure networks as well as communication properties, protocols, data formats, and concrete technologies. Hardware includes mobile devices or device components. Mobile software deals with the characteristics and requirements of mobile applications.
Cloud Computing, the use of computing resources (hardware and software) that are delivered as a service over a network (typically the Internet). The name comes from the use of a cloud‐shaped symbol as an abstraction for the complex infrastructure it contains in system diagrams. Cloud computing entrusts remote services with a user’s data, software, and computation.
Platform‐Based Design (PBD), an embedded computing design methodology that consists of a sequence of design/development steps that leads the initial high‐level description of a digital system to its final implementation. Each step is a refinement process that transforms the design from a higher level description to a lower level description that is progressively closer to the final implementation.
Software Framework, an abstraction in which software providing generic functionality can be selectively changed by user code, thus providing application‐specific software. A software framework is a universal, reusable software platform used to develop applications, products, and solutions. Software Frameworks include support programs, compilers, code libraries, an application programming interface (API), and tool sets that bring together all the different components to enable development of a project or solution.
Autonomic Computing is a paradigm born as a response to the increasing complexity of managing computing systems. It faces the problem by introducing a series of self‐* properties (self‐configuration, self‐healing, self‐optimization, and self‐protection) into complex systems, through which such systems can be capable of performing several self‐management actions without any human intervention.
Activity Recognition aims to recognize the actions and goals of one or more agents from a series of observations on the agents’ actions and the environmental conditions. Since the 1980s, this research field has captured the attention of several computer science communities due to its strength in providing personalized support for many different applications and its connection to many different fields of study such as medicine, human–computer interaction, or sociology. Specifically, we are mainly interested in sensor‐based single‐user and multiuser activity recognition that integrates the emerging area of sensor networks with novel data mining and machine learning techniques to model a wide range of human activities.

Specifically, this book is organized into 12 chapters:

Chapter 1, Body Sensor Networks (BSNs), covers the state‐of‐the‐art about wearable sensor nodes, network architecture/protocols/standards, and applications/systems.
Chapter 2, BSN Programming Frameworks, analyzes the state‐of‐the‐art about the most known software frameworks (CodeBlue, Titan, RehabSPOT, and others) for programming BSN applications/systems.
Chapter 3, Signal Processing In‐Node Environment, describes in detail the SPINE framework (http://spine.deis.unical.it) from architectural and programming perspectives.
Chapter 4, Task‐Oriented Programming, discusses task‐oriented programming of BSN applications through SPINE2.
Chapter 5, Autonomic BSNs, illustrates how to make BSNs autonomic, by using SPINE*, an extension of SPINE2.
Chapter 6, Agent‐oriented BSNs, presents the use of the Agent paradigm for programming BSN systems. Specifically, the MAPS (Mobile Agent Platform for SunSPOT) framework is used to design and implement agent‐based BSNs.
Chapter 7, Collaborative BSNs, provides an introduction of methods and architectures to make BSNs interact with each other for supporting multiuser BSN applications.
Chapter 8, Integration of BSNs and Wireless Sensor Networks, covers gateway‐based solution for interoperability between BSNs and infrastructural WSNs (e.g. building indoor sensor networks). This would enable “invisible” interaction between BSN‐worn people and the surrounding environment.
Chapter 9, Integration of Wearable and Cloud Computing, presents an architecture for the integration of BSNs and the Cloud, called BodyCloud, based on Google App Engine. It is crucial now to move the data acquired or preprocessed on the human body to the cloud for storing and nonreal‐time analysis purposes.
Chapter 10, Development Methodology for BSN Systems, describes a SPINE‐based methodology for the development of BSN systems. The methodology guides the BSN system developer from requirement analysis to implementation and deployment.
Chapter 11, SPINE‐based BSN Applications, presents several applications developed through SPINE in different application domains (Activity Recognition: recognition of human postures and movements, Emotion Recognition: recognition of stress and fear, Handshake Detection: collaborative recognition of two people’s handshake, and Rehabilitation: real‐time computation of extension angles of elbow/knee).
Chapter 12, SPINE at Work, provides a quick yet effective reference for BSN programmers interested in developing their applications using the SPINE framework. The chapter provides the necessary information for setting up the SPINE environment so as to start programming as well as insights on how the framework itself can be customized and extended.