Introduction to the Internet of Things PDF

INTRODUCTION TO THE INTERNET OF THINGS DLBINGEIT01_E INTRODUCTION TO THE INTERNET OF THINGS MASTHEAD Publisher: IU Internationale Hochschule GmbH IU International University of Applied Sciences Juri-Gagarin-Ring 152 D-99084 Erfurt Mailing address: Albert-Proeller-Straße 15-19 D-86675 Buchdorf [email protected] www.iu.de DLBINGEIT01_E Version No.: 001-2024-0209 Reza Vahidnia © 2023 IU Internationale Hochschule GmbH This course book is protected by copyright. All rights reserved. This course book may not be reproduced and/or electronically edited, duplicated, or dis- tributed in any kind of form without written permission by the IU Internationale Hoch- schule GmbH (hereinafter referred to as IU). The authors/publishers have identified the authors and sources of all graphics to the best of their abilities. However, if any erroneous information has been provided, please notify us accordingly. 2 TABLE OF CONTENTS INTRODUCTION TO THE INTERNET OF THINGS Introduction Signposts Throughout the Course Book............................................. 6 Suggested Readings............................................................... 7 Learning Objectives............................................................... 9 Unit 1 Internet of Things — Fundamentals 11 1.1 The Internet of Things—Basics and Motivation.................................. 12 1.2 Evolution of the Internet—Web 1.0 to Web 4.0................................... 20 Unit 2 Social and Economic Significance 27 2.1 Innovations for Consumers and Industry....................................... 28 2.2 Implications on People and the World of Work.................................. 32 2.3 Data Protection and Data Security............................................. 37 Unit 3 Communication Standards and Technologies 43 3.1 Network Topologies.......................................................... 44 3.2 Network Protocols........................................................... 48 3.3 Technologies................................................................ 54 Unit 4 Data Storage and Processing 63 4.1 Networked Storage with Linked Data and RDF(S)................................ 64 4.2 Analysis of Networked Data Using a Semantic Reasoner......................... 68 4.3 Processing of Data Streams with Complex Event Processing...................... 71 4.4 Operation and Analysis of Large Data Clusters Using NoSQL and MapReduce...... 73 Unit 5 Design and Development 77 5.1 Software Engineering for Distributed and Embedded Systems.................... 78 5.2 Architectural Patterns and Styles for Distributed Systems........................ 84 5.3 Platforms: Microcontrollers, Mono-board Computers, and One-Chip Systems...... 91 3 Unit 6 Applicability 97 6.1 Smart Home/Smart Living.................................................... 98 6.2 Ambient Assisted Living..................................................... 102 6.3 Smart Energy/Smart Grid.................................................... 103 6.4 Smart Factory.............................................................. 105 6.5 Smart Logistics............................................................. 107 Appendix List of References............................................................... 112 List of Tables and Figures........................................................ 120 4 INTRODUCTION WELCOME SIGNPOSTS THROUGHOUT THE COURSE BOOK This course book contains the core content for this course. Additional learning materials can be found on the learning platform, but this course book should form the basis for your learning. The content of this course book is divided into units, which are divided further into sec- tions. Each section contains only one new key concept to allow you to quickly and effi- ciently add new learning material to your existing knowledge. At the end of each section of the digital course book, you will find self-check questions. These questions are designed to help you check whether you have understood the con- cepts in each section. For all modules with a final exam, you must complete the knowledge tests on the learning platform. You will pass the knowledge test for each unit when you answer at least 80% of the questions correctly. When you have passed the knowledge tests for all the units, the course is considered fin- ished and you will be able to register for the final assessment. Please ensure that you com- plete the evaluation prior to registering for the assessment. Good luck! 6 SUGGESTED READINGS GENERAL SUGGESTIONS Firouzi, F., Chakrabarty, K., & Nassif, S. (2020). Intelligent Internet of Things: From device to fog and cloud.Springer. http://search.ebscohost.com.pxz.iubh.de:8080/login.aspx?dire ct=true&db=cat05114a&AN=ihb.49660&site=eds-live&scope=site Hanes, D., Salgueiro, G., Grossetete, P., Barton, R., & Henry, J. (2017). IoT fundamentals: Networking technologies, protocols, and use cases for the Internet of Things. Cisco Press. UNIT 1 Choudhury, N. (2014). World Wide Web and its journey from Web 1.0 to Web 4.0. Interna- tional Journal of Computer Science and Information Technologies, 5(6), 8096— 8100.Available online. UNIT 2 McKinsey Global Institute. (2015). The Internet of Things: Mapping the value beyond the hype: June 2015 executive summary. McKinsey & Company. Available online. The Consumer Goods Forum, Capgemini, & Intel. (2016). Making the connection: How the Internet of Things engages consumers and benefits business.Available online. UNIT 3 Ding, J., Nemati, M., Ranaweera, C., & Choi, J. (2020). IoT connectivity technologies and applications: A survey. IEEE Access, 8, 67646—67673. http://search.ebscohost.com.pxz. iubh.de:8080/login.aspx?direct=true&db=edseee&AN=edseee.9057670&site=eds-live& scope=site UNIT 4 Cai, H., Xu, B., Jiang, L., & Vasilakos, A. V. (2017). IoT-based big data storage systems in cloud computing: Perspectives and challenges. IEEE Internet of Things Journal, 4(1), 75 —87.http://search.ebscohost.com.pxz.iubh.de:8080/login.aspx?direct=true&db=edsee e&AN=edseee.7600359&site=eds-live&scope=site 7 UNIT 5 Richards, M., & Ford, N. (2020). Fundamentals of software architecture: An engineering approach. O’Reilly.Chapter 8 http://search.ebscohost.com.pxz.iubh.de:8080/login.aspx?direct=true&db=cat05114a&AN =ihb.48806&lang=de&site=eds-live&scope=site Sommerville, I. (2016). Software engineering (Global edition, 10th ed.). Pearson.Chapter 3h ttp://search.ebscohost.com.pxz.iubh.de:8080/login.aspx?direct=true&db=cat05114a& AN=ihb.49267&site=eds-live&scope=site UNIT 6 Hassan, R., Qamar, F., Hasan, M. K., Aman, A. H. M., & Ahmed, A. S. (2020). Internet of Things and its applications: A comprehensive survey. Symmetry, 12(10), Article 1674.ht tp://search.ebscohost.com.pxz.iubh.de:8080/login.aspx?direct=true&db=edsdoj&AN= edsdoj.2d3757948668425998204cb8418372e1&site=eds-live&scope=site 8 LEARNING OBJECTIVES The Internet of Things (IoT) is an interdisciplinary field that brings various expertise and technologies, such as software engineering, data communication, information technology (IT), and data science, together to create value by connecting the unconnected. Analyzing the IoT data captured from connected devices creates new insights for people to improve their quality of life and for businesses to revolutionize their business models and increase their profitability. You will begin your Introduction to the Internet of Things by studying the key compo- nents of the IoT ecosystem and IoT value chain, as well as IoT use cases and applications in various verticals and industry segments. Reviewing the challenges, requirements, and benefits of IoT solutions is necessary to understand their social and economic impact. The IoT network capabilities are typically limited compared to IT networks, resulting in the need for special technologies, protocols, standards, and security considerations. You will find that recent advances in IoT have led to the development of innovative use cases and applications. These are facilitated by new data communication, storage, and processing technologies, which are the main building blocks of the IoT layer model studied in this course. Different IoT applications and use cases require different software design and develop- ment considerations. This course book will examine distributed and embedded systems. It will also explore various platforms and software architectural styles and patterns that are used to develop IoT solutions for a wide range of use cases in segments such as smart homes, ambient assisted living, smart energy, smart factory, and smart logistics. 9 UNIT 1 INTERNET OF THINGS — FUNDAMENTALS STUDY GOALS On completion of this unit, you will be able to … – define the Internet of Things (IoT) ecosystem and its key elements. – identify the components of the IoT value chain. – list the significant challenges and benefits of IoT solutions. – classify IoT clusters as verticals, use cases, and applications. – describe the characteristics and limitations of each evolutionary phase of the web. 1. INTERNET OF THINGS — FUNDAMENTALS Introduction In the world of the Internet of Things (IoT), just about any “thing” you can imagine—a sim- ple temperature sensor, a lamp that can be switched on and off remotely, or very advanced drones—can be connected to the internet with the capability of exchanging data or interacting with people and other smart objects. The information collected from these smart devices can be used and analyzed to create insight for businesses, automate pro- cesses, enable new services and applications, and ultimately enhance the quality of life. IoT impacts every aspect of our lives. It has changed the way we monitor our health, how we get to work, or even do our grocery shopping. The number of IoT use cases in various industries and verticals, such as energy, retail, healthcare, agriculture, transportation, and industrial automation are growing rapidly. There are currently more than 14 billion IoT device connections, and estimations predict that this will exceed 30 billion by 2025 (Lueth, 2020). The rapid growth of IoT and the massive amount of internet data exchange needing to be stored, processed, and analyzed will bring its own challenges (e.g., scaling the myriad devices, security, and privacy). It will also create new opportunities for device manufacturers, service providers, application developers, and other key players and stakeholders in the IoT value chain. Forecasts predict that these combined opportunities will create a global IoT market of $1.5 trillion by the end of 2025 (Lueth, 2020). 1.1 The Internet of Things—Basics and Motivation The internet has undergone several evolutionary phases since its inception. During the first internet phase of the mid-1990s, it offered a basic connectivity that gave people access to web services and email. The second evolutionary phase realized the networked economy by enabling e-commerce and digitizing businesses to increase efficiency and profit. This was when suppliers, vendors, and consumers became connected, with wide- spread access to online shopping. The third phase digitized personal interactions through social media, cloud services, and mobile phone applications. IoT, or the “digitized world”, is considered the fourth evolutionary phase of the internet, where new services and appli- cations are enabled by connecting various devices. In this phase, the four major compo- nents of the IoT ecosystem—things, data, people, and processes—are connected to create new values and have greater impact on businesses and communities. In fact, in the IoT phase of the internet, data are created by sensors and devices rather than people (Hanes et al., 2017). 12 Figure 1: Evolutionary Phases of the Internet Source: Reza Vahidnia (2022), based on Hanes et al. (2017). IoT versus M2M The term machine-to-machine (M2M) communication and IoT are sometimes used inter- changeably. However, these terminologies are quite different. The isolated machines of the M2M world do not necessarily require an active internet connection; instead, they can connect directly to other devices (point-to-point) that use the same protocols. Thus, the scalability and interoperability of M2M networks is limited. Conversely, IoT networks are inherently scalable and support almost any device with an active internet connection. The IoT is more software-based and the volume of data in these networks is usually much larger than hardware-based M2M networks. IoT Ecosystem The multifaceted, broad world of IoT is viewed as an umbrella of use cases, technologies, and protocols. This creates an ecosystem comprised of four main elements—things, peo- ple, data, and process—that work together to achieve the goals of a connected world (Fir- ouzi et al., 2020). 13 Figure 2: The Internet of Things Ecosystem Source: Reza Vahidnia (2022), based on Firouzi et al. (2020); SlideModel (n.d.). Things A device or thing must have at least the following four characteristics to be considered an IoT device or smart object: 1. Sensor(s) to measure the physical parameters and collect data (e.g., temperature or humidity sensors), and/or actuator(s) to interact with the physical world and perform a task (e.g., an electric valve). 2. A communication unit that connects the smart object to the internet, typically ZigBee through wireless technologies, such as Wi-Fi, ZigBee, or Bluetooth. This is a low-cost, low- 3. A power source for the sensors, communication, and processing systems. Most IoT power wireless technol- ogy mainly used in mesh devices are battery-powered, requiring very low amounts of energy consumption for networks and smart long-term use. home IoT applications. 4. A processing unit, typically a low-cost microcontroller, to process and analyze the data collected from the sensors and to control other systems, such as power and com- munication units. Data This refers to information collected by sensors and transmitted to the cloud, as well as commands issued to the smart object and actuator. The multitude of IoT devices generate huge amounts of raw data that need to be cleaned, securely stored, processed, and ana- lyzed. To reserve bandwidth and make correct decisions at the right time, the most time- sensitive data gathered by the sensors are processed closest to the IoT devices (i.e., edge of the network). Less time-sensitive data can be transmitted to the central data bases (e.g., cloud) for long-term storage and historical analysis. 14 People People develop, operate, and benefit from IoT networks. The analyzed IoT data are inter- preted by people, giving them insights to enhance their businesses and collaborate more effectively with each other. As the IoT sensors collect personal data, it is important that individuals are aware of the privacy implications that arise from this. Process The data gathered and analyzed by IoT sensors can be utilized to optimize industrial pro- cesses, making them more efficient and intelligent. What is the IoT Value Chain? IoT solutions consist of several components and building blocks. Each component is developed by a range of companies and plays several roles that add a certain value for the solution’s end user. The IoT value chain consists of several elements, as outlined in the fol- lowing. Hardware The hardware or the smart item block consists of a combination of sensors and actuators that capture and act upon data and communication hardware (i.e., chipsets and SIMs used to connect the device to the network). Chipset and smart object manufacturers, such as Bosch, Siemens, and Samsung, are some big players in this component of the value chain (Mackenzie & Rebbeck, 2020). Connectivity The connectivity network (e.g., 3G/4G or satellites) delivers data captured by sensors to the backend (i.e., servers that collect data). Mobile Network Operators (MNOs), such as Deutsche Telekom and T-Mobile, belong to the players adding connectivity value to IoT solutions (Mackenzie & Rebbeck, 2020). Platform The cloud or on-premises backend systems (servers) store data received from IoT devices. The software platform (IoT platform), which is basically the central backbone for the IoT, is used to fill the gap between hardware and applications by providing a combination of ready-to-use features for developers. These features are used to manage devices (e.g., device status monitoring and remote firmware upgrade), analyze data, and manage actions to improve business processes based on real-time sensor data. Advanced end-to- end IoT platforms also provide external application programming interfaces (APIs) and software development kits (SDKs) for third-party systems. An API is a collection of proce- dures and functions used to create applications to access the features and data of another service or application. An SDK is a set of tools used to facilitate the creation of applica- 15 tions. Microsoft Azure, Amazon Web Services (AWS), Bosch IoT Suite, and IBM Watson are some examples of IoT platforms offered by major cloud service providers (Mackenzie & Rebbeck, 2020). Application The system integrators are developers who design and manage IoT applications specific to almost every industry and vertical. Salesforce is an example of an IoT application devel- Customer relationship oped to offer customer relationship management (CRM) applications and automation management software (Mackenzie & Rebbeck, 2020). This refers to a process used by businesses to administer their interac- Figure 3: The Internet of Things Value Chain tions with the customers and manage their accounts. Source: Reza Vahidnia (2022), based on Flaticon (n.d.-a); Flaticon (n.d.-b); Free Icons Library (n.d.- a);Free Icons Library (n.d.-b). The application component typically brings the highest value (up to 40 percent) to IoT applications. Therefore, mobile network operators, traditionally meant to focus on con- nectivity, try to play a solution provider role by establishing partnerships with cloud serv- ice providers and application developers to increase their share in the IoT market. IoT Taxonomy and Benefits IoT has many use cases with different benefits across industries and segments. Before addressing IoT benefits, let us discuss IoT clusters. IoT use cases (e.g., smart metering) in a specific industry segment (vertical) usually need similar data storage, processing, and analysis capabilities. The IoT use cases belonging to a specific industry segment are cate- gorized as an IoT vertical with unique regulatory entities, procedures, protocols, and standardization organizations (Dian & Vahidnia, 2020a). For instance, smart grid and smart meter IoT use cases increase efficiency and reliability in the energy vertical (Alotaibi et al., 2020). In the transportation vertical, in-transit visibility can be enhanced by adopting the connected vehicle and asset tracking use cases. Smart irrigation and product monitoring IoT use cases in the agriculture vertical can increase crop productivity. The safety and experiences of citizens can be improved using smart street lighting, traffic control, and environmental monitoring use cases in smart city initiatives (Campisi et al., 2021). More- over, IoT use cases, such as connected factories, can be used to reduce human error and operational costs in the industry segment. IoT can also be used to automate home sys- tems and increase security. In the healthcare vertical, data-driven decisions can be made in the health monitoring and activity recognition use cases (Dian et al., 2020). 16 Each IoT use case may have multiple IoT applications using similar solutions and software. For example, in IoT connected vehicle use cases, there may be applications for infotain- ment, telematics, and fleet management. Various IoT clusters are illustrated below. Figure 4: Internet of Things Clusters Source: Reza Vahidnia (2022), based on Dian & Vahidnia (2020a). In general, IoT can provide the following benefits for organizations in different verticals and industry segments (Firouzi et al., 2020): efficiency. Analyzed data from connected IoT sensors result in streamlined processes and reduced resource consumption. Resources, activities, and processes can be man- aged more effectively by analyzing IoT data. transparency. The status of processes (e.g., products in a production line) and the loca- tion of objects (e.g., assets, parcels, or vehicles) can be identified using IoT data. Online tracking of supply quantities and environmental monitoring are also obtained by deploying IoT sensors. automation and accuracy. The IoT rule engines and artificial intelligence (AI) support accurate data-driven decisions that significantly reduce human intervention and errors. Predictive analytical models can be applied to IoT data to help businesses diagnose, predict, and resolve unforeseeable issues. safety and responsiveness. Hazardous situations and environments can be quickly and carefully detected, and immediate actions can be taken to address safety issues. IoT Challenges Any emerging technology with advantages for humans comes with challenges that need to be overcome. IoT is not an exception and has its own technical and business implemen- tation and deployment challenges. Many organizations do not have a clear view of the architectural design, implementation, and operation of their IoT network because of the large number of different sensors and actuators as well as the large number of different communication protocols and the associated complexity. The major IoT challenges are described in the following subsections. 17 Scalability In typical IT networks, the maximum number of endpoints that need to be connected to the internet are approximately a few thousand computers, cellphones, printers, servers, and more. The scale of an IoT network, however, could easily be millions of devices that need to be connected to the internet. Imagine a nationwide smart metering use case, where millions of electricity and gas meters are installed in all households. Or consider national post or fleet management services, where hundreds of thousands of assets and vehicles need to be tracked and monitored in real time. Now imagine how challenging it could be to install, configure, upgrade, support, and manage this massive number of things in such IoT networks. Typical IT networks may use Internet Protocol version 4 (IPv4), which is a 32-bit address space in conjunction with technologies, such as network Network address trans- address translation (NAT) to identify each endpoint in the network. Nonetheless, scala- lation bility requirements of IoT networks cannot be met by IPv4. The IPv6 address space (128 This refers to a technol- ogy where a public bits) could be the only solution for many IoT use cases which allow up to 3.4·1038IP address is assigned to addresses. several computers in a private network. Interoperability In typical IT networks, when an endpoint device does not comply with the network requirements, the network administrators may simply ask the device owner to upgrade their firmware or hardware; otherwise, their access to the network will be denied. How- ever, that is not the case in IoT networks. IoT devices and sensors come from different ven- dors, with various standards and communication protocols. An IoT network architect can- not ask a factory to replace thousands of expensive legacy sensors or devices with new devices that are compatible with the IoT network requirements. Instead, tunneling and/or protocol translation mechanisms must be deployed to support legacy protocols (e.g., RS232 serial communication protocol). Security Sensitive data gathered by IoT devices must be securely protected when they are gener- ated by the sensor, processed at the edge, transmitted to the cloud, stored in the data- base, accessed by people, applications and other services, and analyzed by the analytics software. The challenge in securing IoT networks is that the massive number of IoT devices may come from different vendors and manufacturers, use different communication tech- nologies and protocol, and therefore, require different security measures at each layer. IoT network architects should take a holistic, network-level approach to security. IoT devices are often physically exposed to the world. To mitigate security breaches at the edge (physical) layer, these devices need to be tamper-proof, properly authenticated (i.e., their identity should be confirmed), and authorized (establish the rights and privileges of the device). At the communication layer, all data exchanged between the smart object and the cloud (backend application) must be encrypted to avoid security attacks, e.g., man-in- the-middle, where an attacker can secretly relay or alter communication between two endpoints. 18 One must comply with local data protection legislation to ensure that all data in the cloud and databases are stored securely and correctly, with any anomalous behavior immedi- ately detected using rule-based security policies. Moreover, various IoT use cases are vul- nerable to different security threats, requiring different security considerations. For instance, in connected car use cases, malicious cyber attackers may wish to take control of cars, whereas in a cardiac health monitoring system, attackers could affect pacemaker data. Privacy IoT sensors collect sensitive personal and business data. Personal data can range from location to vital health information, to shopping patterns. For businesses and factories, IoT data have monetary value and businesses can be endangered if data are shared with competitors. Individuals and businesses are both extremely concerned about privacy and want to know whether collected data are shared and with whom. Big data and analytics The high volume of IoT data are generated by various sensors in different types and for- mats. Data can be as simple as a temperature sensor or as rich as an electroencephalogra- phy (EEG) signal. The immense types of data need to be properly prepared and structured before they can be stored, analyzed, and represented. Legal Multiple entities within the ecosystem play different roles in the implementation and deployment of IoT solutions. Consider a connected vehicle use case in which hardware and car manufacturers, mobile network operators, cloud services providers, application developers and third parties, and transportation agencies are involved. The first legal challenge for developing such connected vehicle solutions is intellectual property (IP), or the share of each entity in this ecosystem. Another legal challenge could be related to the operation of the use case. Imagine a scenario where a collision occurs due to a latency (delay) of communication technology between two connected cars. It is crucial to be able to detect whether the telecommunication company, the car manufacturer, the regulatory body, the hardware manufacturer, or the application developer is liable for this malfunc- tion and to what extent. Limited resources IT networks enjoy the deployment of powerful computers and gigabit connections. Con- versely, most IoT devices are simple, inexpensive, and require low power, with limited memory, power, and processing capabilities. IoT sensors often have very little data to transmit. According to James Brehm & Associates, 86 percent of all IoT devices use less than three MB of data per month (CommScope, 2017). IoT devices can have a transmission range of up to several kilometers, resulting in constrained and lossy networks with lower data rates (as low as a few bits per second) in low-power wide-area networks (LPWAN). 19 Human resistance Some employees may fear that their organizational role may be endangered or replaced by IoT solutions and automatization. To combat this resistance, employees should be edu- cated and assured that IoT solutions can create new opportunities and act as an aid, not an obstacle. IoT must overcome all the aforementioned challenges to become a ubiquitous technology in daily life. Moreover, organizations and businesses need to train and hire skilled IoT staff and personnel to take full advantage of the benefits offered by implementing IoT. 1.2 Evolution of the Internet—Web 1.0 to Web 4.0 The World Wide Web (WWW), commonly referred to as the web, is an interconnection of public webpages, available via the internet and connected using hyperlinks. The web is a service offered by the internet and consists of the following elements: The Hypertext Transfer Protocol (HTTP), which is designed to govern data exchange between web servers and web clients. The Uniform Resource Location (URL), or Uniform Resource Identifier (URI), which is used as the web address to access a resource on the internet. The HyperText Markup Language (HTML), which is the standard language used to create and publish web documents and display them in a web browser. The web was first developed by Timothy Berners-Lee in 1989 to enable data sharing between international scientists at the European Laboratory for Particle Physics (CERN) in Geneva, Switzerland (Nedeva & Dineva, 2012). Since 1990, when Berners-Lee developed the first web server, browser, and webpage on his computer at CERN, significant progress has been made, both in terms of the web and its related technology. The web has now gone through several evolutionary stages, with profound impacts on both society and business. The first generation of the web, known as Web 1.0 or syntactic web, was infor- mation-centered and static. The people-centered Web 2.0 enabled human interactions through social media platforms. In the machine-centered Web 3.0, machines were able to read and understand data shared between services and applications. In the agent-cen- tered Web 4.0, artificial intelligence (AI) enables human-machine symbiotic interactions to realize a connected world. 20 Figure 5: Evolution of the Web Source: Reza Vahidnia (2022). Web 1.0: Web of Cognition Web 1.0, the web of cognition or Syntactic Web, was the first generation of the web (1990— 2000). Businesses used it to deliver content and share information via read-only web- pages. In Web 1.0, people were limited to searching the web for information and reading it. The passive nature of static webpages provided users with little to no possibility of interacting with or contributing to webpages via feedback or comments. Information was disseminated to users via a push model. Web 1.0 pages used HTTP as the communication protocol and were developed in HTML (Solanki & Dongaonkar, 2016). The characteristics and limitations of Web 1.0 are summarized in the table below (Khanzode & Sarode, 2016). Table 1: Characteristics and Limitations of Web 1.0 Characteristics Limitations Content is read-only. There is no machine compatible content (only humans can read the web pages). Information is available to anyone at any time. The content of the website is updated and man- aged by the webmaster only. Static webpages are published via HTML. Dynamic representation is not possible. Source: Reza Vahidnia (2022), based on Khanzode & Sarode (2016). 21 Web 2.0: Web of Communication The second generation of the web (2000—2010) was defined in 2004 by Dale Dougherty as a read-write web (Project Information Literacy, 2010). In Web 2.0, also known as Social Web, the idea was to assemble and manage the social interactions of many people across the globe. In essence, Web 2.0 signaled the transition of the internet into a communication platform, a revolution for businesses. In the people-centric Web 2.0, transactions were bilateral, with people participating in creating web content—collaboratively reading and writing on the web—helping gather collective intelligence. Some of the outstanding fea- tures of the Web 2.0 technologies include flexibility in designing, creativity in reusing, and collaboration in modifying web content, as well as social media platforms, such as Linke- dIn, Facebook, and Instagram. Web technologies and standards, such as Extensible HTML (XHTML), JavaScript, Document Object Model (DOM), and Ajax were major drivers of these Web 2.0 features, enabling an interactive experience for users (Solanki & Dongaonkar, 2016). The characteristics and limitations of Web 2.0 are summarized in the table below. Table 2: Characteristics and Limitations of Web 2.0 Characteristics Limitations The platform is web-based. The iteration cycle of changes is constant. Software and businesses can be “architected.” Ethical issues arise. It facilitates social networking and collective Interconnectivity between platforms is limited. knowledge production. Source: Reza Vahidnia (2022), based on Khanzode & Sarode (2016). Web 3.0: Web of Cooperation In the third generation of the web (2010—2020), also known as Semantic Web, a machine readability feature was added to web documents to allow machines to understand, inter- pret, and respond to semantically structured human requests. The common framework provided by semantic web enables data to be used in various applications and shared across communities and enterprises more easily. The emergence of Web 3.0 improved data management, supported mobile internet, and organized social media collaborations. In Web 3.0, data are not owned, but rather shared, by services or application devices rep- resenting different views of the same data. For instance, when a user searches for a spe- cific flight on the web, these data are available and used by other services and applica- tions to show the user other useful information, such as nearby hotels or car rental companies at the destination. The figure below depicts the development of Semantic Web in a layered architectural model, also known as semantic cake, where each layer uses the services of the layers below (Berners-Lee, 2000). 22 Figure 6: Layered Architecture of the Semantic Web Source: Reza Vahidnia (2022), based on Berners-Lee (2000). The following are building blocks of the layered architecture of Web 3.0: URI/Internationalized Resource Identifier (IRI): naming of resources (any things) by a unique string that can be used as an identifier or address. Usually, HTTP URIs are used. Extensible Markup Language (XML): markup language for representing hierarchically structured data in the format of a text file that can be read by both humans and machines Resource Description Framework (RDF): model for formulating logical statements about arbitrary things (resources), which is based on directed graphs. Logical state- ments are triples – which consist of subject, predicate, and object – and relate subject and object via the predicate. Ontology: vocabulary for a specific area of knowledge that describes terms or classes that can be related to each other. They can be arranged in a term or class hierarchy with superclass and subclass. Instances can be created as objects of previously defined terms/classes (e.g., Berlin as an instance of the class city). Ontologies contain inference and integrity rules (i.e., rules for making inferences and guaranteeing their validity) and enable the derivation of new knowledge (new propositions) from existing knowledge by logical reasoning (inference). Logic: formalization of arguments (statements) with the aim of checking the arguments (statements) for their validity. Usually, the first-order predicate logic is used, which is based on the propositional logic. Proof: verification of the automatic conclusions established in the logic layer by search- ing the web for rules and ontologies until the above conclusions can either be con- firmed or disproved using a heuristic engine Trust: use of digital signatures in order to determine whether the proof should be trus- ted or not The characteristics and limitations of Web 3.0 are summarized in the table below. 23 Table 3: Characteristics and Limitations of Web 3.0 Characteristics Limitations Software as a service (SaaS) business model Lack of ability to eliminate semantically duplicated terms Open-source software platform Vague user queries Distributed database Inconsistencies arise when combining various sour- ces Intelligent web The content producer can deceive the consumer Source: Reza Vahidnia (2022), based on Khanzode & Sarode (2016). Web 4.0: Web of Integration In Web 4.0, also referred to as symbiotic web or smart web, intelligent machines can engage with the real world and interact with humans using mind-controlled interfaces. Artificial intelligence, alongside semantic and reasoning technologies, make Web 4.0 serv- ices proactive, autonomous, collaborative, and self-learning. Smart web services, such as virtual reality (VR), access web databases through intelligent agents to present adaptive contents. These software agents can reason and collaborate with other such agents on the internet to perform tasks on behalf of humans. Evolution of the Internet and Web The four evolutionary stages of the internet have some temporal mismatches with the four phases of the web. However, each phase of both evolutions has had significant busi- ness and social impacts compared to the stage before (Guarda et al., 2017). We are cur- rently in the fourth generation of both the internet (Internet of Things) and web (symbiotic web), where an “always on” world can be realized. SUMMARY Over the last decades, there has been significant progress in web tech- nologies. Development began with Web 1.0, where static pages could only be read by humans with little or no interaction, and transitioned to Web 4.0, where (thanks to AI) proactive and autonomous web services collaborate with each other and smart machines read webpages and perform tasks on behalf of humans. Almost concurrently with the evolu- tion of web, the internet has gone through four evolutionary phases, creating a world where people, devices, and processes are connected and can share data with each other to create an Internet of Things eco- system. The data collected by intelligent devices in the ecosystem are analyzed to increase efficiency, safety, transparency, and accuracy, and optimize processes for businesses and people. 24 To benefit from IoT and realize end-to-end solutions, the hardware, con- nectivity, platform, and application components of the IoT value chain are developed separately or collaboratively by various businesses and entities. However, implementing and deploying any IoT solution has its own challenges. Scalability and interoperability are two major chal- lenges that need to be addressed by IoT solution architects because IoT use cases typically require a massive number of smart “things” from var- ious manufacturers and vendors to connect to the internet. Moreover, it is crucial to securely transfer and store the IoT data and make sure the information is only shared and accessed by legitimate entities. Handling and analyzing the vast data collected by thousands of sensors, as well as legal issues related to deploying IoT use cases, are major hurdles affect- ing IoT implementation in daily life. 25 UNIT 2 SOCIAL AND ECONOMIC SIGNIFICANCE STUDY GOALS On completion of this unit, you will be able to … – identify the impact of IoT on consumers, retailers, and industries. – describe the role of IoT in the Fourth Industrial Revolution. – explain how IoT can transform how people live and work. – compare the security vulnerability of IoT systems with traditional IT networks. – summarize the mechanisms used to protect the security and privacy of IoT data. 2. SOCIAL AND ECONOMIC SIGNIFICANCE Introduction Digitizing the real world has unleashed countless possibilities. By remotely sensing the physical parameters and controlling objects, digitization has inspired a wave of innovation and created a vision of a different future for people, businesses, and governments. The Internet of Things (IoT) has the capability to entirely transform and digitize businesses, automate industry processes, and even make them autonomous in certain areas. The evolution of powerful and low-cost connectivity and data analytics technologies will capture the full potential of the Internet of Things and create substantial socio-economic benefits that are based on data-driven decision-making. IoT can improve quality of life in various aspects, from healthcare to safety, to shopping experiences, transportation, and education, to name just a few. Connected devices facilitate businesses in creating new rev- enue models, enhancing productivity, monitoring and tracking their assets, and improving machine performance. Governments can deliver better services to citizens and reduce costs through smart city initiatives. Crop field monitoring, irrigation automatization, and cattle management sys- tems at smart farms can improve both productivity and water management. All these examples prove that IoT innovations will impact almost every aspect of our lives. McKinsey Global Institute estimates that in 2025, the yearly economic impact of IoT appli- cations could exceed $3.9 trillion, both in economic growth and cost savings (McKinsey Global Institute, 2015). Moreover, IoT will generate millions of job opportunities in various social and industrial segments. 2.1 Innovations for Consumers and Industry Consumers and Retailers IoT can disruptively change consumer interactions with retailers and products in many ways. Connected devices assist people by automating household chores and providing more control over products and appliances. It is estimated that in 2025, household IoT applications will generate more than $350 billion in economic impact. Additionally, IoT will reduce the time required for household chores by 17 percent (McKinsey Global Insti- tute, 2015). Imagine a smart refrigerator that knows what groceries are needed, does the shopping, and tracks the delivery of the items on behalf of the consumer. Or a smart dia- per that alerts the caregiver when the baby needs to be changed. 28 IoT will give consumers pleasant online and in-person shopping experiences while provid- ing retailers with better insights about shoppers’ in-store behavior. The analyzed data gathered by video cameras, Bluetooth and Wi-Fi beacons, and motion sensors will give Beacons retailers useful information about customers’ in-store traffic patterns, visitor demograph- These are small, inexpen- sive devices that con- ics and dwell times, as well as products of interest. The insight obtained by this informa- stantly transmit a signal tion, combined with input from external sources and social media, can be used to improve for other devices to store assortments and in-store customer engagement. This intelligence helps retailers to probe. use the store’s space to effectively position products, target segmented groups of custom- ers, and provide them with special offers and coupons through automated marketing campaigns. Analyzing data captured by various sensors in stores also helps retail managers dynami- cally determine rental fees of shelves within stores. Beacons can work as virtual assistants, guiding customers in-store and providing them with personalized recommendations based on their profile information and purchase history. However, these innovative IoT services should only be offered under one strict condition: Services must not violate cus- tomer privacy. Besides in-store improvements, IoT can positively affect behind-the-scenes processes in the supply chain, like manufacturing, logistics, inventory, and replenishments. When all the components of a supply chain, such as inventory, delivery vehicles, and distribution centers are connected to the internet, bottlenecks can be identified and remediated. This increases the efficiency and productivity at all levels of the supply chain. Through analyzing data collected about consumer behavior, social media conversations, Google searches, and marketplace trends, supply chains will have better visibility and can be more responsive and prepared for any unforeseen situations, such as a sudden surge or drop in demand. Consumers and retailers can benefit from a list of retail-relevant IoT inno- vations, as outlined in the following paragraphs ([x]cube LABS, 2020). Smart shelves Fitted with radio frequency identification (RFID) tags and weight sensors, smart shelves Radio frequency identi- can identify misplaced items, detect potential theft, and notify staff when the shelves need fication This refers to a small sys- restocking. These intelligent shelves can also communicate with shoppers’ smartphones, tem that includes a radio informing them about alternative products and special offers. transponder, receiver, and transmitter that are used to identify tags Automated checkout attached to objects. Automated checkout allows the tagged contents of shopping carts to be scanned and paid for using customers’ mobile payment applications as they exit the store, avoiding cus- tomer frustration and waiting in long lines. With automated checkout, it is estimated that cashier staff requirements could be reduced by 75 percent by 2025, resulting in $150— $380 billion savings. By 2025, automated checkout is expected to cut checkout wait times by 40—80 percent, potentially leading to an economic benefit of $30—$135 billion of saved time (McKinsey Global Institute, 2015). 29 In-store layout optimization IoT sensors and analytics software can determine shoppers’ in-store behavior and help retailers optimize their shop’s physical layout. It is estimated that layout optimization of stores could increase productivity by five percent by 2025, resulting in a total potential economic impact of $79—$158 billion (McKinsey Global Institute, 2015). Smart signage On-site digital displays enhance the in-store customer experience by delivering dynamic promotions and targeted, location-based advertising in real-time. Personalized discounts IoT sensors can determine how many times a customer who has signed up for a loyalty program has entered a venue and what they have looked at during their in-store and online visits. To increase purchases, beacons can communicate with customers’ cell- phones when within range and send personalized offers and discounts to the previously installed application. Supply chain management IoT sensors, such as temperature, humidity, vibration, GPS, and RFID tags obtain useful information about how goods and products are handled and their location in the supply chain. The processed information is then used to deliver higher quality products to the customers, faster. Smart inventory Real-time monitoring and optimization of inventory levels is important when attempting to reduce costs and ensure retailer profitability. It is estimated that by 2025, when tradi- tional rule-based and periodic reordering is replaced with automatic IoT technologies and inventory replenishment, there will be a yearly economic impact of $5—$15 billion (McKin- sey Global Institute, 2015). Customer satisfaction IoT sensors can monitor and analyze customer behavior and capture feedback to provide fresh insights for businesses and industries about consumers’ experiences with shopping or consuming a specific product. Industrial Internet of Things and Connected Factories Industry has gone through several revolutionary phases since Industry 1.0, when steam power was first applied to mechanical production in the late eighteenth century. The late- nineteenth and early-twentieth centuries experienced the second revolutionary stage of industry, Industry 2.0, in which electric power led to mass production facilities. The third generation of industry began with the emergence of computers and IT networks in the 30 early 1970s, when industrial operations became automated. In 2016, IoT was considered the main driver of the Fourth Industrial Revolution by the World Economic Forum (Schwab, 2016). In Industry 4.0, intelligent machines equipped with smart sensors can communicate with each other, resulting in connected factories where decisions are made intelligently, failures are predicted, and machines have self-healing features to minimize human intervention. The figure below shows a summary of the four industrial revolutions. Figure 7: The Industrial Revolutions Source: Reza Vahidnia (2022), based on Vecteezy (n.d.-a); Vecteezy (n.d.-b); Vecteezy (n.d.-c); Vec- teezy (n.d.-d). In traditional, disconnected industrial enterprises and factories there is little visibility into operations that could cause inefficiencies and safety issues. This is because different com- ponents, such as plant floors, suppliers, and front offices operate in independent silos. To resolve these issues, the Industrial Internet of Things (IIoT) retools factories with IoT tech- nologies to enhance operational productivity and reliability through interconnecting smart sensors, actuators, and robots with industrial control systems such as Human Machine Interfaces (HMIs), Supervisory Control and Data Acquisition (SCADA), and Pro- Human Machine Inter- grammable Logic Controllers (PLCs). faces These are the key points of contact between oper- Utilizing smart sensors, internet-protocol-enabled (IP) machines, cameras, and robots ators and processes or transmitting real-time diagnostic and informational data would yield improvements in the devices. Supervisory Control and efficiency and reliability of industrial operations. Machine-to-human connections in a Data Acquisition smart factory bring real-time sensor data directly to the operators, saving them time as This is a combination of there is no need for movement between the plant floor and central control room. More- software and hardware used to control industrial over, immediate decisions can be made to fix any production line problems, which in turn processes. 31 increase productivity. Predictive maintenance scheduling is another important feature of IIoT in smart factories. For instance, the IoT-enabled air compressors of the German com- pany Kaeser can predict when maintenance is required (Plant Services, n.d.). Implementing real-time location systems (RTLS) through the attached RFID tags in smart factories facilitates communication between items on assembly lines and the network about their status and location. The information captured from these tiny sensors is ana- lyzed to identify bottlenecks in production lines, helping operators take appropriate action immediately. In large corporations, increasing productivity is typically the main purpose of deploying IIoT technology. Industrial IoT, however, can bring many other benefits for small and medium-sized enterprises (SMEs), creating value and new business models in the follow- ing dimensions (Wylde et al., 2020): improved products by making existing products “smart” and creating new data services and products improved customer service through integration with dealers improved engineering by reducing the time required to develop an idea and introduce it to the market improved operations through optimized planning and integration with suppliers more efficient management through production asset intelligence and activity synchro- nization Although IIoT brings many significant benefits for industries in Industry 4.0, it comes with its own challenges. Cyber security threats are a major barrier to adopting IIoT technology, as manufacturers fear losing vital business data to competitors. 2.2 Implications on People and the World of Work Over the past few decades, industrial technologies, innovations, and advancements have transformed people’s lives for the better. In a connected world, where IoT-enabled devices sense the environment and transmit the information to advanced data analysis tools, more and more innovative approaches can be introduced to further improve human wel- fare and address existing challenges, from climate change and air pollution to safety and healthcare issues. IoT Implications on People and Lifestyles Deploying IoT technologies has numerous positive impacts on people’s quality of life at home, work, and on the go. The following are some areas where a collection of connected “things” enhanced with artificial intelligence can play a key role in improving people’s lives. 32 Smart city Urban life can be made more attractive, safe, and convenient by using IoT technologies. Smart cities deploy numerous smart technologies to support their infrastructure, includ- ing smart street lighting and energy efficient buildings, smart waste management and on- demand waste pick-ups, smart snow removal, smart traffic control, smart public transpor- tation with digital bus routes and smart parking meters, smart irrigation for parks and fountains, and environmental and air pollution monitoring systems. These IoT-enabled services increase efficiency in managing cities, reducing traffic jams, improving air and water quality, and optimizing energy use. Key elements of smart city services, where infor- mation and communications technology (ICT) providers can play a role, are illustrated in the graphic (Groupe Speciale Mobile Association [GSMA] Smart Cities, 2013). Figure 8: Key Elements of Smart City Services Source: Reza Vahidnia (2022), based on Groupe Speciale Mobile Association Smart Cities (2013). These services are facilitated by ICT providers, such as mobile network operators, and are offered through the collaboration of different stakeholders, including the government, citizens, industry, and scientific centers. These smart city stakeholders are shown in the next figure (GSMA Smart Cities, 2013). The city of Yinchuan, China is one example of a smart city, where many of the aforementioned initiatives have already been deployed. Many other cities across the world have also installed city-wide sensors and infrastructure for various IoT purposes. 33 Figure 9: Smart City Stakeholders Source: Reza Vahidnia (2022), based on Groupe Speciale Mobile Association Smart Cities (2013). Environmental health Chronic water and air pollution in big cities endanger the lives of millions of citizens. IoT- enabled sensors can monitor air and water quality in certain areas and alert citizens when the quality reaches dangerous levels. Analyzing the sensor’s data with advanced anomaly detection machine learning algo- rithms can help identify the source of pollution. For example, Drayson Technologies has installed a network of environmental monitoring sensors on a fleet of vehicles and bicy- cles in London, UK, where hundreds of air pollution-related deaths are reported yearly. The data collected by these sensors are transmitted to cellphones via Bluetooth, allowing the company to generate a heat map illustrating the city’s air quality in real time (Stone, 2017). There are other innovative actions in place around the world: A California-based start-up called Aclima partnered with Google to equip its fleet of Google Street View cars with special environmental monitoring sensors, creating an air pollution map of Oakland, California (Aclima, 2021). Smart agriculture Deploying IoT technology, such as soil sensors and drone imagery in agriculture, can effec- tively reduce fertilizer and water consumption by up to 40 percent and improve product quality for consumers (Kranz, 2018). Moreover, continuously monitoring and tracking agri- cultural products throughout the supply chain, from the farm to the grocery store, minimi- zes crop loss and boosts productivity. IoT technology improves operations management, providing farmers with useful data about markets and therefore facilitating waste reduc- tion. 34 Healthcare Smart wearables and connected health devices are finding more and more applications in our daily life. They collect and analyze health-related data, such as heart rate, pulse, and blood pressure; monitor sport and daily activities; track patients; and are used to monitor environmental parameters for safety issues. In some cases, wearables are also capable of making smart decisions and providing users or healthcare providers with a response. The figure below shows some smart wearables developed for various applications (Dian et al., 2020). Figure 10: Smart Wearables Source: Reza Vahidnia (2022), based on Flaticon (n.d.-d); Flaticon (n.d.-l); Flaticon (n.d.-m); Flaticon (n.d.-s); Flaticon (n.d.-t); Flaticon (n.d.-v). IoT technology is widely used to address profound health challenges. For example, IoT sensors and cameras are used to curb COVID-19 outbreaks by measuring temperatures, detecting face masks, and tracking infected patients (Peerzade, 2021). Early diagnosis of breast cancer is another example. Sensors embedded in breast patches can monitor any temperature change in breast tissue and transmit data to the wearer’s cellphone (Maddox, 2017). The data are shared with the healthcare provider and any abnormal patterns in temperature that could indicate early-stage breast cancer can be identified using machine learning and predictive analysis. Beside these positive implications and opportunities, the advent of IoT could pose signifi- cant security, safety, and privacy challenges to people’s lives. Since most IoT sensors are exposed to the real world, they are vulnerable to security threats, where attackers have access to the real connected world. If IoT devices and communications are not secure enough, malicious users may interrupt the normal operation of smart city traffic lights or even disrupt an implanted heart pacemaker. Moreover, people are concerned about their 35 personal information being shared with unauthorized entities. With the emergence of brain-computer interfaces, hacking a person’s brain could be a major security and privacy issue. Many people and communities are also worried about the electromagnetic safety hazards of IoT devices and the effect of radio frequency (RF) radiation on their mental and physical health. For example, in smart schools, parents might be worried about the RF radiation of the Wi-Fi modems and the adverse health effects on their children. Residents of apart- ment buildings may be reluctant to have smart meters installed in their buildings. There- fore, before implementing and adopting any large scale IoT solutions, people should be educated about the implications of IoT on their daily life so they can be aware of all possi- ble risks. It should be made clear that any technological opportunity entails its own risks, and IoT is not exempt. However, because of the progress of IoT technologies and strat- egies, the benefits of IoT far outweigh the potential risks. These are mitigated to the point that it is costlier not to use IoT technology (Rainie & Anderson, 2017). IoT Implications on the World of Work As with all disruptive and rapid technology changes, businesses and companies need to pay special attention to deploying IoT if they want to stay agile. Otherwise, the risk of fall- ing behind competitors using IoT is high. According to International Data Corporation, it is estimated that worldwide business spending on IoT hardware, software, services, and connectivity needed to launch IoT solutions will exceed $1.2 trillion in 2022 (as cited in Randstad, 2020). With IoT, businesses can create new revenue streams and transform their operational models. However, the deploying IoT sensors in business brings its own security and pri- vacy implications. Therefore, it is crucial that businesses take proper security measures and educate staff about possible issues arising from using IoT in the workplace. IoT provides organizations and businesses with the means to engage with their employees by supporting well-being, improving health and fitness, and strengthening morale using wearables. In business, IoT can also help recruiters find the talent and capable employees they need through psychometric assessment methods or gamification. IoT’s influences on the business world are listed below (Ganatra, 2020). Overwhelming data The huge amount of data generated by smart devices in the workplace requires busi- nesses to invest in their computing and cloud storage capabilities and enhance their big data analysis tools. Enhanced need for security Data are considered the currency of the future and need to be protected by cybersecurity experts against any type of breach. However, compared to demand, there is a lack of such cybersecurity professionals, meaning that organizations must train special staff for this purpose. 36 Wearables IoT sensors worn by employees can transform an office environment by empowering staff with mixed reality and improve employee safety by constantly monitoring environmental parameters and individual health signals. Wearables can enhance teamwork by creating virtual offices using augmented reality (AR) and virtual reality (VR), allowing employees in different locations to collaborate more effectively. Microsoft’s HoloLens smart glasses are an example of such wearables used in office spaces to display information that is mixed with the real world. HoloLens can also simulate a virtual world by using advanced optics and IoT sensors, including depth; video and red, green, blue (RGB) cameras; gazing and gesture sensors; as well as holographic technology (Microsoft, n.d.). 2.3 Data Protection and Data Security In traditional IT networks, data reside in physically secure infrastructure and on enterpri- ses’ dedicated servers. In scalable IoT initiatives—whether it is smart cities, healthcare, or any other vertical—massive amounts of data need to be collected from a wide range of sensors exposed to the real world. These data are processed at the edge of the network, stored in clouds and distributed databases, analyzed, interpreted, and presented by big data tools and technologies. Security threats in IoT systems may be like those in tradi- tional IT networks; however, considering the scale and diversity of IoT devices, communi- cation media, protocols, platforms, and software applications, the overall impact of a security breach in IoT could be much more significant than IT networks. Any breach, damage, or misuse of IoT data may have devastating and negative impacts on the individuals, organizations, or even entire critical infrastructures and social sectors. Therefore, it is crucial to deploy a scalable and robust security mechanism to protect the data when they are collected by sensors, processed at the edge, stored in databases, and analyzed by IoT platforms, and secure the data when they are shared with other applica- tion software and entities. In other words, when designing an IoT system, end-to-end security should be considered and addressed throughout the entire device life cycle, from initial design through operational environment. The figure shows the layer model of an end-to-end IoT solution, as well as the required security measures at each layer (IoT Ana- lytics, 2015). 37 Figure 11: Layer Model of End-to-End Internet of Things Solutions Source: Reza Vahidnia (2022), based on Internet of Things Analytics (2015); Flaticon (n.d.-q); Flaticon (n.d.-t); Flaticon (n.d.-x). In layer one (Things Layer), it is important to physically protect the IoT hardware, making it tamper-proof. Devices must be authenticated and authorized before being connected to the network. IoT devices should be assigned proper access permission using techniques, such as biometrics, grid codes, one-time passwords (OTPs), or security tokens. In layer two (Communication Layer), the data transmitted in the communication medium should be encrypted end-to-end to ensure that information cannot be altered or accessed by unauthorized users while in transit or stored in memories. There are many data encryp- tion techniques available, such as private-key cryptography (symmetric) and public key cryptography (asymmetric). The privacy of data stored in databases on layer three (Cloud Services Layer) should be carefully managed and protected. Moreover, users (people and applications) requesting access must be identified to avoid unauthorized access to layer four data (Application Layer). In general, IoT systems are highly vulnerable to security risks and usually more dif- ficult to secure compared to traditional IT networks due to the following reasons: Most IoT sensors are highly mobile, dynamic, and have no defined perimeter. Inexpensive and simple IoT devices with limited power and computing resources may not have the complex capabilities to easily deploy the existing security mechanisms, such as authentication, end-to-end encryption, and role-based access control. Con- versely, given the large amount of IoT data, protecting the core IoT network with sophis- ticated security techniques may result in network congestion, making the network vul- nerable to denial-of-service attacks. IoT devices, communication technologies, application protocols, and platforms are het- erogeneous, with diverse security requirements. 38 IoT devices may not be Internet Protocol (IP)-enabled and therefore cannot connect to the internet. IoT devices can be exposed to the real world and not physically protected. IoT objects can be controlled and managed by third parties. The IoT metadata collected by wearables and medical well-being sensors on the body are typically rich enough to infer individuals’ activity, personal interests, location, habits, and preferences. Once these sensitive data are collected by the sensors, they are uploaded to the cloud or transferred to other devices, such as mobile phones, and then forwarded to third parties. Therefore, it is crucial to protect the privacy of data at rest (when stored in databases), in transit (while it is being transmitted in the communication medium), and when being accessed by users and applications. This ensures that the individuals’ identi- ties are not compromised and only authorized users and applications have access to the data. Users should be assured that their personal data are fully protected while they enjoy the benefits of IoT in their daily life. Otherwise, privacy issues could be a huge barrier to adopting IoT. For example, people may refuse to buy and drive connected cars if they are unsure which third parties have access to their location, driving habits, and other informa- tion, and for what purpose. For businesses and organizations, the IoT data can reveal profit margins, supply chain sources, and vital information about staff and business models. Therefore, it is essential for both individuals and businesses that their data are fully protected by all entities and platforms involved in data acquisition, management, and analytics. IoT systems should have proper measures and access control techniques in place to authenticate the users and applications to determine who has the right to access the data and at what level. Cyber security threats and their severity have grown over the years. This means that more advanced, dynamic, and sophisticated data protection techniques are required to provide end-to-end security for IoT systems. In other words, given the complexity of IoT systems, there is no static, “one-size-fits-all” security and privacy technique to address the different cyber security risks for various IoT solutions. Instead, security and privacy techniques should be robust, on-demand, and customized to protect data at rest or in transit in differ- ent IoT systems. For example, in healthcare IoT solutions, medical devices and wearables must be authorized and authenticated carefully, even if this process takes a long time. Conversely, in a vehicle-to-vehicle IoT initiative where several authentication processes occur simultaneously, this operation should be fast and efficient to support real-time response, which is critical in such a use case. Therefore, the authentication overhead at both the transmitter and receiver vehicles should be minimized by taking advantage of Graphics Processing Units (GPUs) used in chipsets installed in the vehicles. In many IoT solutions, special attention should be paid to minimize the power consumed by devices. For example, a drone flying over a smart farm capturing data from multiple sensors needs to save energy and cannot wait too long for sensors to be authenticated and generate encryption keys. Therefore, special security techniques should be designed and developed for such use cases so that sensors can start generating encryption keys while the drone is approaching. These examples show that there should be a careful trade-off between the security measures deployed in IoT devices and other factors, such 39 as cost, energy consumption, and response time. Expertise and a careful analysis of the IoT solution, its security, and privacy requirements are needed to ensure the optimal use of IoT hardware when deploying security techniques. Requirements of IoT Data Security and Privacy In IoT systems, data surface is expanded from the cloud to IoT gateways and aggregation nodes at the fog level, and sensors/actuators at the edge of the IoT network. The figure shows the IoT stack from the data management and computer perspective. The most time-sensitive data are typically processed at the edge and less time-sensitive data are passed to the cloud for big data analytics and long-term storage. Therefore, the data could be at rest in the databases or in transit in any communication medium between the edge, fog, and cloud levels. Figure 12: Internet of Things—Data Management and Computer Stack Source: Reza Vahidnia (2022). 40 The data generated by a single sensor may be of little value. However, when data from var- ious sensors are combined, they will be valuable and vulnerable to cybersecurity and pri- vacy threats. To protect sensitive IoT data across the IoT data management stack, the fol- lowing requirements must be met (Bertino, 2016): confidentiality. Unauthorized users must be denied access to the data. Cryptographic techniques are used to digitally sign the data to make sure that they can only be accessed by authorized users. However, cryptographic techniques require high comput- ing capabilities which may not be available in simple, inexpensive IoT devices. integrity. Access control techniques are used to ensure that only authorized users can modify data. availability. Authorized users should be assured that data are available for them when they need them. trustworthiness. This ensures that data are up to date, free from error, and originated from reputable sources. Data quality techniques are used to automatically detect and fix data errors. Provenance methods detect the origin and source of the data, and repu- tation techniques assess the reputation of data sources. Protection of IoT data and compliance with legal and technical standards can involve some issues and implications for people and communities. There could be a transparency concern that the captured data are further processed and used for other purposes than those initially stated. Some people may be concerned about losing control over data pro- cessing in the IoT devices, gateways, and the cloud. Another issue is the difficulty in get- ting individual and company consent to use data for various purposes. Lack of transpar- ency and granularity in IoT services may lead to entire service refusal if the users do not consent to a specific data processing aspect of the IoT system. SUMMARY Digital transformation and IoT will significantly impact almost every aspect of people’s lives. With IoT innovations, online and in-person shopping experiences are more efficient and productive: They are more convenient for consumers, retailers can manage their stores better, they can gain more insights about the personal interests and in-store behav- iors of their customers, and industries are automated. Smart cities improve quality of life through effectively controlling traffic and streetlights, monitoring the environment, and managing waste and snow removal. Wearables and medical IoT devices can constantly moni- tor individuals’ vital health signals, such as heart rate, pulse, or blood pressure, to improve their well-being. In smart farms, IoT sensors moni- tor soil condition, manage irrigation, and track products to save water and cut waste. 41 Deploying IoT in companies and enterprises also has implications for the workplace. It will enable businesses to create new revenue streams and transform their operational models. Applying IoT in various aspects of people’s lives comes with some chal- lenges, including data security and privacy. Any breach or misuse of IoT data may have significant negative impacts on individuals, businesses, or communities that could otherwise benefit from IoT. People and busi- nesses should be assured that their IoT data are carefully protected against cybersecurity attacks and are not shared with unauthorized users and applications. 42 UNIT 3 COMMUNICATION STANDARDS AND TECHNOLOGIES STUDY GOALS On completion of this unit, you will be able to … – list the characteristics of different network topologies used in IoT networks. – identify the application protocols used for IoT communication. – describe the specifications of various IoT access technologies. 3. COMMUNICATION STANDARDS AND TECHNOLOGIES Introduction The Internet of Things (IoT) networks’ capabilities are typically limited compared to IT net- works. Most wireless IoT networks are lossy and constrained, which significantly restricts the speed of data exchange. Conversely, the massive number of IoT nodes and sensors require special provisioning and configuration which may increase the cost of connectiv- ity. Therefore, the IoT networks must be designed in such a way that connectivity fees are minimized. In some IoT use cases, such as mining and forestry, the IoT sensors need to be deployed in rural and remote areas where access to electricity or regular cellular commu- nication may be limited. Use cases like these require low-power communication technolo- gies that can cover wide areas. The inexpensive and tiny IoT sensors have limited processing and power capabilities, which restrict their ability to use any communication technology and protocol. As a result, various communication standards, technologies, and protocols have been developed to address the IoT needs. Some of these lightweight technologies and protocols are suitable for massive IoT use cases where reducing the hardware and connectivity costs is crucial. Conversely, other technologies are designed for mission-critical IoT, such as drone control or automatic guided vehicles (AGV), where the reliability and availability of the network is imperative. An IoT network architect should be aware of the advantages and disadvan- tages of different technologies, standards, and protocols to adequately select the right technology for each use case. 3.1 Network Topologies In construction projects, the architect must adhere to certain standards and plan the building based on whether it is used for residential or commercial purposes. In a similar vein, an IoT network architect needs to carefully plan the IoT network and consider the security policies and design practices based on the requirements of the IoT use case and application. Failure to carefully design the network may result in difficulties when manag- ing, troubleshooting, scaling, and adapting it. How the various elements of an IoT network communicate with each other is decided by the network topology. Different communication topologies have their own specifications in terms of implementation cost and complexity, power consumption, and reliability. The IoT use case determines these specifications and consequently the network topology that is most suitable for the IoT sensors and elements to communicate with each other and the cloud. 44 A peer-to-peer IoT network can be established when both nodes have the capabilities to implement all protocol stack functions. Nonetheless, in many IoT use cases, such as room temperature monitoring, the inexpensive IoT sensors with limited network capabilities transmit the collected data to a central node, called the gateway. The gateway oversees the overall network coordination and takes care of the full protocol stack functions, such as security, overflow control, and congestion control. In such a scenario, the sensor, called a Reduced Function Device (RFD), implements a subset of protocol functions to perform the simple task of communicating with the gateway. The RFD cannot be a coordinator, nor can it establish a direct communication with other RFDs in the network. Different commu- nication technologies offer two distinct connectivity structures, as explained in the subse- quent paragraphs. Point-to-Point Communication In a point-to-point communication structure, a single node or sensor can exchange data with only one gateway. Put simply, a dedicated link is provided between the two devices (the smart object and the IoT gateway), which allows an individual session to be estab- lished between them. A common example of such communication type is a TV remote control. Point-to-Multipoint Communication In a point-to-multipoint connectivity structure, which is used in most IoT applications and technologies, a single node (the gateway) can communicate with multiple nodes of the network at the same time. In other words, the capacity of the transmission medium and channel is shared between different IoT nodes. In a spatially-shared connection (fre- quency division), several IoT nodes can use the communication channel at the same time. In a time-shared connection (time division), the IoT nodes need to take turns to communi- cate with the gateway. Figure 13: Communication Structures Source: Reza Vahidnia (2022), based on Flaticon (n.d.-j); Flaticon (n.d.-n). 45 The physical topology of a network is the way in which a network is physically laid out with two or more links connecting multiple devices to each other. This means that the network topology represents the geometric relationship of the links and nodes of the network (the organization of the nodes). The four famous network topol- ogies used in IT networks include mesh, star, ring, and bus. Mesh and star topologies are the two most common topologies in IoT networks (BehrTech, 2020). Mesh Topology In a mesh network topology, there is a point-to-multipoint communication link between a device and other devices in the network so that the data can hop from one node to another before it reaches the gateway. The sensor nodes in a mesh network can serve either as an endpoint, to transmit their own captured data, or repeaters, which relay data received from other nodes or sensors. In a full mesh network, all homogeneous nodes are fully interconnected to each other and have relaying function. In a full mesh network, all nodes can implement the full protocol functions and serve as a coordinator. In partial mesh networks, only selected nodes have the repeater function to relay the mes- sages. The RFD sensor nodes are the endpoints that can only transmit their data to the sensor nodes with relaying functions. Figure 14: Mesh Network Topology Source: Reza Vahidnia (2022), based on BehrTech (2020). Full mesh topology is uncommon in IoT networks, because in most IoT applications, the messages captured by the endpoint sensors do not need to be distributed among other nodes. Rather, the data should be transmitted to the destination (gateway). Moreover, since in full mesh topology all sensor nodes need to be in their range, the effective cover- age of the network is limited. 46 Many short-range IoT access technologies, such as ZigBee and Z-Wave, benefit from mesh topology to extend their coverage (BehrTech, 2020). For example, while the maximum transmission range of a ZigBee module could be tens of meters, an area as large as a com- mercial building or industrial field can be covered using mesh topology and connecting multiple repeaters and relaying nodes. The range extension usually comes at the price of lower data rates because it takes time for the intermediate nodes to relay the messages. For larger areas, the installation costs of mesh networks will be expensive due to the many relaying nodes required to implement the network. Moreover, managing and maintaining such large mesh networks could be another challenge to scaling networks. Since the relay- ing nodes should be constantly awake to listen to the messages they receive, the power consumed by such networks can be high. Another disadvantage of a large mesh network is its vulnerability to security attacks, where breaching a single relaying node (a possible point of attack) of the network may endanger the entire network. An important quality of short-range IoT technologies (e.g., ZigBee or Z-Wave) that use mesh network is their self-healing capabilities, which allows the data to be rerouted through other links if a relaying node fails. Self-healing enhances the robustness of the network. Mesh topology is usually suitable for consumer marketplaces or IoT use cases like smart homes, where the required coverage and the number of sensors are limited. For example, Wi-Fi mesh can be used to extend the coverage of the Wi-Fi signal in a smart home applica- tion. Star Topology In a star network topology, which is simpler and cheaper to deploy than mesh topology, every sensor node establishes a point-to-point communication link with a central coordi- nating node (i.e., a hub or gateway) that typically has the full protocol stack functions. The following figure shows the star topology. 47 Figure 15: Star Network Topology Source: Reza Vahidnia (2022), based on BehrTech (2020). Star topologies usually have higher security than mesh topologies because the sensor nodes are not directly connected to each other. Basically, the communication between the endpoints is coordinated by a sophisticated gateway that can be well equipped with all necessary security mechanisms. One disadvantage of star topology is its dependence on a single hub. If the gateway or the hub fails, the entire network will collapse. Another drawback of the star networks is that the coverage is limited to the maximum transmission range of the gateway and sensor nodes. However, thanks to low power wide area (LPWA) IoT access technologies, the cov- erage of the star networks can extend to tens of kilometers in rural and line-of-sight areas. In LPWA star networks, the endpoint devices can go into deep sleep mode, which saves them considerable amounts of energy to the point that a cell coin battery can run a small IoT sensor for several years. This is in contrast to mesh topology, in which sensor nodes must always be awake and ready to relay the messages of other nodes. Star topology is suitable for massive IoT, where thousands of smart devices distributed over geographically dispersed locations should be connected to the internet (e.g., oil and gas, mining, or forestry use cases). Scaling, management, and configuration of large star networks is much easier than mesh networks. 3.2 Network Protocols IoT network protocols are a set of rules used by IoT applications to ensure that the mes- sages transmitted by an IoT smart object are read and understood by another IoT device or node. IoT protocols also ensure optimum security of the exchanged information between 48 the connected devices. There are different application-layer IoT protocols depending on the IoT devices and their applications. The application layer is an interface between the user and the IoT smart object. A certain IoT use case may require a specific application layer data communication protocol and an IoT-centric communications stack (Vahidnia & Dian, 2021). Different IoT use cases have different requirements. Therefore, only certain protocols from the several available IoT protocols may be suitable for each use case. For instance, some protocols are specially developed to meet the requirements for fast and reliable business transactions, while others are lightweight, low-power protocols optimized to meet the requirements of data collection, such as sensor updates in constrained networks (Naik, 2017). Each IoT protocol has its own advantages and disadvantages when dealing with dif- ferent IoT use cases. Three common and key IoT protocols are Message Queue Telemetry Transport (MQTT), Constrained Application Protocol (CoAP), and Advanced Message Queuing Protocol (AMQP). Some IoT applications also use the HyperText Transfer Protocol (HTTP). However, since HTTP is a heavy and power-hungry protocol, it is more suitable for communication on browser-based clients that typically function on electronic devices with high capabili- ties (e.g., smartphones). Figure 16: loT Network Protocols Source: Reza Vahidnia (2022). Message Queue Telemetry Transport (MQTT) Protocol MQTT is a light, low-power and highly scalable publish/subscribe (also called pub/sub) IoT network protocol. In MQTT, the messages the publishers send are relayed through a cen- tral server, called the broker, to the subscribers. MQTT is the ideal protocol for tiny IoT smart objects with limited bandwidth and power capabilities, such as temperature sen- sors and water flow meter applications. This lightweight protocol is mostly used for con- strained networks that provide connections to IoT devices in remote areas. Unlike HTTP where the client needs to pull the data from the server, in MQTT the broker (server) pushes the data to the subscribers (clients). Another difference between HTTP and MQTT is the speed of the protocol. MQTT is faster and has less overhead. Additionally, the power consumption of MQTT is lower than HTTP. 49 MQTT supports multiple publishers and subscribers. Scaling the IoT network is straightfor- ward, and new devices can be easily added to the network without needing to touch or change the existing infrastructure. MQTT is a device agnostic protocol. This means that IoT devices from various vendors and manufacturers with different communication standards do not need to be compatible with each other to communicate. They join the network through the broker and all communications are done there. Figure 17: MOTT Protocol Source: Reza Vahidnia (2022), based on Flaticon (n.d.-j); Flaticon (n.d.-q). To better understand the performance of MQTT, consider the broker as a post office. IoT devices publish their new messages directly to the post office (broker). Then, the post office (broker) forwards the message to other IoT devices (subscribers) that need a copy of the message. In MQTT, the topic plays the role of the address in the post office scenario. In other words, IoT devices that want to receive a specific message subscribe to the topic. The MQTT has case sensitive topics that are simple strings that can have hierarchy levels separated by a slash (e.g., home/room/sensors/temperature). In MQTT, the interconnection of the IoT objects is through the broker, i.e., there is no direct communication between the IoT devices in the network. The broker acts like a hub and relays the messages between the publishers and subscribers. In this bidirectional and many-to-many communication model, multiple IoT devices may publish/subscribe to dif- ferent topics that are handled by the broker. This communication model allows the IoT smart objects to transmit their sensor data and receive the control and configuration com- mands at the same time. MQTT quality of service In MQTT, three different levels can be defined to ensure the quality of service (QoS) and message reliability (Vahidnia & Dian, 2021). QoS level zero. In this unreliable fire-and-forget model, the messages are delivered with no duplication or acknowledgment. 50 QoS level one. In this reliable model, the messages are delivered at least once with pos- sible duplications. QoS level two. In this reliable model, data are delivered without duplications. MQTT connection As mentioned earlier, MQTT is an application layer protocol. In a Transmission Control Protocol/Internet Protocol (TCP/IP) network layer model, the application protocols use TCP/IP network layer the services from the layer below (i.e., the transport layer). MQTT is basically a TCP/IP- model These are a widely used based protocol that uses the TCP services. Hence, MQTT is considered a lossless, reliable, set of reliable and con- and connection-oriented protocol where a TCP connection first needs to be established nection-oriented commu- before data can be transferred between the two nodes in the network. Once the data nication protocols on the internet. transfer phase is complete, the connection is torn down (terminated). The figure below explains the publisher-broker three-phase handshake communication in MQTT. Figure 18: MOTT Publisher-Broker Handshake Source: Reza Vahidnia (2022), based on Flaticon (n.d.-j). 51 Security of MQTT Most simple and inexpensive IoT devices do not have the necessary network capabilities to establish a secure connection. In MQTT, the burden of securing the connection is on the shoulders of the broker. In the MQTT architecture, the broker does this by decoupl

Introduction to the Internet of Things PDF

Document Details

Tags

Related

Summary

Full Transcript

Upgrade to continue