Introduction

This portion of the book spans the contents of the entire book. It is designed to provide readers with an idea about the various concepts required for utilizing IoT in real life. We present, here, various conceptual questions, which one comes across in real life; they can be answered using the concepts, protocols, and methodologies covered in the previous chapters of this book. This portion also provides an informal guideline for teachers and instructors to formulate questions relating to IoT and its technologies.

The questions provided in here are significantly different from the ones covered at the end of each chapter. They require the reader to delve deeper into the concepts and fundamentals of the covered technologies, which will require some effort on the reader's part. At times, the readers will have to consult other books, online resources, and materials. This exercise will enable them to explore, beyond the introductory level, the various concepts covered in this book.

Questions

• Q1 What are the roles of end-users and the service provider in cloud computing?

• Q2 What is a hypervisor?

• Q3 Let there be a task of 480 TB from a host machine that needs to be processed in the cloud. The service provider has three data centers (DC), DC1, DC2, and DC3, which are capable of processing the task. The following table describes the capacity of different data centers: To which of the DCs do you think it best to allocate the task? Justify your answer.

• Q4 Let there be four fog nodes F1, F2, F3, and F4, connected with several IoT devices. The data generated from the IoT devices can be processed either in single or multiple fog nodes. The fog nodes, F1, F2, F3, and F4, can accommodate up to 20 GB, 90 GB, 28 GB, and 46 GB respectively. Further, these fog nodes, F1, F2, F3, and F4 take 12 μs; 18 μs; 28 μs and 22 μs to complete 10 GB of task. A cloud architecture is also available to execute different tasks; it can process 20 GB of task in 8 μs. An IoT device requires to process 246 GB of data. The time taken to transmit data packets to the cloud is longer than the time taken to transmit data packets to fog nodes. However, the device is associated with a critical application.

Learning Outcomes

After reading this chapter, the reader will be able to:

• List common connectivity protocols in IoT

• Identify the salient features and application scope of each connectivity protocol

• Understand the terminologies and technologies associated with IoT connectivity

• Determine the requirements associated with each of these connectivity protocols in real-world solutions

• Determine the most appropriate connectivity protocol for each segment of their IoT implementation

Introduction

This chapter outlines the main features of fifteen identified commonly used and upcoming IoT connectivity enablers. These connectivity technologies can be integrated with existing sensing, actuation, and processing solutions for extending connectivity to them. Some of these solutions necessarily require integration with a minimal form of processing infrastructure, such as Wi-Fi. In contrast, others, such as Zigbee, can work in a standalone mode altogether, without the need for external processing and hardware support. These solutions are outlined in the subsequent sections in this chapter.

IEEE 802.15.4

The IEEE 802.15.4 standard represents the most popular standard for low data rate wireless personal area networks (WPAN) [1]. This standard was developed to enable monitoring and control applications with lower data rate and extend the operational life for uses with low-power consumption. This standard uses only the first two layers—physical and data link—for operation along with two new layers above it: 1) logical link control (LLC) and 2) service-specific convergence sublayer (SSCS). The additional layers help in the communication of the lower layers with the upper layers. Figure 7.1 shows the IEEE 802.15.4 operational layers. The IEEE 802.15.4 standard was curated to operate in the ISM (industrial, scientific, and medical) band.

The direct sequence spread spectrum (DSSS) modulation technique is used in IEEE 802.15.4 for communication purposes, enabling a wider bandwidth of operation with enhanced security by the modulating pseudo-random noise signal. This standard exhibits high tolerance to noise and interference and offers better measures for improving link reliability. Typically, the low-speed versions of the IEEE 802.15.4 standard use binary phase shift keying (BPSK), whereas the versions with high data rate implement offset quadrature phase shift keying (O-QPSK) for encoding the message to be communicated. Carrier sense multiple access with collision avoidance (CSMA-CA) is the channel access method used for maintaining the sequence of transmitted signals and preventing deadlocks due to multiple sources trying to access the same channel.

Learning Outcomes

After reading this chapter, the reader will be able to:

• Understand the importance of interoperability in IoT

• List various interoperability types

• Identify the salient features and application scope of each interoperability type

• Understand the challenges associated with interoperability in IoT

• Comprehend the importance of real-world use of interoperability frameworks in IoT

Introduction

The introduction of billions of connected devices under the IoT environment, which may extend to trillions soon, has contributed massively to the evolution of interoperability. As more and more manufacturers and developers are venturing into IoT, the need for uniform and standard solutions is felt now more than ever before [1]. Figure 9.1 shows the various facets of interoperability in IoT. Interoperability is considered as the interface between systems or products—hardware, software, or middleware—designed in such a manner that the connecting devices can communicate, exchange data, or services with one another seamlessly irrespective of the make, model, manufacturer, and platform.

The urgency in the requirement for interoperability and interoperable solutions in IoT arose mainly due to the following reasons:

• (i) Large-scale Cooperation: There is a need for cooperation and coordination among the huge number of IoT devices, systems, standards, and platforms; this is a long-standing problem. Proprietary solutions are seldom reusable and economical in the long run, which is yet another reason for the demand for interoperability.

• (ii) Global Heterogeneity: The network of devices within and outside the purview of gateways and their subnets are quite large considering the spread of IoT and the applications it is being adapted to daily. Device heterogeneity spans the globe when connected through the Internet. A common syntax, platform, or standard is required for unifying these heterogeneous devices.

• (iii) Unknown IoT Device Configuration: Device heterogeneity is often accompanied by further heterogeneity in device configurations. Especially considering the global-scale network of devices, the vast combinations of device configurations such as data rate, frequencies, protocols, language, syntax, and others, which are often unknown beforehand, further raise the requirement of interoperable solutions.

• (iv) Semantic Conflicts: The variations in processing logic and the way data is handled by the numerous sensors and devices making up a typical IoT implementation, makes it impossible for rapid and robust deployment. Additionally, the variations in the end applications and their supported platform configurations further add to the challenges.

Overview and Goals

Internet of Things (IoT) is rapidly gaining a foothold in the technology sector; it has managed to emerge as a highly sought-after field of study and research in computing sciences and electronics. The vastly interdisciplinary nature of the areas to which IoT can be applied has managed to pique the interest of the whole world. IoT finds diverse use in domains spanning industrial, military, as well as regular consumer applications. The versatility of IoT and its ability to connect anything make it one of the most demanded technologies of the modern age. The involvement of people from vastly diverse and distinct backgrounds, all point to the need for a concise repository of information on this new technology.

The Internet hosts much information on IoT, which is in the form of theory, tutorials, courses, implementations, and others. However, these discussions are so scattered that even professionals in this field have trouble obtaining integrated and concise information on IoT.

IoT is a new paradigm for connecting “things” in order to automate a system. In the context of IoT, “things” include computers, cell phones, medical devices, vehicles, wearables, and other appliances and devices for daily use. These “things” tend to be heterogeneous, which results in the development and existence of a vast number of communication solutions and protocols, which vary distinctly from each other. Consequently, communications among these “things” is a challenging issue in IoT. Another major challenge in IoT is the dynamic nature of “mobile things,” which generally follow a decentralized architecture. Due to this decentralized communication and control structure, the connectivity and data transmission dynamically changes with time, in turn resulting in a new set of challenges.

Pedagogical Aids

We have included various pedagogical aids to help the reader swiftly grasp the contents and the treatment of the various topics covered in this book. We have provided a set of conceptual questions at the end of this book. For solving these questions, the reader must have completely grasped the concepts covered in this book.

Additionally, we provide visual presentations of the chapters covered in this book so that it can be used as a teaching aid in colleges and universities. Each chapter of this book has the following pedagogical components:

• Learning outcomes, which gives an initial glimpse into the chapter.

Learning Outcomes

After reading this chapter, the reader will be able to:

• List common communication protocols in IoT

• Identify the salient features and application scope of each communication protocol

• Understand the terminologies and technologies in IoT communication

• Determine the requirements associated with each of these communication protocols in real-world solutions

• Determine the most appropriate communication protocol for their IoT implementation

Introduction

Having covered the various connectivity technologies for IoT in the previous chapter, this chapter specifically focuses on the various intangible technologies that enable communication between the IoT devices, networks, and remote infrastructures. We organize the various IoT communication protocols according to their usage into six groups: 1) Infrastructure protocols, 2) discovery protocols, 3) data protocols, 4) identification protocols, 5) device management protocols, and 6) semantic protocols. These protocols are designed to enable one or more of the functionalities and features associated with various IoT networks and implementations such as routing, data management, event handling, identification, remote management, and interoperability. Figure 8.1 outlines the distribution of these IoT communication protocol groups [3].

Before delving into the various IoT communication protocols, we outline some of the essential terms associated with IoT networks that are indirectly responsible for the development of these communication protocols.

Constrained nodes

Constrained nodes is a term associated with those nodes where regular features of Internet-communicating devices are generally not available. These drawbacks are often attributed to the constraints of costs, size restrictions, weight restrictions, and available power for the functioning of these nodes. The resulting restrictions of memory and processing power often limit the usage of these nodes. For example, most of these nodes have a severely limited layer 2 capability and often lack full connectivity features and broadcasting capabilities. While architecting their use in networks and networked applications, these nodes require special design considerations. The issues of energy optimization and bandwidth utilization are dominant work areas associated with these nodes [1].

Constrained networks

Constrained networks [2], [1] are those networks in which some or all of the constituent nodes are limited in usage aspects due to the following constraints:

• limited processing power resulting in restrictions on achieving smaller duty cycles.

Learning Outcomes

After reading this chapter, the reader will be able to:

• Explain the chronology for the evolution of Internet of Things (IoT)

• Relate new concepts with concepts learned earlier to make a smooth transition to IoT

• List the reasons for a prevailing universal networked paradigm, which is IoT

• Compare and correlate IoT with its precursors such as WSN, M2M, and CPS

• List the various enablers of IoT

• Understand IoT networking components and various networking topologies

• Recognize the unique features of IoT which set it apart from other similar paradigms

Introduction

The modern-day advent of network-connected devices has given rise to the popular paradigm of the Internet of Things (IoT). Each second, the present-day Internet allows massively heterogeneous traffic through it. This network traffic consists of images, videos, music, speech, text, numbers, binary codes, machine status, banking messages, data from sensors and actuators, healthcare data, data from vehicles, home automation system status and control messages, military communications, and many more. This huge variety of data is generated from a massive number of connected devices, which may be directly connected to the Internet or connected through gateway devices. According to statistics fromthe Information Handling Services [7], the total number of connected devices globally is estimated to be around 25 billion. This figure is projected to triple within a short span of 5 years by the year 2025. Figure 4.1 shows the global trend and projection for connected devices worldwide.

The traffic flowing through the Internet can be attributed to legacy systems as well as modern-day systems. The miniaturization of electronics and the cheap affordability of technology is resulting in a surge of connected devices, which in turn is leading to an explosion of traffic flowing through the Internet.

One of the best examples of this explosion is the evolution of smartphones. In the late 1990’s, cellular technology was still expensive and which could be afforded only by a select few. Moreover, these particular devices had only the basic features of voice calling, text messaging, and sharing of low-quality multimedia.

Learning Outcomes

After reading this chapter, the reader will be able to:

• Interpret the foundation of IoT

• List the basic terminologies and technologies associated with wireless sensor networks (WSN)

• List the basic terminologies and technologies associated with machine-to-machine communications (M2M)

• List the basic terminologies and technologies associated with cyber-physical systems (CPS)

• Differentiate between WSN, M2M, and CPS

• Relate new concepts with concepts learned before to make a smooth transition to IoT

Introduction

Before delving into the details of the Internet of Things (IoT), a discussion on the base technologies, which make up the crux of IoT, is required. A majority of these technologies, before the IoT era, were used separately for sensing, decision making, and automation tasks. The range of application domains of these technologies extended from regular domains like healthcare, agriculture, home monitoring, and others to specialized domains such as military and mining. Some of these precursor technologies still being used and often re-engineered for IoT are wireless sensor networks (WSN), machine-to-machine (M2M) communications, and cyber physical systems (CPS). All of these precursor paradigms have their distinct signatures and application scopes. A basic overview of these precursor technologies is covered in the subsequent sections in this chapter.

Wireless Sensor Networks

Wireless sensor networks (WSN), as the name suggests, is a networking paradigm that makes use of spatially distributed sensors for gathering information concerning the immediate environment of the sensors and collecting the information centrally. Here, the sensors are not standalone devices but a combination of sensors, processors, and radio units—referred to as sensor nodes—sensing the environment and communicating the sensed data wirelessly to a remote location, which may or may not be connected to a backbone network. Figure 3.1 shows the block diagram of the various standard components of a typical WSN node [4]. The exact specifications of each of these blocks vary depending on the implementation requirements and the network architect's choice.

Figure 3.2 shows a typical WSN implementation, where the master node aggregates data from multiple slave nodes, forwards it to a remote server utilizing access to the Internet through cellular connectivity. The stored data on the server can be visualized by a user or a subscriber to the system from anywhere in the world over the Internet. WSNs mainly follow a system of communication known as master–slave architecture.

Learning Outcomes

After reading this chapter, the reader will be able to:

• List the salient features of transducers

• Differentiate between sensors and actuators

• Characterize sensors and distinguish between types of sensors

• List the multi-faceted considerations associated with sensing

• Characterize actuators and distinguish between types of actuators

• List the multi-faceted considerations associated with actuation

Introduction

A major chunk of IoT applications involves sensing in one form or the other. Almost all the applications in IoT—be it a consumer IoT, an industrial IoT, or just plain hobby-based deployments of IoT solutions—sensing forms the first step. Incidentally, actuation forms the final step in the whole operation of IoT application deployment in a majority of scenarios. The basic science of sensing and actuation is based on the process of transduction. Transduction is the process of energy conversion from one form to another. A transducer is a physical means of enabling transduction. Transducers take energy in any form (for which it is designed)—electrical, mechanical, chemical, light, sound, and others—and convert it into another, which may be electrical, mechanical, chemical, light, sound, and others. Sensors and actuators are deemed as transducers. For example, in a public announcement (PA) system, a microphone (input device) converts sound waves into electrical signals, which is amplified by an amplifier system (a process). Finally, a loudspeaker (output device) outputs this into audible sounds by converting the amplified electrical signals back into sound waves. Table 5.1 outlines the basic terminological differences between transducers, sensors, and actuators.

Sensors

Sensors are devices that can measure, or quantify, or respond to the ambient changes in their environment or within the intended zone of their deployment. They generate responses to external stimuli or physical phenomenon through characterization of the input functions (which are these external stimuli) and their conversion into typically electrical signals. For example, heat is converted to electrical signals in a temperature sensor, or atmospheric pressure is converted to electrical signals in a barometer. A sensor is only sensitive to the measured property (e.g., a temperature sensor only senses the ambient temperature of a room). It is insensitive to any other property besides what it is designed to detect (e.g., a temperature sensor does not bother about light or pressure while sensing the temperature).

Learning Outcomes

After reading this chapter, the reader will be able to:

• Understand the concept of cloud computing and its features

• Understand virtualization, different cloud models, and service-level agreements (SLAs)

• Identify the salient features of various cloud computing models

• Understand the concept of sensor-clouds

Introduction

Sensor nodes are the key components of Internet of Things (IoT). These nodes are resource-constrained in terms of storage, processing, and energy. Moreover, in IoT, the devices are connected and communicate with one another by sharing the sensed and processed data. Handling the enormous data generated by this large number of heterogeneous devices is a non-trivial task. Consequently, cloud computing becomes an essential building block of the IoT architecture. This chapter aims at providing an extensive overview of cloud computing. Additionally, Check yourself will help the learner to learn different concepts are related to cloud computing.

Cloud computing is more than traditional network computing. Unlike network computing, cloud computing comprises a pool of multiple resources such as servers, storage, and network from single/multiple organizations. These resources are allocated to the end users as per requirement, on a payment basis. In cloud computing architecture, an end user can request for customized resources such as storage space, RAM, operating systems, and other software to a cloud service provider (CSP) as shown in Figure 10.1. For example, a user can request for a Linux operating system for running an application from a CSP; another end user can request for Windows 10 operating system from the same CSP for executing some application. The cloud services are accessible from anywhere and at any time by an authorized user through Internet connectivity.

Cloud computing comprises a shared pool of computing resources, which are accessible dynamically, ubiquitously, and on-demand basis by the users. This shared pool of resources includes networks, storage, processor, and servers. These resources are accessible by multiple users through a regular command-line terminal at the same or different time instants. The services of cloud computing are based on the pay-per-use model. The concept is the same as paying utility bills based on consumption. In cloud computing, a user pays for the cloud services as per the duration of their resource usage.

Learning Outcomes

After reading this chapter, the reader will be able to:

• Understand the concept of Fog computing and its features

• List the salient features of a Fog computing architecture

• Identify the requirements of a Fog computing architecture

• Understand the importance of Fog computing through real-life use cases and identify its application scope in IoT

Introduction

In the Internet of Things (IoT), billions of Things connect through wired or wireless connectivity. These Things or devices produce a huge volume of data, which are typically transmitted to the cloud for analysis. Sometimes, the process of transmitting the data to the cloud and analyzing the data may consume a significant amount of time, which is undesirable for time-critical applications, such as healthcare. Therefore, providing real-time services is a major challenge in cloud computing. The concept of Fog computing was introduced considering the high latencies involved in cloud computing. This chapter aims to provide an insight into the fog computing architecture and how it reduces the latency of data processing and transmission.

In the IoT architecture, physical devices transmit data to the cloud. After specific processing of these data, service is provided to the end user applications. Fog computing follows a distributed architecture, which enables processes to execute near the edge of the devices to avoid service latency. A fog layer is an intermediate layer between the physical IoT devices and the cloud. The term, fog computing, was coined by Cisco.

In Figure 11.1, we depict the difference between the basic architecture of cloud and fog computing. In traditional cloud architecture, all data from the devices are transmitted to the cloud directly and then, processed in order to receive the final end user services. These end user services may be in the form of results of analytics, visualizations, and other such processed information. In fog computing architecture, time-sensitive data from different devices are transmitted to the fog devices at the fog layer. Further, data are processed to serve an end user application. Moreover, based on the requirements, the data are transmitted to the cloud layer for long-term storage.

Essential characteristics in fog computing

A fog computing platform resides between cloud and physical IoT devices; it provides processing, storage, and networking services.

Learning Outcomes

After reading this chapter, the reader will be able to:

• Understand the common hardware platforms, sensors, and actuators used in IoT

• Assess the importance of each sensor or hardware in various applications

• Understand the code structure required to operate these hardware and sensors /actuators connected to them

• Relate the IoT hardware and sensors according to the requirements of their applications

Introduction to Arduino Boards

Arduino is an open-source platform which comprises both hardware and software modules. The hardware are typically processor boards with analog and digital pins for various functions. Arduino IDE (integrated development environment) is the software that is used to communicate to the hardware and program them to perform specific functions. We cover the process right from the installation of Arduino IDE to loading codes to the processor boards.

Arduino vs. Raspberry Pi: Choosing a board

An Arduino board has a mounted micro controller chip to perform simple tasks. The board can execute single tasks at a time, in a repetitive manner. The main focus area of Arduino is integration of hardware to the board rather than complex computation or multitasking. For example, the Arduino board can sense the ambient temperature of an area and display it on an output console. Raspberry Pi is a more complex system, much like a minicomputer. It has an advanced processor, uses Linux-based operating systems, and is able to perform multiple tasks with complex computation. Raspberry Pi gives the option forWi-Fi connectivity and access point formation, something which would require external components to be connected on to a regular Arduino board. The choice of selecting Raspberry Pi or Arduino depends solely on one's application. If users want to perform rudimentary, basic level tasks then Arduino is the best option; it avoids any kind of complexities. If the intended application needs multiple tasks to be done with complex computations, Raspberry Pi is a much more suitable choice as it has pretty much everything integrated on-board. Once the users decide which board to choose, they can narrow down the options based on their requirement and the available features on the board such as operating voltage, RAM size, clock speed, storage, and additional wireless connectivity modules.

Arduino installation and setup

Installer files are available for different variants of the operating system on the official website of Arduino. Download the installer (Figure 16.1) and execute the file.

Learning Outcomes

After reading this chapter, the reader will be able to:

• Describe the common analytical tools and machine learning algorithms used with IoT data

• Assess the importance and applicability of each algorithm

• Understand the operating principle of each of these analytical methods

• Assess the performance of various analytical and learning algorithms and methods through the use of various performance metrics

• Relate to the uses of various learning algorithms through examples

Introduction

In previous chapters, we learned that sensors are an intrinsic part of IoT. These sensors collect data from the environment and serve different IoT-based applications. The raw data from a sensor require processing to draw inferences. However, an IoTbased system generates data with complex structures; therefore, conventional data processing on these data is not sufficient. Sophisticated data analytics are necessary to identify hidden patterns. In this chapter, we discuss a few traditional data analytics tools that are popular in the context of IoT applications. These tools include k-means, decision tree (DT), random forest (RF), k-nearest neighbor (KNN), and density-based spatial clustering of applications with noise (DBSCAN) algorithms. Before discussing these algorithms, let us understand some of the basics related to machine learning (ML).

Machine learning

The term “machine learning” was coined by Arthur Lee Samuel, in 1959. He defined machine learning as a “field of study that gives computers the ability to learn without being explicitly programmed”.

ML is a powerful tool that allows a computer to learn from past experiences and its mistakes and improve itself without user intervention. Typically, researchers envision IoT-based systems to be autonomous and self-adaptive, which enhances services and user experience. To this end, different ML models play a crucial role in designing intelligent systems in IoT by leveraging the massive amount of generated data and increasing the accuracy in their operations. The main components of ML are statistics, mathematics, and computer science for drawing inferences, constructing ML models, and implementation, respectively.