1 Introduction to Internet

CHAPTER of Things
LEARNING OBJECTIVES
The chapter will help you acquire an in-depth understanding of
" Definition of IoT.
"History and growth of Internet and IoT.
." Application areas and focus of IoT.
" Characteristics of IoT.
. Things in IoT (What makes it IoT?).
"IoT stack.
"Enabling technologies.
"Levels of IoT.

1.1 Introduction and Definition of Internet of Things

one of the popu
The term Internet of Things (IoT) has emerged over the past few years asdiscussions
terms. Today, IoT figures prominently in technology due to
lar 'technology buzz' awareness about IoT, its architectural
its rapid growth. This book aims at imparting complete and selection, cloud com
aspects, protocols involved, software contribution, hardware usage
including examples from smart
puting and data analytics. The book concludes with case studies
going to introduce loT through a
city, smart water, and smart healthcare. Broadly, the book is
practitioner's approach.
crux of all the definitions remains the
There are multiple ways one could define IoT. But the
same.

devices embedded
IoT refers to the interconnection via the Internet of computing
receive data.
ineveryday objects, enabling them to send and
NOTE
2 Introduction to Internet of Things

everything and everywhere. Hence, loT can also be defined as the analysis of data
Today, data is subsequently after the interchange of data. One can
to generate a meaningful action, triggered
why loT is the buzzword.
refer to Fig. 1.1to get a preliminary idea about

INTERNETof
THG
Figure 1.1 Internet of (every) thing.
The scope of IoT is not just limited to getting the devices connected or networked, but rather it
is more about the exchange of meaningful information from one device to another to acquire
purposeful result.
Data is everything. However, to draw accurate results, raw data has to be interpreted accurately
which requires other technological parameters to collaborate. If someone just gets the data from
the sensors without understanding how to use or where to use, it is futile. Hence, interpretation
holds value.

It is important to note that loT is not asingle technology, but rather acombination of technolo
gies and domain knowledge. Hence, amechanical engineer has to work with an electricalengi
neer and a computer science engineer for building a complete loT product.

IoT is not owned by any one engineering branch. IoT is a reality when multiple
domains come together.
REMEMBER
 1.1 Introduction and Definition of Internet of Things 3

Let us see how Internet has accelerated the pace of technological developmentin recent history.
Aquick summary 1S presented below:

1990 World Wide Web was born. 3,00,000 desktop computers were connected to the Internet by
then. The count is very insignificant compared to the numbers now.
2000 300+ million desktop computers (laptops too) connected to the Internet. The growth was
phenomenal and faster than expected.
2016 Majority of the people own smart phones: thev are connected to the Internet 24 x 7
about 2.5 billion mobile phones connected to the Internet.
2020 Refrigerators, kettles, water heaters, cameras, lights, home equipment and other elec
tronic devices will be connected to the Internet and there would be 15 billion device
connections. About 4 billion navigation systems would be embedded within the Inter
net. Connected cars have already become a reality. About 500 million wearables,
watches, shoes, etc. and 100s of millions of implants in human/animals would talk to
Internet.

2016
2000s

1990s 2020

Figure 1.2 Growth of Internet - from nowhere to "now here".

Figure 1.2 indicates the growth of Internet in brief.Considering the advancement, soon one's life
Would be governed entirely by Internet and IoT in the near future. The following scenario is an
Idea how you would lead your life when everything in it is connected to the Internet:
1. When you wake up, the water heater would have hot water ready.
2. During your bath,the microwave would have cooked your food.
3. When you lock the door of your house, the car doors would open for you.
4. Your AC would adjust based on your body temperature/room temperature.
5. Parking slots would be informed while your car is nearing your office.
6. The pillbox would you at the time of your tablet intake.
 4
Introduction to Internet of Things

7. While nearing your house, the house lights would turn on with appropriate brightness
8. Your workout room lights/ambience would be automatically set. Heart beat/km al..
would be updated.
9. Your refrigerator would order milk/egg on need.
10. The bedroom ACwould automatically switch on for you to sleep.

This is just aglimpse of what allcan change. Nothing is impossible with the Internet and JoT i
place. The onlychallenge is that we need to be up to date with the requirements and know how
to use it.
IoT enables objects (which eventually are "the things") to be sensed while controlling them
remotely, which facilitates better interaction between the physical world and computers. This
improves the efficiency and accuracy with minimal human intervention.
Figure 1.3 shows IoT design goals the designer should consider while building an application, It
should fulfill all the requirements shown in the figure.

Any time - Any device Anyone Any service/Any business

IeT
Any Network (Non
Anywhere/Any location
homogenous)

Figure 1.3 loT design goals.

1.2 loT Growth - AStatistical View

From the "CISCO BSG 2011" presentation (Fig. 1.4), one can see that IoT has a prodigious
growth. CISCO BSG has predicted that by 2020, people would be using 50 billion smart
objects. Also, a surprising fact is that each person would have six objects connected to the
Internet. The projection for 2020 indicates a steep growth in the use of smart objects in com
parison to 2003.
Not only CISCO, butmany other major forums and organizations realized the potential of Iol
development and have made or planned huge investments. Also, development of skill compe
tence in IoT-related areas is being undertaken by most organizations as they see huge business
prospects in these areas.
 1.3 Application Areas of loT 5

50
~6 things online per person 50
40 Billion

Rapid adoption rate of digital infrastructure smart objects

5 x faster than
Billions
devices
of 30 electricity and telephony

20

Inflection point
10
6.307 World population
6.721 6.894 7.347 7.83
0
2003 2008 2010 2015 2020

Figure 1.4 CISCO BSG - loT growth prediction.

1.3 Application Areas of loT

Chapter 7iscompletely devoted to applications of IoT. As a prelim, Fig. 1.5 showcases the scope
and application areas of IoT. It vividly demonstrates the huge opportunities of IoT. IoT can
be used to build applications for agriculture, assets tracking, energy sector, safety and security
sector, defence, embedded applications, education, waste management, healthcare products,
telemedicine, smart city applications, etc.

Vehicle, asset, pet and Agriculture and Energy sector

Safety/Vigilance/
human tracking resources, including Security systems
water

loT (Internet of Things)

Embedded applications/ Healthcare applications Telemedicine Smart city applications

applications
Mobile applications

Figure 1.5 loT application sectors.

6 Introduction to Internet of Things

1.4 Characteristics of loT

The characteristics of IoT are as follows:

1, Connectivity: This is an important aspect/requirement of IoT infrastructure. Things

in IoT should be connected to the IoT infrastructure. Anyone, anywhere, anytime
connectivity should be guaranteed at all times. Without connection, nothing makes
sense.

2. Intelligence and identity: The extraction of knowledge (i.e., what is to be inferred) from
the generated data is very important. For example, sensors generate data, but that data
will only be useful if it is interpreted properly. Each IoT device has a unique identity
(e.g., an IP address). This identification is helpful in tracking the equipment and at times
for querying its status.
3. Scalability: The number of elements (devices) connected to IoT zone is increas
ing day by day. Hence, an IoT setup should be capable of handling the massive
expansion. The data generated as outcome is enormous and it should be handled
appropriately.
4. Dynamic and self-adapting (complexity): IoT devices should dynamically adapt them
selves to the changing contexts or scenarios. Consider a camera meant for surveillance. It
should be adaptable to work in different conditions and different light situations (morn
ing, afternoon, or night).
5. Architecture: IoT architecture cannot be homogeneous in nature. It should be hybrid,
supporting different manufacturer's products to function in the IoT network.
6. Safety: There is the danger of sensitive personal details ofa user getting compromised
when all his/her devices are connected to the Internet. This could cause a loss to the user.
Hence, data security is a major challenge. Besides, the equipment involved in the huge
IoT network may also be at risk. Therefore, equipment safety is also critical.
The next section discusses the "things" in Internet of Things.

1.5 Things inloT

In the context of loT, things refer to avariety of devices. At times, even humans in the loop become
athing, For something to qualify as a"thing" it requires identity of its existence. The "thing" in
anetwork can monitormeasure; for example, a temperature sensor could be a thing. Things are
capable of exchanging data with other connected devices in the system. The data could be stored in
acentralized server (or cloud), processed there anda control action could be initiated. The devices
involved in getting this accomplished are known as things (Fig. 1.6).
Some of the famous "things" that you would already be aware of are temperature sensors, pres
sure sensors, humidity sensors, etc. The data from these sensors is collected (real time) and sent
to the cloud (rather than a local server). Based on the data analysis, the control action wold
be taken. For example, switching off the water heater remotely when the water is heated asper
requirement. Not just sensors, the following can also be called as things:
 1.5 Things in loT 7

Figure 1.6 "Things" in Internet of things.

1, Industrial motors.
2 Wearables (e.g., watch).
3. Vehicles.
4. Shoes.
5. Heart monitoring implants (e.g., pacemaker, ECG real-time tracking).
6. Biochip transponders (for animals in farms).
7. Automobiles with built-in sensors (automobile feature real-time monitoring).
8. Food/perishables quality measuring.

«THINGS"= HARDWARE + SOFTWARE + SERVICE

REMEMBER

Now, let us consider the example of IoT-based home automation. In this case, the things" could
be the following (Fig. 1.7):
 Introduction to Intermet of Things

Figure 1.7 loT-based home automation.

1 Lighting control and automation devices.

2. Ventilation devices.
3. Air conditioning [heating, ventilation and air conditioning (HVAC)] systems.
4. Appliances such as washer/dryer.
5. Air purifiers.
6. Ovens or refrigerators/freezers that use Wi-Fi for remote monitoring.
7. Security cameras.
8. Smart phones.
This list is certainly not exhaustive.

List the things that could be in smart retail environment.

FLASH
QUIZ

1.6 loT Stack

This section
As with any other digital technology, loT has a stack with clearly defined layers.
explains loT stack in detail.
layer. OSI
Before discussing IoTstack, let us first recall the Open System Interconnection (OSI) Layer 1
model defines a networking framework to implement protocols in seven layers, where
 1.6 loT Stack 9

the physical layer and layer 7 is the application layer. Each layer has the protocols defined
clearly with appropriate hardware/software workino from it. We have identified the following
seven layers in loT stack.

1. Layer 1(Physical or sensor layer): This layer is concerned about the physical com
ponents, which mainly includes sensors. In this layer, the sensors are the core com
ponent. lemperature sensor, pressure sensor. humidity sensor,etc. can all be referred
to as phySical layer components (Fig. 1.8), While considering industrial automation,
PLC, actuator, etc. are regarded as physical layer components.This layer is responsible
for data collection (i.e., sensing happens here). Choosing an appropriate sensor
is the challenge in this layer since there are many sensors available in the market that
capable of performing the same tasks but at different costs. Hence, the selection of
sensors is important.

Sensors
(Temperature, Hurnidity, Accelerator, Presssure etc.)
Industrial Automation
(PLC/Actuator/Pneumatic Control)

Figure 1.8 Layer 1: Sensor layer (physical layer);data collection happens here.
2. Layer 2 (Processing and control action layer): This layer is very significant; it com
prises of the core components for IoT. The micrOcontrollers or processors are found
in this layer (Fig. 1.9); the data is received by the microcontrollers from the sensors.
A variety of development kits are available in the market; one can easily spot Arduino,
NodeMCU, PIC, ARM development boards. Operating systems play a major role too
and Android, IOS, Linux can very well execute the task. The data collected from the
sensors is processed in this layer. To determine if the data is meaningful a microcontroller
should be present.
10 Introduction to Internet of Things

Microcontrollers:
PIC/ARM/INTEL boards

DEV Kits:
Arduino/Rasberry
OS:
RTOS, Linux, Android, loS

Figure 1.9 Layer 2: Processing and control action layer; microcontrollers/

processors work in this layer.
3. Layer 3 (Hardware interface layer): The next layer in the stack is the hardware interface
layer. The hardware components and communication standards such as RS232, CAN,
SPI, SCI, IC, etc. Occupy this layer. Allthese components ensure flawless communica
tion (Fig. 1.10).

Components for Communication:

Seria/Parallel standards, RS 232, USB,
1PC, SPI, CAN, routers, etc.

Figure 1.10 Layer 3: Hardware interface layer; handshake happens here.

4. Layer 4(RF layer): Whenever one talks about loT, RF inevitably gets discussed.
It plays a major role in the communication channel - be it short range or long rang
 1.6 loT Stack 11

The protocols used for communication and transport of data based on

this layer. Some famous and common protocols are Wi-Fi, RF are listed in
This layer (Fig. 1.11)can also include Li-F (not dependent NFC, RFID, Bluetooth, etc.
on RF); which are effective
alternates for RF protocols.

Classification:
Short Range (Low Bandwidth, High Bandwidth)
Long Range (Low Bandwidth, High Bandwidth)
Examples:
NFC, RFID, Wi-Fi, Bluetooth, BLE, Li-Fi, LTE

SWASMINARAYA

BLE
N))
NFC
(LIBRARY)
Figure 1.11 Layer 4: RF layer.
5. Layer 5 (Session/Message layer): Session management is as important in IoT as it is in
general networking, which is guided by OSI layer. There are protocols which oversee
how messages (data) are broadcasted to the cloud. Layer 5 (session layer) has the mes
saging protocols as MQTT, COAP, etc. and also other protocols such as SSH and FTP
(Fig. 1.12). The protocols for messaging with software coding have been demonstrated in
Chapter 3.
Session/Message layer
(Messaging is the content)

Some Protocols:
MQTT,COAP, HTTP,FTP, SFTP,SSH

Figure 1.12 Layer 5: Session/message layer;, messaging is the content.

12 Introduction to Internet of Things

6. Layer 6 (User experience layer): This layer is fully concerned with the end user expe-
rience (Fig. 1.13). When a product is designed, it should showcase rich UI features
designs which provide a pleasing experience while using the service/system or nr
uct. Object-oriented programming languages, scripting languages, analytics tools, etc. all
should be included in this layer.

Technologies:
Object Oriented, Procedure Oriented

DBMS, SQL
Analytics Tools/Software from vendors

Figure 1.13 Layer 6: User experience and visualization layer.

7. Layer 7 (Application layer): Everything comes to perfection at this layer
(Fig. 1.14). This layer talks about the possible applications that can be built with the
support of the rest of the layers. It can range from a simple automation application
to smart city application.

Smart Home, Smart City,

Smart Parking, Smart Energy,
Smart Retail, Smart Agriculture

Figure 1.14 Layer 7: Application layer.

After learning about the layers, it is now easier for readers us to relate them with andestination
app
tion, for example, vegetable quality monitoring during transport from source to the
 1.6 loT Stack 13

Temperature Humidity Odour

Analog-to-Digital IFTTT
Converter USON
NodeMCU Adafruit cloud Google sheets JSON

Connected to Wi-Fi

Location displayed in Android app for displaying

Google maps grades and locations

Connected to Internet

Figure 1.15 Architecture of a smart vegetable quality tracking system using loT.
using IoT. This scenario is presented as a case study in Chapter 2In this application, you will be
able to identify all the above-mentioned layers. To set the tempo, the architecture diagram of
the smart vegetable quality tracking application for transport is presented in Fig. 1.15. One can
understand that it starts with the sensor layer, where temperature, humidity, etc. are measured.
Then the data goes tothe microcontroller; from there it goes to the cloud (Adafruit) and here
the messaging protocols take effect. Adafruit is the cloud storage platform used in this appli
cation. The procedure for setting up the same is covered in detail in Chapter 5. RF protocols
such as Wi-Fi are used for transport and communication. User experience is taken care with the
Android application designed. The complete architecture, functioning and results are presented
in Chapter 7.
The consolidated view of all the seven layers in the IoT stack is presented in Fig. 1.16.
Application layer
User experience

Session layer

RF layer

Hardware
interface layer
Processing and
control layer

Sensor layer

Figure 1.16 loT layers.

14 Introduction to Internet of Things
1.7 Enabling Technologies
As already discussed, IoT is a collection/group of many
technologies/devices,
sensors, embedded systems (1.e.,the boards), data analytics, mobile and mobile The
simplest of
aspects and protocols involving cloud storage (computing) have all
gies. There are many more things which can be listed. In
general, become enabling
enabling Internet,Steecchunolrityo-
fall under one of the following categories:
1. Technologies that help in acquiring/sensing data.
2. Technologies that help in analysing/processing data.
technologies/devices
3. Technologies that help in taking control action.
4. Technologies that help in enhancing
security/privacy.
In the following subsections, we will
discuss some of the enabling
technologies.
1.7.1 Sensors
Sensors are at the heart of
any IoT application. As the name suggests,
ronment and retrieve data. Sensors are
ple of a sensor is a simple the starting point of any IoT they senseAntheexam:envi:
temperature monitoring application such as aapplication.
temperature sensor); it fetches data for us
Details about sensors and their to operate on. thermometer (.e,
Sensors could be
interfacing with computing units have been analog or digital.
Chapter 2.
discussed in
Some examples of sensors that
(Fig. 1.17): could be regarded as enabling
technologies
are as foloNS
1.
Camera used in home security
2
3.
Weather tracking system uses systems.
4.
Vehicle health monitoring sensors
On Board temperatuELM327
Diagnostics (OBDs) such as
mation from an automobile to
re/humidity/musedoisture
keep track of speed, tyre sensors.
are pressure, etc.
detect for collecting all critical infor
5.
6.
Vibration sensors are used to track theanomalies.
Water quality quality of
7.
level, etc.
PIR sensor is used in
is
monitored through sensors buildimeasure
that ngs/structuPH,res.turbidity, chloride
pedestrian signal operation with human presence detection.
1.7.2 Cloud Computing
The next
covered intechnology that is
detail in Chapter highly significant init IoT
5. In this is cloud Cloud computing 5
storage plays a major role in loT. As a chapter, is discussed ascomputing.
than expected because per its relevance to loT
Data
it serves as an data storage option, cloud has
popula
much more storage
affordable, effective and efficientgrown
medium for data
 1.7 Enabling Technologies 15

Vibration sensor Temperature Moisture sensor Camera

sensor

OBD - ELM327 PlR sensor PH sensor

BNE280

Pressure, Temp
Humidity- BME280

Figure 1.17 Sensors that act as enabling technologies.

All the case studies which are cited in Chapter 7 use Adafruit as it was very
much affordable and
easier to learn and handle. Cloud services are categorized as follows:

machines
1. laaS (Infrastructure-as-a-Service): In this cloud service, one can choose virtual
provides vir
over physical machines. In other words, it is a form of cloud computing that
machines, select
tualized computing resources over the Internet. The users manage the
the OS and underlying applications, and pay per their use.
2 PaaS (Platform-as-a-Service): This is a cloud computing model in which the cloud ser
software tools needed for
vice provider (a third-party provider) delivers hardware andprovider hosts the hardware
application development to users over the Internet. A PaaS and maintain the
and software on its own infrastructure. Users have to build, manage
applications as per their requirement.
software application is
3. SaaS (Software-as-a-Service): In this model, a completea service. This service can
provided to the user. It can also be called application as
be availed by paying a monthly, yearly, etc., subscription.
services, Azure and
Some well-known service providers in the market are Amazon web
Adafruit.
16 Introduction to Internet of Things

1.7.3 Big Data Analytics

function or operation we get more data.
Data is everywhere, and from every IoT is all
collecting data from various sensory nodes and
the application a success.
handling the huge data is
fundamental to about
make
The biggest challenge with big data is its volume, variety, speed (velocity) at which it
its veracity. These are fondly referred as the 4Vs of big
data. Chapter 6on data
details of big dataanalytics. Big data is majorly governed by
the following: analyticscomeS and
provides
1. Scale (Volume): Huge volume of data is generated every minute. Storage bas ho
reduced. Cloud storage ne
inexpensive and hence, cost-related challenges have
ware storage both have become affordable because OI the tremendous growth in
semiconductor industry.
2. Complexity (Variety): Data no longer comes trom one single source. Moreover. it comas
syStemt
in different formats (e.g., audio, video, text and image) and has to be interpreted
atically.This becomes a huge challenge.
rate at which at
3. Speed (Velocity): It is the rate at which new data is being created. The
is generated is very fast. Also, data dynamics changes very frequently. Nowadays, data
ata
can come from anywhere from fitbit watches to refrigerators. All the data pours in
the actual data
very high speed, which makes it very challenging to not miss and oversee
from the noise.
data we now rely
4. Data in doubt (Veracity): It is perhaps the one hidden secret of all the
dynamically and ambi
on. How accurate is all this data anyway?The data's nature alters
to proces
guity is often seen (incomplete data). Hence, it would be pretty challenging
this unstable data.

partial list to answer this is as follows:

So the question is: "Who is generating all this data?"A

1. Sensors from security systems.

2. Sensors from weather monitoring systems.
3. Sensors from car/navigation systems.
4. Sensors from water quality monitoring systems.
5. Datafrom wearables (e.g., bands).
6. Data from industrial equipment (e.g., motor health).
7.
Sensors from bridges/roads about traffic density and other factors.
8. Social media (e.g., tweets, photo uploads, etc.). the
oneof
IS
When it comes to IoT, it is data, which is everything. Hence, data analytics
all about
enabling technologies for building a complete and comprehensive IoT application.

1.7.4 Embedded Computing Boards

real-
to
boards
component to bring IoT design
An embedded computing board a very important computing
these are linked with the
ity, From the proof of concept to the prototype, all
 1.7 Enabling Technologies 17

Most of the computing boards available in the market are driven by microcontrollers or proces
sors. Some of the boards are as follows:

1. Raspberry Pi.
2. Arduino (many variants).
3. NodeMCU.
4, Intel Edison.
5. Intel UP Squared* Grove* IoT Development Kit.

All these boards are small, yet smart. Also, the cost involved is very minimal and one can get
these boards for less than a hundred dollars. Figure 1.18 shows some of these boards.

NodeMOCU
Arduino Uno Intel UP Squared loT Dev Kit
(one of the variants)

Raspberry Pi3

Figure 1.18 Computing boards.

1.7.5 Communication Protocols

Protocols are the pillars for good IoT infrastructure and hence are very important in com
munication. Data exchange happens through these protocols, which take care of the following:
1. Addressing.
2. Format of the messages.
3. Message security (encryption and decryption).
4. Routing.
5. Flow control.
6. Error monitoring.
18 Introduction to Internet of Things
7. Sequencing,
8. Retransmission guidelines.
9. Segmentation of the data packets.
Adetailed analysis of the protocols shall be presented in the Chapters 3and 4,
1.7.6 User Interfaces
All devices must have an intuitive user interface. IoT devices/services should be desiena.
such a way that accessing and handling the services are easier and comfortable for the
In most cases, the end user shall be provided with "mobile application or web end user.
application should be consistent and not clumsy. application? The
1.8 loT Challenges
As with any application, while building an IoT application one faces many challenges - both
technical and non-technical. The following are some of the challenges.

1. Security/Personnel safety: This is one of the most significant challenges to confront.

Since a number of devices are used in IoT, user data becomes more vulnerable to theft.
So, it becomes necessary to make sure that the data is safe. Even the best social media
websites get into trouble for misusing acustomer's data. Poor security features can let
attackers damage the whole network.
People's personal safety is also aconcern and challenge. The implants and wearables
used by people should be safe and the devices should not cause any physical damage to
the person using it. Since many devices are in the loop, if one device gets attacked, thet
the rest of the devices could also become vulnerable.
2. Privacy: One could be tracked/monitored by anyone, as we are connected 24 x7to the
Internet. At times, one could be tracked without the host's permissions in place. So,there
is a threat on user data and raises aquestion on user privacy: "How do we ensure that the
data that is sensed/collected from the user is with their permission?"
3. Data extraction with consistency from complex environments: It is a huge challenge to
senselextract data from complex environments. Let us consider the following questior
"How do we sense the data input (temperature, humidity, etc.) during commute ou
a vehicle?" Let us assume that a highly temperature-sensitive material is being ua
ported, where the measurement is always expected to be accurate. Variation in tem-
perature could also damage the products being transported, | particularlyif medicines
critical
drugs are being transported. In such a case,the temperature maintained is very
and should be accurately monitored. In this case, if the temperature is abouttochang,
then the corrective action has to be taken to ensure that the medicines/drugs are
damaged. Also, in loT-based applications, the Internet is needed and is mandalo
providing
most places. In hilly regions, which are technically not so up-to-date terrains, cloud
the
Internet 24 x 7 may not be an easy task. Hence, data extraction and storage inextracting
could be more challenging. Extracting data inside a room is different from
data fronm an open environment.
 1.9 loT Levels 19

4. Connectivity: This is a serious challenge that the loT world must acknowledge. Since
the Internet is itself agiant collection of networks and devices and IoT is a part of it,
requirement of wired and wireless connectivity is anecessity. The usage of frequency/
spectrum is also to be noted (2.4 GHz band is the optimal band everywhere). There are
spectrum regulations to be followed based on the country for which the application is
being developed. Hence, understanding the connectivity requirements is important.
5. Power requirements: All the loT devices require power and most of them are battery
operated. Even though we now have long-lasting batteries that are economical, demand
for power is on the rise. Hence, usage of green power sources such as solar and wind
should be motivated. If the power requirements are met appropriately, IoT can be even
more powerful.
6. Complexity involved: IoT is not easy. It needs a lot of different domains to integrate
into a cohesive system. There is very limited expertise available in the market, but the
growth is very rapid. The toolkits, software and hardware are not abundant and real skill
is required to build an application. Over the next few years, IoT would get more and
more technology experts to work.
7. Storage: Cloud is becoming mandatory for the data to be stored and analysed. The chal
lenge with respect to this aspect is connected to the following points:
a. Which cloud do we use (private, public, or hybrid)?
b. How do we identify the service provider?
c. How much does it cost?
d. Do we really need cloud?

More on cloud computing challenges and the selection of cloud service provider is discussed in
Chapter 5.

1.9 loT Levels

Based on the architectural approach, IoT can be classified in five levels: Level 1to Level 5. In this
section, we will explain all these levels with appropriate examples for better clarity.
1.9.1 Level 1
What is Level 1 IoT application all about? As one could guess, it is of minimal complexity and
the easiest to build.

The application has one sensor a device to sense. lt could be a temperature sensor, pressure
sensor, etc. The data sensed is stored locally and the data analysis is done locally. Monitoring/
control is done through an application (.apk or webapp). This is used for simple applications
involving limited or no complexity. Data generated in this level application is not huge (i.e., not a
big data). Al the control happens through the Internet. ASimple example scenario is presented
in Fig. 1.19, where the temperature sensor senses the room temperature and the data is stored
and analysedlocally. Basedon the analysis, the control action can be triggered through mobile
application or it can help in monitoring the status.
20 Introduction to Internet of Things

Temperature
Air conditioner sensor Data collection and
analysis
Control and monitoring
action (On/Off
Control)
Figure 1.19 Level 1-loT application.
1.9.2 Level 2
The second level is slightly more complex than the previous level. Here, the data isS
minous and hence, cloud storage is preferred. The frequency of sensing done by themore volu-
sensor is
faster. This means that the sensing happens faster and the number of times
sensing is done would
be much more than Level 1. The analysis is carried out locally, while cloud is meant for
storaea
only. Based on the data analysis, the control action can be triggered through the
or mobile application. Some examples are agriculture web application
based on odour, etc. Figure 1.20 shows the Level 2 IoTapplications,
application
room freshening solutions
of an air conditioner. The
sensor reads the room temperature at a better pace and rate than
to the cloud for storage. Analysis is done locally and the Level1; the data then goes on
application.
action is triggered through the mobile

Air conditioner
Temperature
sensor
Big data (Bigger than Cloud Control and monitoring
Level 1) action (On/Off
Data analysis done control)
locally

Figure 1.20 Level 2-loT application.

1.9.3 Level 3
In Jevel 3 also, since the data is
and the data is stored on the voluminous, frequency of sensing done by the sensor iscloud.
faster

Based on the data analysis, thecloud.The difference is that the analysis is carried out on the
or mobile application. Some control action can be triggered through the web application
examples are agriculture applications, room freshening solutions
based on odour, etc., where analysis of data occurs in the cloud.
Level 3 IoT application. Figure 1.21 shows an example
 1.9 loT Levels 21

Air conditioner Big data (Bigger than Cloud (Data

Temperature analysis Control and monitoring
Level 1) action (On/Off control)
sensor done here)

Figure 1.21 Level 3-loT application.

1.9.4 Level 4
With every passing level, the volume of data increases and hence the rate at which it is sensed
also increases. At this level, multiple nodes are present which are independent of each other.
These nodes upload data to the cloud. As an example see Fig. 1.22. AIl the sensors upload the
read sensory inputs. Here, cloud storage is preferred as data is huge. Analysis is also carried out
on the cloud and based on the analysis carried out, the control action shall be triggered through
a web application or mobile application.

Temperature
sensor

Moisture sensor
Data collection and Control and monitoring
analysis action (0n/Off
control)

Pressure, temp,
humidity - BME280

Figure 1.22 Level 4-loT application.

1.9.5 Level 5
At this level, the amount of data is extensive and is sensed much faster. Multiple nodes are involved
in the applications categorized under Level 5 and these nodes are independent of each other. The
sensing of data and its storage is the same as in all the previous levels. When an application is com
pletely cloud oriented, it is computationally intensive in real time. Based on the data analysis, the
control action can be triggered through web application or mobile application as in all other levels.
Figure 1.23 shows an example of Level 5 IoT application.
 22 Introduction to Internet of Things

Temperature
sensor Data
Data

Moisture sensor Data collection and

analysis
Control and monitoing
Data

Coordinator node action (On/Of

Control)

Pressure, temp,
humidity - BME280

Figure 1.23 Level 5-loT application.

IEEE has identified about 140 existing standards and projects which can well be
connected to loT and which stay very relevant. One can get complete informatian
FACT
about these standards from: http:/standards.ieee.org/innovate/iot/

Figure 1.24 illustrates IoT in a nutshel.

Sense

Transport

Store

Analyse

Control

Share

Figure 1.24 loT in anutshell.

 1.11 Wireless Sensor Network versus loT 23

European Telecommunications Standards (ETSI) has defined something very

close to IoT's definition. Machine-to-machine communication can be defined
as the communication between entities (more than one) which do not require
manual/human intervention. They are independent. This is automation which
NOTE enables a machine to take decisions on its own without manual intervention.

1.10 Cyber Physical System versus loT

An important question is if IoT is the same as Cyber Physical System (CPS). The answers avail
able are not concrete and hence it is really challenging to find the differences between these two
concepts. There isa misconception that both the terms are the same.
In Section 1.1, we have learned the definition of IoT. It is all about things" getting identified
through an address or identifier. The "thing" can also be accessed from anywhere, anytime by
an authorised party. The information (i.e., the sensed data) can be as simple as an RFID read.
Hence, the complexity involved in IoT applications is minimal. For complex levels of operation
and to address larger network of things; a new term called Cyber Physical System, or CPS, has
been introduced.
It is important to note that CPS is not IoT. CPS has IoT as one of its components. CPS is more
complex than IoT and is much more challenging. It is a combination of multiple engineering
domains coming together, which includes computer science, electronics, electrical, and mechani
cal engineering. The flight of an aeroplane can be seen as a CPS which involves multiple domains
of enginering. It also has IoT as one of the components. Obviously, the complexity increases
by volumes. CPS is much more autonomous than IoT, taking appropriate decisions as and when
needed. It is not merely about identifying things"; it is more about understanding and taking
decisions in a more dynamic way.

For definition of the loT, IEEE visit: https://iot.ieee.org/images/files/pdf/IEEE_

IoT_ Towards Definition_Internet_of_Things_Revision1_27MAY15.pdf.
NOTE

CPS is mainly concerned about the collaborative activity of sensors or actuators

to achieve a certain goal and to do this CPS uses an IoT system to achieve the
collaborative work of the distributed systems.
REMEMBER

1.11 Wireless Sensor Network versus loT

As the name suggests, wireless sensor networks (WSNs) is a network. The network is built with
multiple autonomous sensors; the sensors could be pressure, moisture, temperature, humidity,
sound, and so on. All the sensed data are passed to acentrally located server. The data passing
24
Introduction to Internet of Things
happens in acoOrdinated pattern. WSN is basically composed of nodes (rom a
thousand). Each node has one or more sensors. In short, we can say that WSN ic allfew to a
dinated data collection. about coorfew-
On the other hand, IoT is much more than just data collection and the
intelligent. systems are more

Summary
1. IoT refers to the interconnection of computing devices embedded in
the Internet, enabling them to send and receive data. everyday objects via
2. IoT is not owned by any one engineering branch. It is a reality when
forces and combine efforts. multiple domains jon
3. IoT is all about providing service to any device, anywhere, anybody, and any network
4. IoT has certain characteristics which are important:
a. Connectivity.
b. Intelligence and identity.
C.
Scalability.
d. Dynamic and self-adapting (complexity).
e. Architecture.
Safety.
5. "Things" refer to variety of devices. At times, even humans in the loop becomes a thing. For
anything to qualify as athingit requires identity. The thing" can monitor, measure, etc; for
example, a temperature sensor could be a thing"
6. One should understand that

"THINGS" = HARDWARE + SOFTWARE + DATA + SERVICE

7. IoT stack has seven layers, starting with sensor layer and ending with application layer just
as OSI.
8. Security/personnel safety, privacy, data extraction with consistency from complex envron
ments, connectivity, power requirements, complexity involved and storage are the maj
challenges we face while building an IoT application.
and architec-
9. IoT application can be classified as Level 1, 2, 3, 4 or 5 based on the complexity
ture involved.
10. IoT is all about sense, transport, store, analyse, control and sharing.

Review Questions
1. Define Internet of Things.
2. What are the things" in IoT?
3. Discuss the characteristics of IoT.
4. What are the challenges in building an application with loT?
 Further Reading 25

5. With neat diagrammatic representation,explain IoT stack with appropriate examples for
each layer.
6. How does the classification of loT as Level 1to 5 happen? Explain.
7. Enumerate the role of cloud in loT.
8. Explain the importance of communication protocols when it comes to IoT.

Further Reading
1. Al-Fuqaha, A., Guizani, M., Mohammadi, M., Aledhari, M. and Ayyash, M. 2015. Internet of
things: A survey on enabling technologies, protocols, and applications. IEEE Communica
tions Surveys & Tutorials, 17(4), 2347-2376.
2. Balakrishna, C. 2012. Enabling technologies for smart city services and applications. In Next
Generation Mobile Applications, Services and Technologies (NGMAST), 2012 6th Interna
tional Conference, 223-227, IEEE.
3. Biswas, A. R. and Giaffreda, R. 2014. IoT and cloud convergence: Opportunities
and challenges. In Proceedings of he 2014 IEEE World Forum on Internet of Things
(WF-loT) (WE-I0T), 375-376.
4. Gazis,V., Goertz, M., Huber, M., Leonardi, A., Mathioudakis, K., Wiesmaier, A. and Zeiger, F.
2015. Short paper: IoT: Challenges, projects, architectures. In Intelligence in Next Generation Net
works (1CIlN), 2015 18th International Conference 145-147 EEE.
5. Gubbi, J, Buyya, R., Marusic, S. and Palaniswami, M. 2013. Internet of Things (loT):
Avision, architectural elements, and future directions. Future Generation Computer Systems,
29(7), 1645-1660.
6. Konstantinidis, E.I, Bamparopoulos, G., Billis, A. and Bamidis, PD. 2015. Internet of things
for an age-friendly healthcare. In MIE, 587-591.
7. Lea, R. and Blackstock, M. 2014. City hub: A cloud-based IoT platform for smart cities. In
Cloud Computing Technology and Science (CloudCom), 2014 IEEE óth International Con
ference, 799-804, IEEE.
8. Lee, I. and Lee, K. 2015. The Internet of Things (IoT): Applications, investments, and chal
lenges for enterprises, Business Horizons, 58(4), 431-440.
9. Liu,Y. and Zhou,G. 2012. Key technologies and applications of internet of things. In Intelli
gent Computation Technology and Automation (1CICTA), 2012 Fifth International Confer
ence, 197-200, IEEE.
10. Van Kranenburg, R. and Bassi, A. 2012. IoT challenges. Communicationsin Mobile
Computing, 1(1), 9.
11. Zhou, J,Cao, Z., Dong, X. and Vasilakos, A.V. 2017 Security and privacy for cloud-based IoT:
Challenges. IEEE Communications Magazine, 55(1), 26-33.