UNIT 1: IOT INTRODUCTION 3.

Explain the drivers behind the network

architecture ?
1.What Is IoT ?  To implement any networking concepts
designing and understanding the network
 The Internet of Things (IoT) describes the
architecture is of utmost importance.
network of physical objects—“things”—that are
 The difference between IT and IoT networks is
embedded with sensors, software, and other
technologies for the purpose of connecting and much like the difference between residential
exchanging data with other devices and systems architecture and stadium architecture.
over the internet.  The key difference between IT and IoT is the
 Basic goal of IoT : “connect the unconnected.” data.
This means that objects not currently joined to the  While IT systems are mostly concerned with
Internet will be connected so that they can reliable and continuous support of business
communicate and interact with people and other applications such as email, web, databases and so
objects. on.
 IoT is all about the data generated by sensors and
how that data is used.
2. Explain the impact of IoT ?
 The essence of IoT architectures thus involves
 Internet of Things Positive Impacts: how the data is transported, collected, analyzed,
and ultimately acted upon.
 Effective communication and Instant  IOT ARCHITECTURAL DRIVERS :
messaging services  Scale
 Increase business interactions, save vital time  Security
 Less complicated banking, transactions, and  Constrained Devices and Networks
shopping
 Data
 Access the latest news from anywhere in the
 Legacy Device Support
world
1. Scale:
 Run online course on virtual assistants using
 The scale of a typical IT network is on the
internet
order of several thousand devices—typically
 Professionals have IoT devices in healthcare,
printers, mobile wireless devices, laptops,
e-Commerce, and AI to help in doing jobs
servers, and so on.
easily
 But when a scale of a network goes from a
 Internet of Things Negative Impacts: few thousand endpoints to a few millions the
IT engineers lack a required skills to design a
 Easy availability of age unsuitable content network that is intended to support millions
materials and information of routable IP endpoints.
 Social networks disturb life personally and 2. Security:
professionally  It has often been said that if World War III breaks
 Stealing data or hacking into IoT devices is out, it will be fought in cyberspace.
easy  The frequency and impact of cyber attacks in
 Using the internet to spread bad scenarios via recent years has increased dramatically.
the IoT device is easy  Protecting corporate data from intrusion and theft
is one of the mainfunctions of the IT department.
 IT departments go to great lengths to protect
servers, applications, and the network, setting up
 defense-in-depth models with layers of security 802.11g, we can simply deny him access to the
designed to protect the cyber crown jewels of wireless network, and he will be forced to upgrade.
the corporation. In OT systems, end devices are likely to be on the
 However, despite all the efforts mustered to network for a very long time.
protect networks and data, hackers still find ways
to penetrate trusted networks. 5. Explain in brief IoTWF standardized
architecture ?
3. Constrained Devices and Networks  A seven-layer IoT architectural reference
 Most IoT sensors are designed for a single job, model published by IoTWF architectural
and they are typically small and inexpensive. committee ( Cisco, IBM, Rockwell
 This means they often have limited power, Automation)
CPU, and memory, and they transmit only when  Edge computing
there is something important.  Data storage
 If an IT network has performance constraints,  Access
the solution is simple: Upgrade to a faster
network.  Using this reference model, we are able to
4. Data :
 IoT devices generate a mountain of data.
 In general, most IT shops don’t really care
much about the unstructured chatty data
generated by devices on the network.
 However, in IoT the data is like gold, as it is
what enables businesses to deliver new IoT
services that enhance the customer experience,
reduce cost, and deliver new revenue
opportunities2
 Although most IoT-generated data is
unstructured, the insights it provides through
analytics can revolutionize processes and create
new business models.
 For ex : Imagine a smart city with a few
hundred thousand smart streetlights, all
connected through an IoT network.
achieve the following:
 However, when all this data is combined, it can
 Decompose the IoT problem into smaller
become difficult to manage and analyze
parts
effectively.
 Identify different technologies at each layer
5. Legacy Device Support :
and how they relate to one another
Supporting legacy devices in an IT organization is
 Define a system in which different parts
not usually a big problem.
can be provided by different vendors
If someone’s computer or operating system is
 Have a process of defining interfaces that
outdated, she simply upgrades.
leads to interoperability
If someone is using a mobile device with an
 Define a tiered security model that is
outdated Wi-Fi standard, such as 802.11b or
enforced at the transition points between
levels
Layer 1: Physical Devices and Controllers Layer  Convert event based data to query based data.
 The various endpoint devices and sensors that Layer 5: Data abstraction layer
send and receive information  Multiple data format and sementic from
 The size of these “things” can range from various source and confirms that the data set
almost microscopic sensors to giant machines is complete and store multiple data using
in a factory. virtualiazation.
 Their primary function is generating data and Layer 6:Application Layer
being capable of being queried and/or  Application may monitor,control,provide
controlled over a network. report based on the analysis of data.
Layer 2: Connectivity Layer Layer 7:Collaboration and Process layer
 The most important function of this IoT layer  Consumes and share the application
is the reliable and timely transmission of data. information
Layer 2 function:  This layer can change business process and
1. Communication between layer 1 devices. deliver the benifit of iot .
2. Reliable delivery of information across the  Upper layer deals with handling and
network processing the IOT data generated by the
3. Switching and routing bottom layer.
4. Translation between protocol
5. Network level security 4. Explain in brief one M2M IoT
Layer 3: Edge Computing Layer
standardized architecture ?
 Fog layer
 In an effort to standardize the rapidly growing
 At this layer, the data reduction and
field of machine-to-machine (M2M)
converting network data flows into
communications, the European
information that is ready for storage and
Telecommunications Standards Institute (ETSI)
processing by higher layers.
created the M2M Technical Committee in 2008.
 One of the basic principles of this reference
 The goal of this committee was to create a
model is that information processing is
common architecture that would help accelerate
initiated as early and as close to the edge of
the adoption of M2M applications and devices.
the network as possible .
 oneM2M’s framework focuses on IoT services,
 Another important function that occurs at
applications, and platforms
Layer 3 is the evaluation of data to see if it
 One of the greatest challenges of IoT
can be filtered or aggregated before being
architecture is dealing with the heterogeneity of
sent to a higher layer.
devices, software, and access methods.
 This also allows for data to be reformatted
 By developing a horizontal platform
or decoded, making additional processing
architecture, oneM2M is developing standards
by other systems easier
that allow interoperability at all levels of the
Layer 3 fuction:
IoT stack.
1. Evaluate and reformat data for processing at high
level.
 This is where the oneM2M common services
2. Filter data to reduce traffic higher level
architecture comes in.
processing
Three major domains:
3. Access data for alerting,Notification or other
 The application layer,
action.
 The services layer,
Layer 4: Data Accumulation Layer
 The network layer
 5. Explain in detail the IoT Data
Management and Compute Stack.
 The data generated by IoT sensors is one of
the single biggest challenges in building an
IoT system
 In the case of modern IT networks, the data
sourced by a computer or server is typically
generated by the client/server communications
model, and it serves the needs of the
application
 In sensor networks, the vast majority of data
Application layer : generated is unstructured and of very little use
 The oneM2M architecture gives major on its own
attention to connectivity between devices and  In most cases, the processing location is
their applications. outside the smart object.
 It includes the application-layer protocols and  A natural location for this processing activity
attempts to standardize northbound API is the cloud.
 definitions for interaction with business  Smart objects need to connect to the cloud,
intelligence (BI) systems. and data processing is centralized.
Services layer  One advantage of this model is simplicity.
 This layer is shown as a horizontal framework  Objects just need to connect to a central cloud
across the vertical industry applications application.
 At this layer, horizontal modules include the  That application has visibility over all the IoT
physical network that the IoT applications run nodes and can process all the analytics
on, the underlying management protocols, and needed.
the hardware.  This model also has limitations.
 Examples include backhaul communications Minimizing latency:
via cellular, MPLS networks, VPNs, and so  Milliseconds matter for many types of
on. industrial systems
 Riding on top is the common services layer.  Analyzing data close to the device that
 This conceptual layer adds APIs and collected the data can make a difference
middleware supporting third-party services Conserving network bandwidth:
and applications  Offshore oil rigs generate 500 GB of data
Network layer weekly.
 This is the communication domain for the  Commercial jets generate 10 TB for every 30
IoT devices and endpoints. minutes of flight.
 It includes the devices themselves and the  It is not practical to transport vast amounts of
communications network that links them data from thousands or hundreds of thousands
 Communications infrastructure include of edge devices to the cloud
wireless mesh technologies, such as IEEE Increasing local efficiency:
802.15.4, and wireless point-to-multipoint  Collecting and securing data across a wide
systems, such as IEEE 801.11ah geographic area with different environmental
conditions may not be useful.
 Collecting and securing data across a wide  Analyzing IoT data close to where it is collected
geographic area with minimizes latency, offloads gigabytes of network
 different environmental conditions may not be traffic from the core network, and keeps
useful. sensitive data inside
 Analyzing both areas in the same cloud system .
may not be necessary for immediate efficiency. The defining characteristic of fog computing are as
follows:
 IoT systems function differently.  Contextual location awareness and low
 Several data-related problems need to be latency
addressed:  Geographic distribution
 Bandwidth in last-mile IoT networks is very  Deployment near IoT endpoints
limited.  Wireless communication between the fog
 When dealing with thousands/millions of and the IoT endpoint
devices, available bandwidth may be on order of  Use for real-time interactions
tens of Kbps per device or even less.
 Latency can be very high.
 Big data is getting bigger. The concept of storing
and analyzing all sensor data in the cloud is
impractical

6 .Explain in brief simplified IoT

architecture ?
 Internet of Things (IoT) technology has a wide
variety of applications and use of Internet of
Fog Computing:
Things is growing so faster.
 Distribute data management throughout the IoT
 Depending upon different application areas of
system, as close to the edge of the IP network as
Internet of Things, it works accordingly as per
possible.
it has been designed/developed.
 The best-known embodiment of edge services in
 But it has not a standard defined architecture of
IoT is fog computing.
working which is strictly followed universally.
 Any device with computing, storage, and
 The architecture of IoT depends upon its
network connectivity can be a fog node.
functionality and implementation in different
 Examples include industrial controllers,
sectors. Still, there is a basic process flow
switches, routers, embedded servers, and IoT
based on which IoT is built.
gateways.
 The whole variety of factors affecting IoT functionalities like malware protection, and
architecture, it’s easier and more effective to filtering also some times decision making
find a reliable provider of IoT solutions. based on inputted data and data management
 This decision will significantly reduce the services, etc.
number of resources spent on the way. 3.Data processing Layer –
There are four layers of IOT architecture:  This is processing unit of IoT ecosystem.
 Here data is analyzed and pre-processed before
sending it to data center from where data is
accessed by software applications often termed
as business applications where data is
monitored and managed and further actions are
also prepared.
 So here Edge IT or edge analytics comes into
picture.
4.Application Layer –
 This is last layer of 4 stages of IoT
architecture.
 Data centers or cloud is management stage of
data where data is managed and is used by
end-user applications like agriculture, health
care, aerospace, farming, defense, etc .

7.Challenges in Internet of things

(IoT)?
1. Application layer
 The Internet of Things (IoT) has fast grown to
2. Data processing layer
be a large part of how human beings live,
3. Network layer
communicate and do business.
4. Sensing layer
 All across the world, web-enabled devices are
1.Sensing Layer –
turning our global rights into a greater
 Sensors, actuators, devices are present in this
switched-on area to live in.
Sensing layer.
 These Sensors or Actuators accepts Security challenges in IoT :
data(physical/environmental parameters), 1. Lack of encryption –
processes data and emits data over network.  Although encryption is a great way to
2.Network Layer – prevent hackers from accessing data, it is
 Internet/Network gateways, Data Acquisition also one of the leading IoT security
System (DAS) are present in this layer. challenges.
 DAS performs data aggregation and  The result is an increase in attacks where
conversion function (Collecting data and hackers can easily manipulate the
aggregating data then converting analog data algorithms that were designed for
of sensors to digital data etc). protection.
 Advanced gateways which mainly opens up 2. Insufficient testing and updating –
connection between Sensor networks and  With the increase in the number of
Internet also performs many basic gateway IoT(internet of things) devices, IoT
 manufacturers are more eager to produce  IoT applications must be developed,
and deliver their device as fast as they can keeping in mind the technological changes
without giving security too much of of the future.
although.  Its development requires a balance of
 Most of these devices and IoT products do hardware and software functions.
not get enough testing and updates and are  It is a challenge for IoT application
prone to hackers and other security issues. developers to ensure that the device and
Design challenge in IoT : IoT platform drivers the best performance
1. Battery life is a limitation – despite heavy device rates and fixings.
 Issues in packaging and integration of 3. Data collection and processing –
small-sized chip with low weight and less  In IoT development, data plays an
power consumption. important role. What is more critical here
 If you’ve been following the mobile space, is the processing or usefulness of stored
you’ve likely see how every yr it looks like data.
there’s no restriction in terms of display  Along with security and privacy,
screen size. development teams need to ensure that
2. Increased cost and time to market – they plan well for the way data is
 Embedded systems are lightly constrained collected, stored or processed within an
by cost. environment.
 The need originates to drive better
approaches when designing the IoT devices 8. Explain the difference between
in order to handle the cost modelling or Services layer and the Network layer for
cost optimally with digital electronic
IoT architecture ?
components.
Services layer:
3. Security of the system –
 The horizontal modules include the physical
 Systems have to be designed and
network that the IoT applications run on, the
implemented to be robust and reliable and
underlying management protocols, and the
have to be secure with cryptographic
hardware. Examples include backhaul
algorithms and security procedures.
communications via cellular, MPLS networks,
 It involves different approaches to secure
VPNs, and so on.
all the components of embedded systems
 This conceptual layer adds APIs and
from prototype to deployment.
middleware supporting third-party services
and applications.
Deployment challenges in IoT :
 One of the stated goals of oneM2M is to
1. Connectivity –
“develop technical specifications which
 It is the foremost concern while connecting
address the need for a common M2M Service
devices, applications and cloud platforms.
Layer that can be readily embedded within
 Connected devices that provide useful front
various hardware and software nodes.
and information are extremely valuable.
 But poor connectivity becomes a challenge Network layer:
where IoT sensors are required to monitor
 It includes the devices and the
process data and supply information.
communications infrastructure which include
2. Cross platform capability –
wireless mesh technologies, such as IEEE
 802.15.4, and wireless point-to-multipoint  A device that provides a usable output in
systems, such as IEEE 801.11ah. response to a specified measurement.
 Also included are wired device connections,  The output of the sensor is a signal which is
such as IEEE 1901 power line converted to a human-readable form like
communications. changes in characteristics, changes in
 The device domain includes the gateway resistance, capacitance, impedance etc.
device, which provides communications up  The sensor attains a physical parameter and
into the core network and acts as a converts it into a signal suitable for processing
demarcation point between the device and (e.g. electrical, mechanical, optical) the
network domains. characteristics of any device or material to
detect the presence of a particular physical
quantity.
UNIT 2: SMART OBJECT
Categories of sensor:
1. Write a short note on smart objects.
 Active or passive:
Sensors can be categorized based on whether
 The concept of smart in IoT is used for
they produce an energy output and typically
physical objects that are active, digital,
require an external power supply (active) or
networked, can operate to some extent
whether they simply receive energy and
autonomously, reconfigurable and has local
typically require no external power supply
control of the resources.
(passive).
 The smart objects need energy, data storage,
 Invasive or non-invasive:
etc.
Sensors can be categorized based on whether
 A smart object is an object that enhances the
a sensor is part of the environment it is
interaction with other smart objects as well as
measuring (invasive) or external to it (non-
with people also.
invasive).
 The world of IoT is the network of
 Contact or no-contact:
interconnected heterogeneous objects (such as
Sensors can be categorized based on whether
smart devices, smart objects, sensors,
they require physical contact with what they
actuators, RFID, embedded computers, etc.)
are measuring (contact) or not (no-contact).
uniquely addressable and based on standard
 Absolute or relative:
communication protocols.
Sensors can be categorized based on whether
 Smart objects are utilized widely to transform
they measure on an absolute scale (absolute)
the physical environment around us to a digital
or based on a difference with a fixed or
world using the Internet of things (IoT)
variable reference value (relative).
technologies.
 Area of application:
 A smart object carries blocks of application
Sensors can be categorized based on the
logic that make sense for their local situation
specific industry or vertical where they are
and interact with human users.
being used.
2. Explain different types of sensor?  How sensors measure:
 sensors are used in the architecture of IOT Sensors can be categorized based on the
devices. physical mechanism used to measure sensory
input (for example, thermoelectric,
 Sensors are used for sensing things and
devices etc.
 electrochemical, piezoresistive, optic, electric,
fluid mechanic, photoelastic). Image Sensors
 What sensors measure:  IoT image sensors are used to convert images
Sensors can be categorized based on their into electronic signals.
applications or what physical variables they  These are then either displayed or become
measure. electronically stored files.
 The most common use of image sensors is in
Types of sensor: digital cameras and IoT WiFi modules.
Temperature Sensors
Level Sensors
 In the past, IoT temperature sensors have been  Level sensors are used to detect the levels of
used for heat, ventilation, and air conditioning certain types of objects.
systems (HVAC), refrigerators, and other
similar devices used for environmental control.  These include liquids, granular materials, and
 However, the emergence of IoT has seen its powders.
role expand.  As you can imagine, this kind of sensor is
 Nowadays, you can find temperature sensors useful and used in many different industries
throughout industries such as manufacturing and applications.
and agriculture.  These include:
 Oil manufacturing
Pressure Sensors  Beverage manufacturing
 Food manufacturing
 An IoT pressure sensor is any device that senses
 Water treatment
pressure and converts it into an electric signal.
 The level of voltage given out by the sensor Gas sensor:
depends on the level of pressure applied.
 Gas sensors monitor and detect changes in the
 These sensors enable IoT systems that monitor air.
systems and devices that are pressure propelled.
 These sensors are vital to our safety as they’re
 If there’s any deviation from standard pressure able to detect the presence of potentially
ranges, the device notifies the administrator of harmful or even toxic gases.
the problem.
 Gas sensors are most commonly used within
the mining, oil and gas, and chemical research.
Humidity Sensors
 IoT humidity sensors measure the amount of Motion Detector Sensors
water vapor in the air.  Not to be confused with proximity sensors,
 In scientific terms, they measure Relative motion detector sensors are used to detect
Humidity (RH). physical movement in a given area.
 This kind of sensor is usually used in addition to  In turn, this then sets off an electronic signal.
IoT temperature sensors when a manufacturing  The most obvious use of this is within the
process requires absolute perfect working security industry.
conditions.  Businesses use motion detector sensors in
 They’re usually found in heating, ventilation, areas where there should be no movement.
and air conditioning (HVAC) systems – in both  These IoT sensors can also be found in many
the home and business settings. devices within modern commercial buildings
 hardware, and so on are critical design
3. Explain the characteristics of smart objects. elements.
Processing unit:
 A smart object has some type of processing unit 5. Write a short note on wireless sensor networks.
for acquiring data, processing and analyzing  Wireless sensor networks are made up of
sensing information received by the sensor(s), wirelessly connected smart objects, which are
coordinating control signals to any actuators, and sometimes referred to as motes.
controlling a variety of functions on the smart  The fact that there is no infrastructure to
object, including the communication and power consider with WSNs is surely a powerful
systems. advantage for flexible deployments, but there
Sensor(s) and/or actuator(s): are a variety of design constraints to consider
 A smart object is capable of interacting with the with these wirelessly connected smart objects.
physical world through sensors and actuators.  limitations of the smart objects in WSNs:
 As described in the previous sections, a sensor o Limited processing power
learns and measures its environment, whereas an o Limited memory
actuator is able to produce some change in the o Lossy communication
physical world. o Limited transmission speeds
o Limited power
Communication device:  Smart objects with limited processing, memory,
 The communication unit is responsible for power, and so on are often referred to as
connecting a smart object with other smart constrained nodes.
objects and the outside world (via the network).  These limitations greatly influence how WSNs
 Power source: Smart objects have components are designed, deployed, and utilized.
that need to be powered. o The fact that individual sensor nodes are
 Interestingly, the most significant power typically so limited is a reason that they are
consumption usually comes from the often deployed in very large numbers.
communication unit of a smart object. o As the cost of sensor nodes continues to
Power source: decline, the ability to deploy highly
 Smart objects have components that need to be redundant sensors becomes increasingly
powered. feasible.
 Interestingly, the most significant power o Because many sensors are very inexpensive
consumption usually comes from the and correspondingly inaccurate, the ability to
communication unit of a smart object. deploy smart objects redundantly allows for
 As with the other three smart object building increased accuracy
blocks, the power requirements also vary o Such large numbers of sensors permit the
greatly from application to application. introduction of hierarchies of smart objects.
 Typically, smart objects are limited in power,
are deployed for a very long time, and are not
easily accessible.
 This combination, especially when the smart
object relies on battery power, implies that
power efficiency, judicious power management,
sleep modes, ultra-low power consumption
6. Difference between sensor and actuator?  Should indoor versus outdoor deployments be
differentiated?
These are the questions asked when discussing wired
and wireless access technologies.
 The simplest approach to answering these types
of questions is to categorize these technologies
as shown in Fig, breaking them down into the
following range

1. Short range:
7. What are the different criteria for connecting  The classical wired example is a serial cable.
smart objects?  Wireless short-range technologies are often
 IoT devices and sensors must be connected to the considered as an alternative to a serial cable,
network for their data to be utilized. supporting tens of meters of maximum distance
 In addition to the wide range of sensors, between two devices.
actuators, and smart objects that make up IoT,  Examples of short-range wireless technologies
there are also a number of different protocols are IEEE 802.15.1 Bluetooth and IEEE 802.15.7
used to connect them. Visible Light Communications (VLC)
Communication criteria:  These shortrange communication methods are
 Range found in only a minority of IoT installations. In
 Frequency Bands some cases, they are not mature enough for
production deployment.
 Power Consumption
 Topology
2. Medium range:
 Constrained Devices
 This range is the main category of IoT access
 Constrained-Node Networks
technologies.
Range :
 In the range of tens to hundreds of meters,
 How far does the signal need to be
many specifications and implementations are
propagated?
available.
 That is, what will be the area of coverage for a
 The maximum distance is generally less than
selected wireless technology?
1 mile between two devices
  Examples of medium-range wireless  These frequencies are used in many
technologies include IEEE 802.11 Wi-Fi, communications technologies for short-range
IEEE 802.15.4, and 802.15.4g WPAN. devices (SRDs).
 Wired technologies such as IEEE 802.3  Unlicensed means that no guarantees or
Ethernet and IEEE 1901.2 Narrowband Power protections are offered in the ISM bands for
Line Communications (PLC) may also be device communications
classified as medium range, depending on  ISM bands for IoT access
their physical media characteristics.  2.4 GHz band as used by IEEE 802.11b/g/n
3. Long range: Wi-Fi  IEEE 802.15.1 Bluetooth
 Distances greater than 1 mile between two  IEEE 802.15.4 WPAN
devices require long-range technologies.  Unlicensed spectrum is usually simpler to
 Wireless examples are cellular (2G, 3G, 4G) deploy than licensed because it does not
and some applications of outdoor IEEE require a service provider.
802.11 Wi-Fi and Low-Power Wide-Area  The frequency of transmission directly
(LPWA) technologies. impacts how a signal propagates and its
 LPWA communications have the ability to practical maximum range.
communicate over a large area without
consuming much power. Power Consumption
 These technologies are therefore ideal for  Powered nodes and battery-powered nodes
battery-powered IoT sensors  A powered node has a direct connection to a
power source, and communications are usually
Frequency Bands :
not limited by power consumption criteria.
 Radio spectrum is regulated by countries
 However, ease of deployment of powered
and/or organizations, such as the International
nodes is limited by the availability of a power
Telecommunication Union (ITU) and the
source, which makes mobility more complex
Federal Communications Commission (FCC).
 Battery-powered nodes bring much more
 These groups define the regulations and
flexibility to IoT devices.
transmission requirements for various
 These nodes are often classified by the required
frequency bands .
lifetimes of their batteries.
 For example, portions of the spectrum are
 A new wireless environment known as Low-
allocated to types of telecommunications such
Power WideArea (LPWA)
as radio, television, military, and so on.
 Battery-powered nodes are often placed in a
 Focusing on IoT access technologies, the
“sleep mode” to preserve battery life when not
frequency bands leveraged by wireless
transmitting
communications are split between licensed and
unlicensed bands.  Wired IoT access technologies consisting of
powered nodes are not exempt from power
 Licensed spectrum is generally applicable to
optimization
IoT longrange access technologies .
 In order to utilize licensed spectrum, users
Topology
must subscribe to services when connecting
 Among the access technologies available for
their IoT devices .
connecting IoT devices, three main topology
 The ITU has also defined unlicensed spectrum
schemes are dominant: star, mesh, and peer-to-
for the industrial, scientific, and medical (ISM)
peer.
portions of the radio bands.
 For long-range and short-range technologies, a  They have to be lightweight with energy-
star topology is prevalent. efficient protocols and less bandwidth
 Star topologies utilize a single central base consumption.
station or controller to allow communications  Using Class 2 devices might reduce
with endpoints . development costs and increase the
 For medium-range technologies, a star, peer-to- interoperability.
peer, or mesh topology is common.
 Peer-to-peer topologies allow any device to Constrained-Node Networks
communicate with any other device as long as  IEEE 802.15.4 and 802.15.4g RF, IEEE
they are in range of each other. 1901.2a PLC, LPWA, and IEEE 802.11ah
access technologies
 Constrained-node networks are often referred
to as low-power and lossy networks (LLNs).
 Low power – battery powered constraints
 Lossy network -- network performance may
suffer from interference and variability due to
harsh radio environments
 Protocols that can be used for constrained-node
networks must be evaluated in the context of the
following characteristics:
 data rate and throughput, latency and
Constrained device: determinism, and overhead and payload.
Class 0: Data Rate and Throughput
 Class 0 devices have constraints in  The data rates available from IoT access
memory(<<10KiB of RAM and <<100KiB of technologies range from 100 bps with protocols
Flash) and processing capabilities. such as Sigfox to tens of megabits per second
 These devices has severe constraints to with technologies such as LTE and IEEE
communicate securely with internet, so they 802.11ac
typically pre-configured and are connected  However, the actual throughput is less
to proxies, gateways, or servers for internet  Technologies not particularly designed for IoT,
communication. such as cellular and Wi-Fi, match up well to IoT
Class 1: applications with high bandwidth requirements
 Class 1 devices can have low power IoT stack  Short-range technologies can also provide
[UDP, CoAP, leight weigh security protocols medium to high data rates that have enough
like DTLS etc] throughput to connect a few endpoints.
 But quite constrained in code space and  For example, Bluetooth sensors that are now
processing capabilities to employing a full appearing on connected wearables fall into this
protocol stack such as using HTTP, TLS, and category.
related security protocols and data  In this case, the solutions focus more on
representations with out a gateway. footprint and battery lifetime than on data rate.
Class 2:  The IoT access technologies developed for
 Class 2 devices are less constrained and can constrained nodes are optimized for low power
perform at par with mobiles phones/notebooks consumption, but they are also limited in terms
in supporting most the protocol stacks. of data rate
 Another characteristic of IoT devices is that a LoRaWAN Layers:
majority of them initiate the communication. Physical Layer
 Semtech LoRa modulation is based on chirp
Latency and Determinism spread spectrum modulation
 Much like throughput requirements, latency  Chirp - Compressed High Intensity Radar
expectations of IoT applications should be Pulse
known when selecting an access technology.  Lower data rate and increase the
 This is particularly true for wireless networks, communication distance Understanding LoRa
where packet loss and retransmissions due to gateways is critical to understanding a
interference, collisions, and noise are normal LoRaWAN system.
behaviors.  A LoRa gateway is deployed as the center hub
Overhead and Payload . of a star network architecture.
 When considering constrained access network  It uses multiple transceivers and channels and
technologies, it is important to review the MAC can demodulate multiple channels at once or
payload size characteristics required by applications. even demodulate multiple signals on the same
 You should be aware of any requirements for IP. channel simultaneously.
 The minimum IPv6 MTU size is expected to be  The data rate in LoRaWAN varies depending
1280 bytes. on the frequency bands and adaptive data rate
(ADR).
8. Explain the characteristics of  ADR is an algorithm that manages the data
LoraWAN. rate and radio signal for each endpoint.
 Low-Power Wide-Area (LPWA) adapted for  An important feature of LoRa is its ability to
long-range and battery powered endpoints handle various data rates via the spreading
 LoRaWAN is unlicensed-band LPWA factor.
technology  Devices with a low spreading factor (SF)
 LoRa was a physical layer, or Layer 1, achieve less distance in their communications
modulation that was developed by a French but transmit at faster speeds, resulting in less
company named Cycleo airtime.
 Later, Cycleo was acquired by Semtech  A higher SF provides slower transmission
 Optimized for long-range, two-way rates but achieves a higher reliability at longer
communications and low power consumptio distances.

MAC Layer
 The MAC layer is defined in the LoRaWAN
specification.
 This layer takes advantage of the LoRa
physical layer and classifies LoRaWAN
endpoints to optimize their battery life and
ensure downstream communications to the
LoRaWAN endpoints.
 The LoRaWAN specification documents three
classes of LoRaWAN devices:
 Class A: acknowledged, and “unconfirmed” signifies that
 This class is the default implementation. the end device does not need to acknowledge.
Optimized for battery powered nodes, it allows  “up/down” is simply a directional notation
bidirectional communications, where a given identifying whether the message flows in the
node is able to receive downstream traffic after uplink or downlink path.
transmitting.  Uplink messages are sent from endpoints to the
 Two receive windows are available after each network server and are relayed by one or more
transmission LoRaWAN gateways.
Class B:  Downlink messages flow from the network
 This class was designated “experimental” in server to a single endpoint and are relayed by
LoRaWAN 1.0.1 until it can be better only a single gateway
defined.  LoRaWAN endpoints are uniquely addressable
 A Class B node or endpoint should get through a variety of methods, including the
additional receive windows compared to following:
Class A, but gateways must be synchronized o An endpoint can have a global end device ID
through a beaconing process or DevEUI represented as an IEEE EUI-64
Class C: address.
 This class is particularly adapted for powered o An endpoint can have a global application ID
nodes. or AppEUI represented as an IEEE EUI-64
 This classification enables a node to be address that uniquely identifies the application
continuously listening by keeping its receive provider, such as the owner, of the end device
window open when not transmitting. o In a LoRaWAN network, endpoints are also
 LoRaWAN messages, either uplink or downlink, known by their end device address, known as a
have a PHY payload composed of a 1-byte MAC DevAddr, a 32-bit address.
header, a variable-byte MAC payload, and a o The 7 most significant bits are the network
MIC that is 4 bytes in length. identifier (NwkID), which identifies the
 The MAC payload size depends on the LoRaWAN network.  The 25 least significant
frequency band and the data rate, ranging from bits are used as the network address
59 to 230 bytes for the 863–870 MHz band and (NwkAddr) to identify the endpoint in the
19 to 250 bytes for the 902–928 MHz band. network.
Security
 LoRaWAN endpoints must implement two
layers of security, protecting communications
and data privacy across the networ

 In version 1.0.x, LoRaWAN utilizes six MAC

message types.
 LoRaWAN devices use join request and join
accept messages for activation and joining the
network.
 The other message types are unconfirmed data
up/down and confirmed data up/down.  A
“confirmed” message is one that must be
 The first layer, called “network security” but Properties:
applied at the MAC layer, guarantees the 1. Standardization and alliances:
authentication of the endpoints by the  It specifies low-data-rate PHY and MAC layer
LoRaWAN network server. requirements for wireless personal area
 Also, it protects LoRaWAN packets by networks(WPAN).
performing encryption based on AES  IEEE 802.15. Protocol Stacks include:
 Each endpoint implements a network session key ZigBee:
(NwkSKey), used by both itself and the  ZigBee is a Personal Area Network task group
LoRaWAN network server. with a low rate task group 4.
 This can be achieved through one of the two  It is a technology of home networking. ZigBee
join mechanisms: is a technological standard created for
Activation by personalization (ABP): controlling and sensing the network.
 Endpoints don’t need to run a join procedure as  As we know that ZigBee is the Personal Area
their individual details, including DevAddr and network of task group 4 so it is based on IEEE
the NwkSKey and AppSKey session keys, are 802.15.4 and is created by Zigbee Alliance.
preconfigured and stored in the end device. 6LoWPAN:
 This same information is registered in the  The 6LoWPAN system is used for a variety of
LoRaWAN network server applications including wireless sensor
Over-the-air activation (OTAA): networks.
 Endpoints are allowed to dynamically join a  This form of wireless sensor network sends
particular LoRaWAN network after data as packets and uses IPv6 – providing the
successfully going through a join procedure. basis for the name – IPv6 over Low power
 The join procedure must be done every time a Wireless Personal Area Networks.
session context is renewed. ZigBee IP:
 During the join process, which involves the  Zigbee is a standards-based wireless
sending and receiving of MAC layer join technology that was developed for low-cost
request and join accept messages, the node and low-power wireless machine-to-machine
establishes its credentials with a LoRaWAN (M2M) and internet of things (IoT) networks.
network server, exchanging its globally unique ISA100.11a:
DevEUI, AppEUI, and AppKey.  It is a mesh network that provides secure
 The AppKey is then used to derive the session wireless communication to process control.
NwkSKey and AppSKey keys Wireless HART:
 It is also a wireless sensor network technology,
9. Explain the IEEE 802.15.4 technology in that makes use of time-synchronized and self-
detail. organizing architecture.
 IEEE 802.15.4 is a low-cost, low-data-rate Thread:
wireless access technology for devices that  Thread is an IPv6-based networking protocol
are operated or work on batteries. for low-power Internet of Things devices in
 This describes how low-rate wireless IEEE 802.15. 4-2006 wireless mesh network.
personal area networks (LR-WPANs) Thread is independent.
function. 2. Physical Layer:
 This standard enables a wide range of PHY
options in ISM bands, ranging from 2.4 GHz to
sub-GHz frequencies.
 IEEE 802.15.4 enables data transmission  frame of data
speeds of 20 kilobits per second, 40 kilobits  Frame for a beacon
per second, 100 kilobits per second, and 250  Frame of acknowledgment
kilobits per second. The fundamental structure  Frame for MAC command
assumes a 10-meter range and a data rate of 4. Topology:
250 kilobits per second.  Networks based on IEEE 802.15.4 can be
 To further reduce power usage, even lower developed in a star, peer-to-peer, or mesh
data rates are possible. topology. Mesh networks connect a large
 IEEE 802.15.4 regulates the RF transceiver number of nodes.
and channel selection, and even some energy  This enables nodes that would otherwise be
and signal management features, at the out of range to interact with each other to use
physical layer. intermediate nodes to relay data.
 Based on the frequency range and data 5. Security:
performance needed, there are now six PHYs  For data security, the IEEE 802.15.4 standard
specified. employs the Advanced Encryption Standard
 Four of them employ frequency hopping (AES) with a 128-bit key length as the basic
techniques known as Direct Sequence Spread encryption technique.
Spectrum (DSSS).  Activating such security measures for
 Both PHY data service and management 802.15.4 significantly alters the frame format
service share a single packet structure so that and uses a few of the payloads.
they can maintain a common simple interface  The very first phase in activating AES
with MAC. encryption is to use the Security Enabled field
in the Frame Control part of the 802.15.4
3. MAC layer: header
 The MAC layer provides links to the PHY  For safety, this field is a single bit which is
channel by determining that devices in the assigned to 1.
same region will share the assigned  When this bit is set, by taking certain bytes
frequencies. from its Payload field, a field known as the
 The scheduling and routing of data packets Auxiliary Security Header is formed
are also managed at this layer. following the Source Address field.
 The 802.15.4 MAC layer is responsible for a 6. Competitive Technologies:
number of functions like:  The IEEE 802.15.4 PHY and MAC layers
 Beaconing for devices that operate as serve as a basis for a variety of networking
controllers in a network. profiles that operate in different IoT access
 used to associate and dissociate PANs scenarios.
with the help of devices.  DASH7 is a competing radio technology
 The safety of the device. with distinct PHY and MAC layers.
 Consistent communication between two Advantages of IEEE 802.15.4:
MAC devices that are in a peer-to-peer IEEE 802.15.4 has the following advantages:
relationship.  cheap cost
 Several established frame types are used by  long battery life,
the MAC layer to accomplish these functions.  Quick installation
In 802.15.4, there are four different types of  simple
MAC frames:  extensible protocol stack
Disadvantages of IEEE 802.15.4: UNIT 3: IP LAYER
IEEE 802.15.4’s drawbacks include:
 IEEE 802.15.4 causes interference and 1. What is the need for optimization in
multipath fading. IP?
 doesn’t employ a frequency-hopping approach.
 Internet of Things will largely be built on the
 unbounded latency
Internet Protocol suite
 interference susceptibility
 In coping with the integration of non-IP
devices, may need to deal with the limits at the
Applications of IEEE 802.15.4: device and network levels that IoT often
IEEE 802.15.4 Applications: imposes.
 Wireless sensor networks in the industry  Therefore, optimizations are needed at various
 Building and home automation layers of the IP stack to handle the restrictions
 Remote controllers and interacting toys that are present in IoT networks.
 Automotive networks  The following concepts take a detailed look at
why optimization is necessary for IP:
The architecture of LR-WPAN Device:  Constrained Nodes
 Constrained Networks
 IP Versions

Constrained Nodes :
 IoT having different classes of devices coexist.
 Depending on its functions in a network, a
“thing” architecture may or may not offer similar
characteristics compared to a generic PC or
server in an IT environment.
 Another limit is that this network protocol stack
on an IoT node may be required to communicate
through an unreliable path.
 IoT constrained nodes can be classified as
follows:
 Devices that are very constrained in
resources, may communicate infrequently to
transmit a few bytes, and may have limited
security and management capabilities:
This drives the need for the IP adaptation
model, where nodes communicate through
gateways and proxies.
 Devices with enough power and capacities to
implement a strippeddown IP stack or non-
IP stack:
In this case, you may implement either an
optimized IP stack and directly communicate
with application servers (adoption model) or go
for an IP or non-IP stack and communicate
 through gateways and proxies (adaptation  Due to the low bandwidth, a constrained
model). network that overreacts can lead to a network
 Devices that are similar to generic PCs in collapse which makes the existing problem
terms of computing and power resources but worse.
have constrained networking capacities, such  Control plane traffic must also be kept at a
as bandwidth: minimum; otherwise, it consumes the
These nodes usually implement a full IP stack bandwidth that is needed by the data traffic.
(adoption model), but network design and  The power consumption in battery-powered
application behaviors must cope with the nodes:
bandwidth constraints.  Any failure or verbose control plane
 In constrained nodes, the costs of computing protocol may reduce the lifetime of the
power, memory, storage resources, and power batteries.
consumption are generally decreasing.  This led to work on optimizing protocols for
 At the same time, networking technologies IoT
continue to improve and offer more bandwidth IP Versions :
and reliability.  IETF has been working on transitioning the
Constrained Networks : Internet from IP version 4 to IP version 6.
 Low-speed connections (like low-speed  The main driving force has been the lack of
modems) demonstrated that IP could run over address space in IPv4 as the Internet has
low-bandwidth networks. grown.
 High-speed connections are not usable by some  IPv6 has a much larger range of addresses that
IoT devices in the last mile. should not be exhausted for the foreseeable
 The reasons include the implementation of future.
technologies with low bandwidth, limited  Today, both versions of IP run over the
distance and bandwidth due to regulated transmit Internet, but most traffic is still IPv4 based.
power, and lack of or limited network services.  Internet of Things has the Internet itself and
 A constrained network can have high latency support both IPv4 and IPv6 versions
and a high potential for packet loss. concurrently.
 Constrained networks are often referred to as  The following are some of the main factors
low-power and lossy networks (LLNs). applicable to IPv4 and IPv6 support in an IoT
 Constrained networks operate between a few solution:
kbps and a few hundred kbps and may utilize a  Application Protocol
star, mesh, or combined network topologies,  Cellular Provider and Technology
ensuring proper operations.  Serial Communications
 In constrained network, it is not unusual for the  IPv6 Adaptation Layer
packet delivery rate (PDR) to oscillate between Application Protocol:
low and high percentages.  IoT devices implementing Ethernet or Wi-Fi
 Large bursts of unpredictable errors and even interfaces can communicate over both IPv4 and
loss of connectivity at times may occur, where IPv6, but the application protocol may dictate the
packet delivery variation may fluctuate greatly choice of the IP version.
during the course of a day.  For example, SCADA protocols such as
Latency and control plane reactivity: DNP3/IP (IEEE 1815), Modbus TCP, or the IEC
 One of the golden rules in a constrained 60870-5-104 standards are specified only for
network is to “underreact to failure.” IPv4.
 So, there are no known production  IPv6 Adaptation Layer:
implementations by vendors of these protocols  IPv6-only adaptation layers for some physical
over IPv6 today. and data link layers for recently standardized
 For IoT devices with application protocols IoT protocols support only IPv6.
defined by the IETF, such as HTTP/HTTPS,  While the most common physical and data link
CoAP, MQTT, and XMPP, both IP versions are layers (Ethernet, Wi-Fi, and so on) stipulate
supported. adaptation layers for both versions, newer
 The selection of the IP version is only technologies, such as IEEE 802.15.4 (Wireless
dependent on the implementation. Personal Area Network), IEEE 1901.2, and
 Cellular Provider and Technology: ITU G.9903 (Narrowband Power Line
 IoT devices with cellular modems are dependent Communications) only have an IPv6
on the generation of the cellular technology as adaptation layer specified.
well as the data services offered by the provider.  This means that any device implementing a
 For the first three generations of data services technology that requires an IPv6 adaptation
GPRS, Edge, and 3G IPv4 is the base protocol layer must communicate over an IPv6-only
version. subnet work.
 Consequently, if IPv6 is used with these  This is reinforced by the IETF routing
generations, it must be tunneled over IPv4. protocol for LLNs, RPL, which is IPv6 only.
 On 4G/LTE networks, data services can use IPv4
or IPv6 as a base protocol, depending on the 2. What is the need for optimizing IP for IoT?
provider.  While the Internet Protocol is key for a
 Serial Communications: successful Internet of Things, constrained nodes
 Data is transferred using either proprietary or and constrained networks mandate optimization
standards-based protocols, such as DNP3, at various layers and on multiple protocols of the
Modbus, or IEC 60870-5-101. IP architecture
 In the past, communicating this serial data over
any sort of distance could be handled by an
analog modem connection.
 However, as service provider support for analog
line services has declined, the solution for
communicating with these legacy devices has
been to use local connections.
 To make this work, you connect the serial port of
the legacy device to a nearby serial port on a
piece of communications equipment, typically a
router.
 This local router then forwards the serial traffic
over IP to the central server for processing.
 Encapsulation of serial protocols over IP The following optimizations technique of IP already
leverages mechanisms such as raw socket TCP available:
or UDP.  From 6LoWPAN to 6Lo
 While raw socket sessions can run over both  Header Compression
IPv4 and IPv6, current implementations are  Fragmentation
mostly available for IPv4 only  Mesh Addressing
  Mesh-Under Versus Mesh-Over Routing  Header compression,
 6Lo Working Group  Fragmentation,
 6TiSCH  Mesh addressing.
 RPL  These headers can be stacked in the adaptation
 Objective Function (OF) layer to keep these concepts separate while
 Rank enforcing a structured method for expressing
 RPL Headers each capability.
 Metrics  Depending on the implementation, all, none, or
 Authentication and Encryption on any combination of these capabilities and their
Constrained Nodes corresponding headers can be enabled.
 ACE
 DICE

From 6LoWPAN to 6Lo :

 In the IP architecture, the transport of IP
packets over any given Layer 1 (PHY) and
Layer 2 (MAC) protocol must be defined.
 The initial focus of the 6LoWPAN working
group was to optimize the transmission of
IPv6 packets over constrained networks such
as IEEE 802.15.4.
 IoT-related protocols follow a similar
process. Figure 3.3 6LoWPAN Header Stacks
 The main difference is that an adaptation
layer designed for IoT may include some Header Compression:
optimizations to deal with constrained nodes  Shrinks the size of IPv6’s 40-byte headers and
and networks. User Datagram Protocol’s (UDP’s) 8-byte
headers down as low as 6 bytes combined in
some cases.
 Header compression for 6LoWPAN is only
defined for an IPv6 header and not for IPv4.
 However, a number of factors affect the amount
of compression, such as implementation of RFC,
whether UDP is included, and various IPv6
addressing scenarios.

 6LoWPAN works by taking advantage of shared

information known by all nodes from their
participation in the local network.
Fig.Comparison of an IoT Protocol Stack Utilizing  In addition, it omits some standard header fields
6LoWPAN and an IP Protocol Stack by assuming commonly used values.
 The 6LoWPAN working group published several
RFCs (Request for Comments by IETF), but
RFC defines frame headers for the capabilities of
  To remedy this situation, large IPv6 packets
must be fragmented across multiple 802.15.4
frames at Layer 2.

Figure 3.4 6LoWPAN Header Compression

 At the top of, a 6LoWPAN frame without any

header compression enabled:
 The full 40-byte IPv6 header and 8-byte UDP Figure 3.5 6LoWPAN Fragmentation Header
header are visible.
 The 6LoWPAN header is only a single byte in  The fragment header utilized by 6LoWPAN is
this case. composed of three primary fields:
 Uncompressed IPv6 and UDP headers leave  Datagram Size:
only 53 bytes of data payload out of the 127- The 1-byte field specifies the total size of the
byte maximum frame size in the case of IEEE unfragmented payload
802.15.4.  Datagram Tag:
 The bottom half of shows a frame where header identifies the set of fragments for a payload.
compression enabled:  Datagram Offset:
 The 6LoWPAN header increases to 2 bytes to field delineates how far into a payload a
accommodate the compressed IPv6 header, and particular fragment occurs.
UDP has been reduced in half, to 4 bytes from  The 6LoWPAN fragmentation header field itself
8. uses a unique bit value to identify that the
 Most importantly, the header compression has subsequent fields behind it are fragment fields as
allowed the payload to more than double, from opposed to another capability, such as header
53 bytes to 108 bytes, which is obviously much compression.
more efficient.  In the first fragment, the Datagram Offset field is
Fragmentation: not present because it would simply be set to 0.
 The maximum transmission unit (MTU) for an  This results in the first fragmentation header for
IPv6 network must be at least 1280 bytes. an IPv6 payload being only 4 bytes long.
 The term MTU defines the size of the largest
protocol data unit that can be passed. Mesh Addressing:
 For IEEE 802.15.4, 127 bytes is the MTU.  The purpose of the 6LoWPAN mesh addressing
 A problem because of IPv6, with a much larger function is to forward packets over multiple
MTU, is carried inside the 802.15.4 frame with hops.
a much smaller one.  Three fields are defined for this header:
  The edge gateway must also implement a
mechanism to translate between the configured
Layer 2 protocol and any IP routing mechanism
implemented on other Layer 3 IP interfaces.
 Mesh-over or route-over scenarios,
 IP Layer 3 routing is utilized for computing
reachability and then getting packets
forwarded to their destination, either inside
or outside the mesh domain.
 Each full-functioning node acts as an IP
 Hop Limit: router, so each link layer hop is an IP hop.
The hop limit for mesh addressing also  When a LoWPAN has been implemented
provides an upper limit on how many times using different link layer technologies, a
the frame can be forwarded. mesh-over routing setup is useful.
Each hop decrements this value by 1 as it is  While traditional IP routing protocols can be
forwarded. Once the value hits 0, it is used, a specialized routing protocol for smart
dropped and no longer forwarded. objects, such as RPL
 Source Address, and Destination Address:
The Source Address and Destination Address 6Lo Working Group:
fields for mesh addressing are IEEE 802.15.4  The 6Lo working group seeks to expand on
addresses indicating the endpoints of an IP this completed work with focus on IPv6
hop. connectivity over constrained node networks.
Mesh-Under Versus Mesh-Over Routing:  While the 6LoWPAN working group initially
 IEEE 802.15.4, IEEE 802.15.4g, and IEEE focused its optimizations on IEEE 802.15.4
1901.2a that support mesh topologies and LLNs,
operate at the physical and data link layers, two  6Lo working group is focused on the
main options exist for establishing reachability following:
and forwarding packets. IPv6-over-foo adaptation layer specifications
 “Mesh-under”: The routing of packets is using 6LoWPAN technologies (RFC4944,
handled at the 6LoWPAN adaptation layer. RFC6282, RFC6775) for link layer technologies:
 “Mesh-over” or “route-over”: utilizes IP  For example, this includes:
routing for getting packets to their destination.  IPv6 over Bluetooth Low Energy
 Mesh-under routing,  Transmission of IPv6 packets over near-field
 the routing of IP packets leverages the communication
6LoWPAN mesh addressing header to route and  IPv6 over 802.11ah
forward packets at the link layer.  Transmission of IPv6 packets over DECT Ultra
 The term mesh-under is used because multiple Low Energy
link layer hops can be used to complete a single  Transmission of IPv6 packets on WIA-PA
IP hop. (Wireless Networks for Industrial Automation–
 Nodes have a Layer 2 forwarding table that they Process Automation)
consult to route the packets to their final  Transmission of IPv6 over Master Slave/Token
destination within the mesh. Passing (MS/TP)
 An edge gateway terminates the mesh-under
domain.
Information and data models such as MIB
modules:
 One example is RFC 7388, “Definition of
Managed Objects for IPv6 over Low-Power
Wireless Personal Area Networks
(6LoWPANs).”
Optimizations that are applicable to more than
one adaptation layer specification:
 For example, this includes RFC 7400,
“6LoWPAN-GHC: Generic Header Fig. 5-7 Location of 6TiSCH’s 6top Sublayer
Compression for IPv6 over Low-Power
Wireless Personal Area Networks  Schedules in 6TiSCH are broken down into
(6LoWPANs).” cells.
 Informational and maintenance publications  A cell is simply a single element in the TSCH
needed for the IETF specifications in this area. schedule that can be allocated for
unidirectional or bidirectional communications
6TiSCH between specific nodes. •
 IEEE 802.15.4e, Time-Slotted Channel  Nodes only transmit when the schedule
Hopping (TSCH), is an add-on to the Media dictates that their cell is open for
Access Control (MAC) portion of the IEEE communication.
802.15.4 standard,  The 6TiSCH architecture defines four
 Devices implementing IEEE 802.15.4e TSCH schedule management mechanisms: • Static
communicate by following a Time Division scheduling
Multiple Access (TDMA) schedule.  Neighbor-to-neighbor scheduling
 An allocation of a unit of bandwidth or time  Remote monitoring and scheduling
slot is scheduled between neighbor nodes. management
 This allows the programming of predictable  Hop-by-hop scheduling
transmissions and enables deterministic,
industrial-type applications. Static scheduling:
 Not like other IEEE 802.15.4  All nodes in the constrained network share a
 To standardize IPv6 over the TSCH mode of fixed schedule.
IEEE 802.15.4e (known as 6TiSCH)  Cells are shared, and nodes contend for slot
 The IEEE 802.15.4e standard defines a time access in a slotted aloha manner.
slot structure, but it does not mandate a  Slotted aloha is a basic protocol for sending
scheduling algorithm for how the time slots are data using time slot boundaries when
utilized. communicating over a shared medium.
 This is left to higher-level protocols like  Static scheduling is a simple scheduling
6TiSCH. mechanism that can be used upon initial
 Scheduling is critical because it can affect implementation or as a fall back in the case of
throughput, latency, and power consumption. network malfunction.
 The drawback with static scheduling is that
nodes may expect a packet at any cell in the
schedule.
  Therefore, energy is wasted idly listening 1. Track Forwarding (TF):
across all cells.  This is the simplest and fastest forwarding
Neighbor-to-neighbor scheduling: model.
 A schedule is established that correlates with  A “track” in this model is a unidirectional
the observed number of transmissions path between a source and a destination.
between nodes.  This track is constructed by pairing bundles
 Cells in this schedule can be added or deleted of receive cells in a schedule with a bundle of
as traffic requirements and bandwidth needs receive cells set to transmit.
change.  So, a frame received within a particular cell
or cell bundle is switched to another cell or
Remote monitoring and scheduling cell bundle.
management:  This forwarding occurs regardless of the
 Time slots and other resource allocation are network layer protocol.
handled by a management entity that can be 2. IPv6 Forwarding (6F):
multiple hops away.  2This model forwards traffic based on its IPv6
 The scheduling mechanism leverages 6top routing table.
and even CoAP in some scenarios.  Flows of packets should be prioritized by
 This scheduling mechanism provides quite a traditional QoS (quality of service) and RED
bit of flexibility and control in allocating (random early detection) operations.
cells for communication between nodes.  QoS is a classification scheme for flows based
on their priority, and RED is a common
Hop-by-hop scheduling: congestion avoidance mechanism.
 A node reserves a path to a destination node 3. Fragment forwarding (FF):
multiple hops away by requesting the  This model takes advantage of 6LoWPAN
allocation of cells in a schedule at each fragmentation to build a Layer 2 forwarding
intermediate node hop in the path. table.
 The protocol that is used by a node to  Fragmentation within the 6LoWPAN protocol.
trigger this scheduling mechanism is not  IPv6 packets can get fragmented at the
defined at this point. 6LoWPAN sublayer to handle the differences
 In addition to schedule management between IEEE 802.15.4 payload size and IPv6
functions, the 6TiSCH architecture also MTU.
defines three different forwarding models.  Additional headers for RPL source route
 Forwarding is the operation performed on information can further contribute to the need
each packet by a node that allows it to be for fragmentation.
delivered to a next hop or an upperlayer  However, with FF, a mechanism is defined
protocol. where the first fragment is routed based on the
 The forwarding decision is based on a pre IPv6 header present.
existing state that was learned from a routing  This increases latency and can be power- and
computation. CPU-intensive for a constrained node.
 There are three 6TiSCH forwarding models: RPL
1. Track Forwarding (TF)  IETF chartered the RoLL (Routing over Low-
2. Fragment forwarding (FF) Power and Lossy Networks) working group to
3. IPv6 Forwarding (6F) evaluate all Layer 3 IP routing protocols and
determine the needs and requirements for
 developing a routing solution for IP smart
objects.
 The new routing protocol should be developed
for use by IP smart objects is IPv6 Routing
Protocol for Low Power and Lossy Networks
(RPL).
 In an RPL network,
 each node acts as a router and becomes part
of a mesh network.
 Routing is performed at the IP layer.
Fig. Example of a Directed Acyclic Graph (DAG)
 Each node examines every received IPv6
packet and determines the nexthop
 A basic RPL process involves building a
destination based on the information
contained in the IPv6 header. destination-oriented directed acyclic graph
(DODAG).
 The constraints of computing and memory that
 A DODAG is a DAG rooted to one destination.
are common characteristics of constrained
nodes, the protocol defines two modes:  In RPL, this destination occurs at a border router
Storing mode: known as the DODAG root.
 All nodes contain the full routing table of the  Observe that that a DAG has multiple roots,
RPL domain. Every node knows how to directly whereas the DODAG has just one.
reach every other node.
Non-storing mode:
 Only the border router(s) of the RPL domain
contain(s) the full routing table.
 All other nodes in the domain only maintain
their list of parents and use this as a list of
default routes toward the border router.
 This abbreviated routing table saves memory
space and CPU.
 When communicating in non-storing mode, a
node always forwards its packets to the border Fig.DAG and DODAG Comparison
router, which knows how to ultimately reach the
final destination.  In a DODAG, each node maintains up to three
 RPL is based on the concept of a directed parents that provide a path to the root.
acyclic graph (DAG). A DAG is a directed  one of these parents is the preferred parent,
graph where no cycles exist. which means it is the preferred next hop for
 This means that from any vertex or point in the upward routes toward the root.
graph, you cannot follow an edge or a line back  The routing graph created by the set of DODAG
to this same point. parents across all nodes defines the full set of
 All of the edges are arranged in paths oriented upward routes.
toward and terminating at one or more root  RPL protocol implementation should ensure that
nodes. routes are loop free by disallowing nodes from
selected DODAG parents that are positioned
further away from the border router.
 Upward routes in RPL are discovered and  For example, nodes implementing an OF with
configured using DAG Information Object Minimum Expected Number of Transmissions
(DIO) messages. (METX) advertise the METX among their
 Nodes listen to DIOs to handle changes in the parents in DIO messages.
topology that can affect routing.  Whenever a node establishes its rank, it simply
 Nodes establish downward routes by advertising sets the rank to the current minimum METX
their parent set toward the DODAG root using a among its parents.
Destination Advertisement Object (DAO) Rank
message.  The rank is a rough approximation of how
 DAO messages allow nodes to inform their “close” a node is to the root and helps avoid
parents of their presence and reachability to routing loops and the count-to-infinity problem.
descendants.  Nodes can only increase their rank when
 In non-storing mode of RPL, receiving a DIO message with a larger version
nodes sending DAO messages report their parent number.
sets directly to the DODAG root (border router),  However, nodes may decrease their rank
and only the root stores the routing information. whenever they have established lower-cost
 In storing mode, routes.
each node keeps track of the routing  While the rank and routing metrics are closely
information that is advertised in the DAO related, the rank differs from routing metrics in
messages. that it is used as a constraint to prevent routing
 The nodes can make their own routing loops.
decisions; in non-storing mode, on the other
hand, all packets must go up to the root to get a RPL Headers:
route for moving downstream.  An IPv6 Routing Header for Source Routes
 RPL messages, such as DIO and DAO, run on with the Routing Protocol for Low-Power and
top of IPv6. Lossy Networks (RPL).
 A new IPv6 option, known as the RPL option
 The RPL option is carried in the IPv6 Hop-
by-Hop header.
 The purpose of this header is to leverage data
plane packets for loop detection in a RPL
instance.
 Source Routing Header (SRH) for use
between RPL routers.
 A border router or DODAG root inserts the
SRH when specifying a source route to
deliver datagrams to nodes downstream in the
mesh network.
Metrics
Fig. RPL Overview  Some of the RPL routing metrics and
Objective Function (OF) : constraints defined in RFC 6551 include the
 An objective function (OF) defines how metrics following
are used to select routes and establish a node’s  Expected Transmission Count (ETX):
rank.
Assigns a discrete value to the number of ACE:
transmissions a node expects to make to deliver a  The Authentication and Authorization for
packet. Constrained Environments (ACE) working
 Hop Count: group is tasked with evaluating the applicability
Tracks the number of nodes traversed in a path. of existing authentication and authorization
Typically, a path with a lower hop count is chosen protocols and documenting their suitability for
over a path with a higher hop count. certain constrained-environment use cases.
 Latency:  ACE working group will focus its work on
Varies depending on power conservation. Paths with CoAP with the Datagram Transport Layer
a lower latency are preferred. Security (DTLS) protocol.
 Link Quality Level:  The ACE working group expects to produce a
Measures the reliability of a link by taking into standardized solution for authentication and
account packet error rates caused by factors such as authorization that enables authorized access
signal attenuation and interference. (Get, Put, Post, Delete) to resources identified
 Link Color: by a URI and hosted on a resource server in
Allows manual influence of routing by constrained environments.
administratively setting  An unconstrained authorization server performs
values to make a link more or less desirable. These mediation of the access.
values can be either statically or dynamically DICE:
adjusted for specific traffic types.  New generations of constrained nodes
 Node State and Attribute: implementing an IP stack over constrained
Identifies nodes that function as traffic aggregators access networks are expected to run an
and nodes that are being impacted by high optimized IP protocol stack.
workloads.  The DTLS in Constrained Environments
High workloads could be indicative of nodes that (DICE) working group focuses on
have incurred high CPU or low memory states. implementing the DTLS transport layer security
Naturally, nodes that are aggregators are preferred protocol in these environments.
over nodes experiencing high workloads.  The first task of the DICE working group is to
 Node Energy: define an optimized DTLS profile for
Avoids nodes with low power, so a battery-powered constrained nodes.
node that is running out of energy can be avoided  In addition, the DICE working group is
and the life of that node and the network can be considering the applicability of the DTLS
prolonged. record layer to secure multicast messages and
 Throughput: investigating how the DTLS handshake in
Provides the amount of throughput for a node link. constrained environments can get optimized.
Often, nodes conserving power use lower
throughput. 3. Explain SCADA protocol in detail ?
This metric allows the prioritization of paths with SCADA
higher throughput.  Supervisory control and data acquisition
Authentication and Encryption on Constrained (SCADA).
Nodes.  Designed decades ago, SCADA is an
 The IETF working groups that are focused on automation control system that was initially
IoT security: implemented without IP over serial links (such
 ACE and DICE.
 as RS-232 and RS-485), before being adapted  DLMS User Association specified a
to Ethernet and IPv4. communication profile based on TCP/IP in
A Little Background on SCADA the DLMS/COSEM.
 SCADA networking protocols, running directly  Serial protocols have adapted and evolved to
over serial physical and data link layers. utilize IP and TCP/UDP as both networking and
 At a high level, SCADA systems collect sensor transport mechanisms.
data and telemetry from remote devices, and to  DNP3 (Distributed Network Protocol) is based
control them. on a master/slave relationship.
 SCADA systems allow global, real-time, data-  The term master is refers to a powerful
driven decisions to be made about how to computer located in the control center of a
improve business processes. utility, and a slave is a remote device with
 SCADA commonly uses certain protocols for computing resources found in a location such as
communications between devices and a substation.
applications.  DNP3 refers to slaves as outstations.
 E.g.Modbus industrial protocols used to  Outstations monitor and collect data from
monitor and program remote devices via a devices that indicate their state, such as whether
master/slave relationship. a circuit breaker is on or off, and take
 Modbus used in building management, measurements, including voltage, current,
transportation, and energy applications. temperature, and so on.
 The DNP3 (Distributed Network Protocol) and  This data is then transmitted to the master when
International Electrotechnical Commission it is requested, or events or alarms and control
(IEC) protocols are found mainly in the utilities commands can be sent in an asynchronous
industry, along with DLMS/COSEM manner.
 ANSI C12 for advanced meter reading (AMR).
 These protocols used decade ago and are serial
based. So, transporting them over current IoT
and traditional networks requires that certain
Adjustments.
Adapting SCADA for IP 92
 Ethernet and IP include the ability to leverage
existing equipment and standards while
integrating seamlessly the SCADA subnetworks
to the corporate WAN infrastructures.
 Assigning TCP/UDP to the protocols, as
following:
 DNP3 (adopted by IEEE 1815-2012)
specifies the use of TCP or UDP on port
20000 for transporting DNP3 messages over
IP.
 The Modbus messaging service utilizes TCP.
 IEC 60870-5-104 is the evolution of IEC
60870-5-101 serial for running over Ethernet Figure 3.12 Protocol Stack for Transporting Serial
and IPv4 using port 2404. DNP3 SCADA over IP
 Connection management links the DNP3 layers
with the IP layers in addition to the configuration
parameters and methods necessary for
implementing the network connection.
 The master side initiates connections by
performing a TCP active open.
 The outstation listens for a connection request
by performing a TCP passive open.
 Master stations may parse multiple DNP3 data
link layer frames from a single UDP datagram,
while DNP3 data link layer frames cannot span
 Scenarios A:
multiple UDP datagrams.
 Both the SCADA server and the RTUs have a
Tunneling Legacy SCADA over IP Networks
direct serial connection to their respective
 End-to-end native IP support is preferred, in the
routers.
case of DNP3.
 The routers terminate the serial connections at
 Otherwise, transport of the original serial
both ends of the link and use raw socket
protocol over IP can be achieved either by
encapsulation to transport the serial payload over
tunneling using raw sockets over TCP or UDP or
the IP network.
by installing an intermediate device that
performs protocol translation between the serial
protocol version and its IP implementation.
 A raw socket connection simply denotes that the
serial data is being packaged directly into a TCP
or UDP transport.
 A socket is a standard application programming
interface (API) composed of an IP address and a
TCP or UDP port that is used to access network
devices over an IP network.
 Scenarios A, B and C in Figure,
 Routers connect via serial interfaces to the  Scenarios B:
remote terminal units (RTUs), which are often  A small change on the SCADA server side. A
associated with SCADA networks. piece of software is installed on the SCADA
 An RTU is a multipurpose device used to server that maps the serial COM ports to IP
monitor and control various systems, ports.
applications, and devices managing automation.  This software is commonly referred to as an
 From the master/slave perspective, the RTUs are IP/serial redirector.
the slaves.  The IP/serial redirector in essence terminates the
 Opposite the RTUs in is a SCADA server, or serial connection of the SCADA server and
master, that varies its connection type converts it to a TCP/IP port using a raw socket
connection.
 connection is present on one side and an IP
connection is used on the other.
 By running protocol translation, the IoT
gateway connected to the RTUs is implementing
a computing function close to the edge of the
network.
 Adding computing functions close to the edge
helps scale distributed intelligence in IoT
networks.
SCADA Transport over LLNs with MAP-T
 Due to the constrained nature of LLNs, the
implementation of industrial protocols should at
Scenarios C: a minimum be done over UDP.
 The SCADA server supports native raw socket  This in turn requires that both the application
capability. servers and devices support and implement UDP.
 Unlike in Scenarios A and B, where a router or  When deployed over LLN subnetworks that are
IP/serial redirector software has to map the IPv6 only, a transition mechanism, such as
SCADA server’s serial ports to IP ports, in MAP-T (Mapping of Address and Port using
Scenario C the SCADA server has full IP Translation), needs to be implemented.
support for raw socket connections.
SCADA Protocol Translation

Fig. DNP3 Protocol over 6LoWPAN Networks with MAP-T

 In figure shows a scenario in which a legacy
 With protocol translation, the legacy serial endpoint is connected across an LLN running
protocol is translated to a corresponding IP 6LoWPAN to an IP-capable SCADA server.
version.Figure shows two serially connected  The legacy endpoint could be running various
DNP3 RTUs and two master applications industrial and SCADA protocols, including
supporting DNP3 over IP that control and pull DNP3/IP, Modbus/TCP, or IEC.
data from the RTUs.  In this scenario, the legacy devices and the
 The IoT gateway in this figure performs a SCADA server support only IPv4.
protocol translation function that enables  MAP-T makes the appropriate mappings
communication between the RTUs and between IPv4 and the IPv6 protocols.
servers, despite the fact that a serial  This allows legacy IPv4 traffic to be forwarded
across IPv6 networks.
  In Figure the IPv4 endpoint on the left side is certifications for IoT constrained nodes and
connected to a Customer Premise Equipment networks.
(CPE) device.  Internet Protocol for Smart Objects (IPSO)
 The MAP-T CPE device has an IPv6 connection Alliance
to the RPL mesh.  Wi-SUN Alliance
 On the right side, a SCADA server with native  Thread
IPv4 support connects to a MAP-T border  IPv6 Ready Logo
gateway
Internet Protocol for Smart Objects (IPSO)
4. Explain the difference between CoAP and Alliance
MQTT protocols ?  The alliance initially focused on promoting IP
CoAP MQTT as the premier solution for smart objects
COAP stands for MQTT stands for Message communications.
Constrained Application query telemetry transport.  Today, it is more focused on how to use IP,
Protocol. with the IPSO Alliance organizing
For communication, it For communication, it uses interoperability tests between alliance members
uses a request-response the publish-subscribe to validate that IP for smart objects can work
prototype prototype together and properly implement industry
It uses asynchronous and It uses only asynchronous standards.
synchronous messaging. mode for messaging. Wi-SUN Alliance
 Wi-SUN’s main focus is on the IEEE
It uses User Datagram It uses the Transmission 802.15.4g protocol and its support for
Protocol (UDP) Control Protocol (TCP). multiservice and secure IPv6 communications
The Heather size of The Heather size of MQTT with applications running over the UDP
CAOP is 4 bytes is 2 bytes transport layer.
It is RESTful based It is not RESTful based  The utilities industry is the main area of focus
It does not have It is mainly used for live for the Wi-SUN Alliance.
persistence support communication and has  The Wi-SUN field area network (FAN) profile
persistence support enables smart utility networks to provide
It will give labels to the It does not have any such resilient, secure, and cost-effective
messages. function. connectivity with extremely good coverage in a
It has a secured system, It is very secure, and its range of topographic environments, from dense
and its usability is in usability is in IoT urban neighborhoods to rural areas.
Utility area networks applications. Thread
It has Low Latency and It has Low Latency and  Thread Group has defined an IPv6-based
NAT issues. NAT issues wireless profile that provides the best way to
connect more than 250 devices into a low-
5. Explain profiles and compliances for IP ? power, wireless mesh network.
 Profile definitions, certifications, and IPv6 Ready Logo
promotion by alliances can help implementers  The IPv6 Ready Logo program has established
develop solutions that guarantee conformance and interoperability testing
interoperability and/or interchangeability of programs with the intent of increasing user
devices. confidence when implementing IPv6.
 Some of the main industry organizations
working on profile definitions and
  The IPv6 Core and specific IPv6 components, Coap message Field:
such as DHCP, IPsec, and customer edge CoAP Description
router certifications, are in place. message
header
6. Explain CoAP protocol and MQTT in detail. Ver It is 2 bit unsigned integer. It mentions
CoAP CoAP version number. Set to one.
 Constrained Application Protocol (CoAP) T It is 2 bit unsigned integer. Indicates
 is to develop a generic framework for message type viz. confirmable (0), non-
resource-oriented applications targeting confirmable (1), ACK (2) or RESET(3).
constrained nodes and networks. TKL It is 4 bit unsigned integer, Indicates
 The CoAP framework defines simple and length of token (0 to 8 bytes)
flexible ways to manipulate sensors and Code It is 8 bit unsigned integer, It is split into
actuators for data or device management. two parts viz. 3 bit class (MSBs) and 5
 The CoAP messaging model is primarily bit detail (LSBs).
designed to facilitate the exchange of
messages over UDP between endpoints, Message 16 bit unsigned integer. Used for
including the secure transport protocol ID matching responses. Used to detect
Datagram Transport Layer Security (DTLS). message duplication.

 CoAP can run over IPv4 or IPv6. However, it is

recommended that the message fit within a
single IP packet and UDP payload to avoid
fragmentation.

 CoAP over Short Message Service (SMS) as

defined in Open Mobile Alliance for
Lightweight Machine-to-Machine (LWM2M)
for IoT device management.
 CoAP communications across an IoT
 A CoAP message is composed of
infrastructure can take various paths.
 a short fixed-length Header field (4 bytes),
 Connections can be between devices located on
 A variable-length but mandatory Token
the same or different constrained networks or
field (0–8 bytes),
between devices and generic Internet or cloud
 Otional field if necessary, and the payload
servers, all operating over IP.
field.
 As both HTTP and CoAP are IP-based
protocols, the proxy function can be located
practically anywhere in the network, not
 necessarily at the border between constrained  The MQTT server (or message broker) accepts
and non-constrained networks. the network connection along with application
 Just like HTTP, CoAP is based on the REST messages, such as Temp/RH data, from the
architecture, but with a “thing” acting as both publishers.
the client and the server.  It also handles the subscription and
 Through the exchange of asynchronous unsubscription process and pushes the
messages, a client requests an action via a application data to MQTT clients acting as
method code on a server resource. subscribers.
 A uniform resource identifier (URI) localized  The application on the right side of Figure is an
on the server identifies this resource. MQTT client that is a subscriber to the Temp/RH
 The server responds with a response code that data being generated by the
may include a resource representation.  publisher or sensor on the left.
 The CoAP request/response semantics include  This model, where subscribers express a desire
the methods GET, POST, PUT, and DELETE. to receive information from publishers, is well
MQTT known.
 Message Queuing Telemetry Transport (MQTT)  The presence of a message broker in MQTT
 Considering the harsh environments in the oil decouples the data transmission between clients
and gas industries, an extremely simple protocol acting as publishers and subscribers.
with only a few options was designed, with  In fact, publishers and subscribers do not even
considerations for constrained nodes, unreliable know (or need to know) about each other.
WAN backhaul communications, and bandwidth  A benefit of having this decoupling is that the
constraints with variable latencies. MQTT message broker ensures that information
 These were some of the rationales for the can be buffered and cached in case of network
selection of a client/server and publish/subscribe failures.
framework based on the TCP/IP architecture  Compared to the CoAP message, MQTT
contains a smaller header of 2 bytes compared to
4 bytes for CoAP.

 An MQTT client can

 act as a publisher to send data (or resource
information) to an MQTT server acting as an
MQTT message broker.
 In Figure the MQTT client on the left side is a  MQTT is a lightweight protocol because each
temperature (Temp) and relative humidity (RH) control packet consists of a 2-byte fixed header
sensor that publishes its Temp/RH data. with optional variable header fields and optional
payload.
 The first MQTT field in the header is Message UNIT 4: DATA AND ANALYTICS
Type, which identifies the kind of MQTT packet
within a message.
FOR IOT
 Fourteen different types of control packets are 1. IoT Data Analytics Overview
specified in MQTT.  The true importance of IoT data from smart
objects is realized only when the analysis of the
data leads to actionable business intelligence and
insights.
 Data analysis is typically broken down by the
types of results that are produced

Four types of data analysis results :

 Descriptive:
 Descriptive data analysis tells you what is
happening, either now or in the past.
 For example, a thermometer in a truck engine
reports temperature values every second.
 From a descriptive analysis perspective, you
can pull this data at any moment to gain insight
into the current operating condition of the truck
engine.
 If the temperature value is too high, then there
may be a cooling problem or the engine may
be experiencing too much load.
 Diagnostic:
 When you are interested in the “why,”
diagnostic data analysis can provide the
answer.
  Continuing with the example of the  Both predictive and prescriptive analyses are
temperature sensor in the truck engine, you more resource intensive and increase
might wonder why the truck engine failed. complexity, but the value they provide is much
 Diagnostic analysis might show that the greater than the value from descriptive and
temperature of the engine was too high, and diagnostic analysis
the engine overheated.  You can see that descriptive analysis is the least
 Applying diagnostic analysis across the data complex and at the same time offers the least
generated by a wide range of smart objects value.
can provide a clear picture of why a problem  On the other end, prescriptive analysis provides
or an event occurred. the most value but is the most complex to
implement.
 Predictive:
 Predictive analysis aims to foretell problems
or issues before they occur.
 For example, with historical values of
temperatures for the truck engine, predictive
analysis could provide an estimate on the
remaining life of certain components in the
engine.
 These components could then be proactively
replaced before failure occurs.
 Or perhaps if temperature values of the truck
engine start to rise slowly over time, this
could indicate the need for an oil change or
some other sort of engine cooling
maintenance. IoT Data Analytics Challenges:

 Prescriptive:  Scaling problems:

 Prescriptive analysis goes a step beyond  Due to the large number of smart objects in
predictive and recommends solutions for most IoT networks that continually send data,
upcoming problems. relational databases can grow incredibly large
 A prescriptive analysis of the temperature very quickly.
data from a truck engine might calculate  This can result in performance issues that can
various alternatives to cost-effectively be costly to resolve, often requiring more
maintain our truck hardware and architecture changes.
 These calculations could range from the cost
necessary for more frequent oil changes and  Volatility of data:
cooling maintenance to installing new cooling  With relational databases, it is critical that the
equipment on the engine or upgrading to a schema be designed correctly from the
lease on a model with a more powerful beginning.
engine.  Changing it later can slow or stop the database
 Prescriptive analysis looks at a variety of from operating.
factors and makes the appropriate  Due to the lack of flexibility, revisions to the
recommendation schema must be kept at a minimum.
  IoT data, however, is volatile in the sense that  on a single node.
the data model is likely to change and evolve  It is common to see clusters of servers that
over time. A dynamic schema is often required consist of dozens, hundreds, or even thousands
so that data model changes can be made daily of nodes for some large deployments.
or even hourly.  The characteristics of big data can be defined by
the sources and types of data.
2. Big Data Analytics Tools and Technology ?  First is machine data, which is generated by
 Big data analytics can consist of many different IoT devices and is typically unstructured data.
software pieces that together collect, store,  Second is transactional data, which is from
manipulate, and analyze all different data types. sources that produce data from transactions on
Generally, the industry looks to the “three Vs” to these systems, and, have high volume and
categorize big data: structured.
 Third is social data sources, which are
Velocity:
typically high volume and structured.
 Velocity refers to how quickly data is being
 Fourth is enterprise data, which is data that is
collected and analyzed. Hadoop Distributed File
lower in volume and very structured
System is designed to ingest and process data
 Hence big data consists of data from all these
very quickly.
separate sources.
 Smart objects can generate machine and sensor
data at a very fast rate and require database or
3. Explain about MPP Shared-Nothing
file systems capable of equally fast ingest
Architecture?
functions.
 Enterprises have used relational databases for
storing structured, row and column style data
 Variety:
types for decades.
 Variety refers to different types of data. Often
 Relational databases are often grouped into a
you see data categorized as structured, semi-
broad data storage category called data
structured, or unstructured.
warehouses.
 Different database technologies may only be
 Though they are the centerpiece of most data
capable of accepting one of these types.
architectures, they are often used for longer-term
 Hadoop is able to collect and store all three
archiving and data queries that can often take
types.
minutes or hours
 This can be beneficial when combining
 Massively parallel processing (MPP) databases
machine data from IoT devices that is very
were built on the concept of the relational data
structured in nature with data from other
warehouses but are designed to be much faster,
sources, such as social media or multimedia
to be efficient, and to support reduced query
that is unstructured.
times .
 To accomplish this, MPP databases take
 Volume:
advantage of multiple nodes (computers)
 Volume refers to the scale of the data.
designed in a scale-out architecture such that
 Typically, this is measured from gigabytes on the
both data and processing are distributed across
very low end to petabytes or even exabytes of
multiple systems .
data on the other extreme.
 MPPs are sometimes referred to as analytic
 Generally, big data implementations scale databases because they are designed to allow for
beyond what is available on locally attached fast query processing and often have built-in
storage disks analytic functions
 As the name implies, these database types analyzing data on the fly—and the data may
process massive data sets in parallel across many come from many different sources, including
processors and nodes trackside sensors, car telemetry, and weather
 An MPP architecture typically contains a single reports.
master node that is responsible for the Comparing Big Data and Edge Analytics
coordination of all the data storage and
processing across the cluster.  From a business perspective, streaming analytics
 It operates in a “shared-nothing” fashion, with involves acting on data that is generated while it
each node containing local processing, memory, is still valuable, before it becomes stale.
and storage and operating independently.  For example, roadway sensors combined with
GPS way finding apps may tell a driver to avoid
a certain highway due to traffic.
 This data is valuable for only a small window of
time.
 Historically, it may be interesting to see how
many traffic accidents or blockages have
occurred on a certain segment of highway or to
predict congestion based on past traffic data.
 However, for the driver in traffic receiving this
information, if the data is not acted upon
immediately, the data has little value.
 From a security perspective, having
instantaneous access to analyzed and
preprocessed data at the edge also allows an
4. Write a short note on edge streaming organization to realize anomalies in its network
analytics. so those anomalies can be quickly contained
 One industry where data analytics is used before spreading to the rest of the network.
extensively is the world of automobile racing.  To summarize, the key values of edge streaming
 For example, in Formula One racing, each car analytics include the following:
has between 150 to 200 sensors that, combined, Reducing data at the edge:
generate more than 1000 data points per second,  The aggregate data generated by IoT devices is
resulting in hundreds of gigabytes of raw data generally in proportion to the number of
per race. devices. The scale of these devices is likely to
 The sensor data is transmitted from the car and be huge, and so is the quantity of data they
picked up by track-side wireless sensors. During generate.
a race, weather conditions may vary, tire  Passing all this data to the cloud is inefficient
conditions change, and accidents or other racing and is unnecessarily expensive in terms of
incidents almost always require an adaptable and bandwidth and network infrastructure.
flexible racing strategy. Analysis and response at the edge:
 As the race develops, decisions such as when to  Some data is useful only at the edge (such as
pit, what tires to use, when to pass, and when to a factory control feedback system).
slow down all need to be made in seconds.  In cases such as this, the data is best
analyzed and acted upon where it is
 Teams have found that enormous insights
generated.
leading to better race results can be gained by
Time sensitivity: 5. Explain OCTAVE AND FAIR in detail.
 When timely response to data is required,  OCTAVE (Operationally Critical Threat, Asset
passing data to the cloud for future and Vulnerability Evaluation) has undergone
processing results in unacceptable latency.  multiple iterations. The version this section
 Edge analytics allows immediate responses focuses on is OCTAVE Allegro, which is
to changing conditions. intended to
Edge Analytics Core Functions  be a lightweight and less burdensome process to
 To perform analytics at the edge, data needs to implement. Allegro assumes that a robust
be viewed as real-time flows. security
 Whereas big data analytics is focused on large  team is not on standby or immediately at the
quantities of data at rest, edge analytics ready to initiate a comprehensive security
continually processes streaming flows of data in review.
motion. Streaming analytics at the edge can be  This approach and the assumptions it makes are
broken down into three simple stages: quite appropriate, given that many operational
 Raw input data:  technology areas are similarly lacking in
 This is the raw data coming from the sensors security-focused human assets. illustrates the
into the analytics processing unit.  OCTAVE Allegro steps and phases.
 Analytics processing unit (APU):
 The APU filters and combines data streams (or
separates the streams, as necessary), organizes
them by time windows, and performs various
analytical functions.
 It is at this point that the results may be acted
on by micro services running in the APU.
 Output streams:
 The data that is output is organized into
insightful streams and is used to influence the
behavior of smart objects, and passed on for
storage and further processing in the cloud.
 Communication with the cloud often happens
through a standard publisher/subscriber
messaging protocol, such as MQTT.

 OCTAVE is a balanced information-focused

process.
 What it offers in terms of discipline and largely
unconstrained breadth, however, is offset by its
lack of security specificity.
 There is an assumption that beyond these steps
are seemingly means of identifying specific
mitigations that can be mapped to the threats and
risks exposed during the analysis process.
 websites and quickly return search results for
FAIR : open source search engines.
 FAIR (Factor Analysis of Information Risk) is a  Initially, the project had two key elements:
technical standard for risk definition from The
Open Group.  Hadoop Distributed File System (HDFS):
 While information security is the focus, much as A system for storing data across multiple nodes
it is for OCTAVE, FAIR has clear applications
within operational technology. MapReduce:
 Like OCTAVE, it also allows for non-malicious A distributed processing engine that splits a l
actors as a potential cause for harm, but it goes to arge task into smaller ones that can be run in
greater lengths to emphasize the point. For many parallel
operational groups, it is a welcome  Much like the MPP and NoSQL systems
acknowledgement of existing contingency discussed earlier, Hadoop relies on a scale-out
planning. architecture that leverages local processing,
 Unlike with OCTAVE, there is a significant memory, and storage to distribute tasks and
emphasis on naming, with risk taxonomy provide a scalable storage system for data.
definition as a very specific target.  Both MapReduce and HDFS take advantage of
 FAIR places emphasis on both unambiguous this distributed architecture to store and process
definitions and the idea that risk and associated massive amounts of data and are thus able to
attributes are measurable. leverage resources from all nodes in the cluster.
 Measurable, quantifiable metrics are a key area  For HDFS, this capability is handled by
of emphasis, which should lend itself well to an specialized nodes in the cluster, including
operational world with a richness of operational NameNodes and DataNodes :
data.
 At its base, FAIR has a definition of risk as the NameNodes:
probable frequency and probable magnitude of  These are a critical piece in data adds, moves,
loss. deletes, and reads on HDFS.
 With this definition, a clear hierarchy of sub-  They coordinate where the data is stored, and
elements emerges, with one side of the taxonomy maintain a map of where each block of data is
focused on frequency and the other on stored and where it is replicated.
magnitude.  All interaction with HDFS is coordinated
 Loss even frequency is the result of a threat through the primary (active) NameNode, with a
agent acting on an asset with a resulting loss to secondary (standby) NameNode notified of the
the organization. changes in the event of a failure of the primary.
 The NameNode takes write requests from
4. Explain the Hadoop big data tool ? clients and distributes those files across the
available nodes in configurable block sizes,
Hadoop usually 64 MB or 128 MB blocks.
 Hadoop is the most recent entrant into the data
 The NameNode is also responsible for
management market, but it is arguably the most instructing the DataNodes where replication
popular choice as a data repository and should occur.
processing engine.
 Hadoop was originally developed as a result of
DataNodes
projects at Google and Yahoo!, and the original  These are the servers where the data is stored at
intent for Hadoop was to index millions of the direction of the NameNode.
 It is common to have many DataNodes in a MapReduce to be responsible only for data
Hadoop cluster to store the data. processing.
 Data blocks are distributed across several nodes
and often are replicated three, four, or more The Hadoop Ecosystem
times across nodes for redundancy.  Since the initial release of Hadoop in 2011,
 Once data is written to one of the DataNodes, the many projects have been developed to add
DataNode selects two (or more) additional incremental functionality to Hadoop and have
nodes, based on replication policies, to ensure collectively become known as the Hadoop
data redundancy across the cluster. ecosystem.
 MapReduce leverages a similar model to batch  Apache Kafka
process the data stored on the cluster nodes.  Apache Spark
 Batch processing is the process of running a  Apache Storm and Apache Flink
scheduled or ad hoc query across historical data  Lambda Architecture
stored in
 the HDFS. A query is broken down into smaller Apache Kafka
tasks and distributed across all the nodes running  Part of processing real-time events, such as those
MapReduce in a cluster. commonly generated by smart objects, is having
 While this is useful for understanding patterns them ingested into a processing engine.
and trending in historical sensor or machine data,  The process of collecting data from a sensor or
it has one significant drawback: time log file and preparing it to be processed and
analyzed is typically handled by messaging
systems.
 Messaging systems are designed to accept data,
or messages, from where the data is generated
and deliver the data to stream-processing engines
such as Spark Streaming or Storm.
 Apache Kafka is a distributed publisher-
subscriber messaging system that is built to be
scalable and fast.
 It is composed of topics, or message brokers,
where producers write data and consumers read
data from these topics.
 the data flow from the smart objects (producers),
through a topic in Kafka, to the real-time
YARN processing engine.
 Introduced with version 2.0 of Hadoop, YARN  Due to the distributed nature of Kafka, it can run
(Yet Another Resource Negotiator) was designed in a clustered configuration that can handle many
to enhance the functionality of MapReduce. producers and consumers simultaneously and
 With the initial release, MapReduce was exchanges information between nodes, allowing
responsible for batch data processing and job topics to be distributed over multiple nodes.
tracking and resource management across the  The goal of Kafka is to provide a simple way to
cluster. connect to data sources and allow consumers to
 YARN was developed to take over the resource connect to that data in the way they would like.
negotiation and job/task tracking, allowing
 combination of the Hadoop ecosystem projects
Apache Spark discussed.
 Apache Spark is an in-memory distributed data  One architecture that is currently being leveraged
analytics platform designed to accelerate for this functionality is the Lambda Architecture.
processes in the Hadoop ecosystem.  Lambda is a data management system that
 The “in-memory” characteristic of Spark is what consists of two layers for ingesting data (Batch
enables it to run jobs very quickly. and Stream) and one layer for providing the
 At each stage of a MapReduce operation, the combined data (Serving).
data is read and written back to the disk, which  These layers allow for the packages discussed
means latency is introduced through each disk previously, like Spark and MapReduce, to
operation. operate on the data independently, focusing on
 However, with Spark, the processing of this data the key attributes for which they are designed
is moved into high-speed memory, which has and optimized. Data is taken from a message
significantly lower latency. broker, commonly Kafka, and processed by each
 This speeds the batch processing jobs and also layer in parallel, and the resulting data is
allows for near-real-time processing of events. delivered to a data store where additional
processing or queries can be run.
Apache Storm and Apache Flink  This parallel data flow through the Lambda
 As you work with the Hadoop ecosystem, you Architecture.
will inevitably notice that different projects are  The Lambda Architecture is not limited to the
very similar and often have significant overlap packages in the Hadoop ecosystem, but due to its
with other projects. breadth and flexibility, many of the packages in
 This is the case with data streaming capabilities. the ecosystem fill the requirements of each layer
 For example, Apache Spark is often used for nicely:
both distributed streaming analytics and batch
processing.
 Apache Storm and Apache Flink are other
Hadoop ecosystem projects designed for
distributed stream processing and are commonly
deployed for IoT use cases.
 Storm can pull data from Kafka and process it in
a near-real-time fashion, and so can Apache
Flink.
 This space is rapidly evolving, and projects will
continue to gain and lose popularity as they
evolve.
Stream layer:
Lambda Architecture
 This layer is responsible for near-real-time
 Ultimately the key elements of a data processing of events.
infrastructure to support many IoT use cases  Technologies such as Spark Streaming,
involves the collection, processing, and storage Storm, or Flink are used to quickly ingest,
of data using multiple technologies. process, and analyze data on this layer.
 Querying both data in motion (streaming) and  Alerting and automated actions can be
data at rest (batch processing) requires a triggered on events that require rapid response
 or could result in catastrophic outcomes if not operational systems may be deemed legacy
handled immediately. systems.
Batch layer:  For example, in a power utility environment, it
 The Batch layer consists of a batch-processing is not uncommon to have racks of old
engine and data store. mechanical equipment still operating alongside
 If an organization is using other parts of the modern intelligent electronic devices (IEDs).
Hadoop ecosystem for the other layers,  In many cases, legacy components are not
MapReduce and HDFS can easily fit the bill. restricted to isolated network segments but have
 Other database technologies, such as MPPs, now been consolidated into the IT operational
NoSQL, or data warehouses, can also provide environment.
what is needed by this layer.  From a security perspective, this is potentially
Serving layer: dangerous as many devices may have historical
 The Serving layer is a data store and mediator vulnerabilities or weaknesses that have not been
that decides which of the ingest layers to query patched and updated, or it may be that patches
based on the expected result or view into the are not even available due to the age of the
data. equipment.
 If an aggregate or historical view is requested, it
may invoke the Batch layer. If real-time Insecure Operational Protocols
analytics is needed, it may invoke the Stream  The structure and operation of most of these
layer. protocols is often publicly available.
 The Serving layer is often used by the data  While they may have been originated by a
consumers to access both layers simultaneously. private firm, for the sake of interoperability, they
are typically published for others to implement.
5. What are the challenges in IoT security?  Thus, it becomes a relatively simple matter to
 The security challenges faced in IoT are by no compromise the protocols themselves and
means new and are not limited to specific introduce malicious actors that may use them to
industrial environments. compromise control systems for either
 The following sections discuss some of the reconnaissance or attack purposes that could lead
common challenges faced in IoT. to undesirable impacts in normal system
operation.
Erosion of Network Architecture
 There is a wide variety in secured network
Device Insecurity
designs within and across different industries.
 Beyond the communications protocols that are
 For example, power utilities have a strong
used and the installation base of legacy systems,
history of leveraging modern technologies for
control and communication elements themselves
operational activities, and in North America
have a history of vulnerabilities.
there are regulatory requirements in place from
 To understand the nature of the device
regulatory authorities, such as North American
insecurity, it is important to review the history of
Electric Reliability Corporation’s (NERC’s)
what vulnerabilities were discovered and what
Critical Infrastructure Protection (CIP)
types of devices were affected.
Pervasive Legacy Systems  A review of the time period 2000 to 2010 reveals
that the bulk of discoveries were at the higher
 Due to the static nature and long lifecycles of
levels of the operational network, including
equipment in industrial environments, many
control systems trusted to operate plants,
 transmission systems, oil pipelines, or whatever several different types of databases, including
critical function is in use. the following:

Document stores:
6. Write a short note on network analytics.  This type of database stores semi-
 Another form of analytics that is extremely structured data, such as XML or JSON.
important in managing IoT systems is network-  Document stores generally have query
based analytics . engines and indexing features that allow for
 Network analytics is concerned with discovering many optimized queries.
patterns in the communication flows from a Key-value stores:
network traffic perspective.  This type of database stores associative arrays
 Network analytics has the power to analyze where a key is paired with an associated value.
details of communications patterns made by  These databases are easy to build and easy to
protocols and correlate this across the network. scale.
 It allows you to understand what should be Wide-column stores:
considered normal behavior in a network and to  This type of database stores similar to a key-
quickly identify anomalies that suggest network value store, but the formatting of the values can
problems due to suboptimal paths, intrusive vary from row to row, even in the same table.
malware, or excessive congestion. Graph stores:
 Another form of analytics that is extremely  This type of database is organized based on the
important in managing IoT systems is network- relationships between elements.
based analytics .  Graph stores are commonly used for social
 Network analytics is concerned with discovering media or natural language processing, where the
patterns in the communication flows from a connections between data are very relevant.
network traffic perspective.  NoSQL was developed to support the high-
velocity, urgent data requirements of modern
 Network analytics has the power to analyze
web applications that typically do not require
details of communications patterns made by
much repeated use.
protocols and correlate this across the network.
 The original intent was to quickly ingest rapidly
 It allows you to understand what should be
changing server logs and clickstream data
considered normal behavior in a network and to
generated by web-scale applications that did not
quickly identify anomalies that suggest network
neatly fit into the rows and columns required by
problems due to suboptimal paths, intrusive
relational databases.
malware, or excessive congestion.
 Similar to other data stores, like MPPs and
Hadoop (discussed later), NoSQL is built to
scale horizontally, allowing the database to span
7. What is NoSQL?
multiple hosts, and can even be distributed
 NoSQL (“not only SQL”) is a class of databases
geographically.
that support semi-structured and unstructured
 Expanding NoSQL databases to other nodes is
data, in addition to the structured data handled
similar to expansion in other distributed data
by data warehouses and MPPs.
systems, where additional hosts are managed by
a master node or process.
 NoSQL is not a specific database technology;
 This expansion can be automated by some
rather, it is an umbrella term that encompasses
NoSQL implementations or can be provisioned
manually.
 This level of flexibility makes NoSQL a good
candidate for holding machine and sensor data
associated with smart objects

8. What is Machine Learning?

 ML is central to IoT.
 Data collected by smart objects needs to be
analyzed, and intelligent actions need to be taken
based on these analyses.
 Performing this kind of operation manually is
almost impossible (or very, very slow and
inefficient).
 Machines are needed to process information fast
and react instantly when thresholds are met
 Ex: advances in self-driving vehicle--abnormal
pattrn recognition in a crowd and automated
intelligent and machine-assisted decision
systemMachine learning is, in fact, part of a
larger set of technologies commonly grouped
under the term artificial intelligence (AI).
 AI includes any technology that allows a
computing system to mimic human intelligence
using any technique, from very advanced logic to
basic “if-then-else” decision loops.
 Any computer that uses rules to make decisions
belong to this group.
 ML is concerned with any process where the
computer needs to receive a set of data that is
processed to help perform a task with more
efficiency.
 ML is a vast field but can be simply divided in
two main categories: supervised and
unsupervised learning
 9. Dstinguish between supervised learning and unsupervised learning?
Supervised learning
Parameters Unsupervised learning

In a supervised learning model, input and In unsupervised learning model, only input
Process
output variables will be given. data will be given

Algorithms are used against data which is not

Input Data Algorithms are trained using labeled data.
labeled
Support vector machine, Neural network, Unsupervised algorithms can be divided into
Algorithms
Linear and logistics regression, random forest, different categories: like Cluster algorithms,
Used
and Classification trees. K-means, Hierarchical clustering, etc.
Computational Unsupervised learning is computationally
Supervised learning is a simpler method.
Complexity complex
Supervised learning model uses training data
Unsupervised learning does not use output
Use of Data to learn a link between the input and the
data.
outputs.
Computational Unsupervised learning is computationally
Supervised learning is a simpler method.
Complexity complex
Supervised learning model uses training data
Unsupervised learning does not use output
Use of Data to learn a link between the input and the
data.
outputs.
Accuracy of
Highly accurate and trustworthy method. Less accurate and trustworthy method.
Results
Real Time
Learning method takes place offline. Learning method takes place in real time.
Learning
Accuracy of Highly accurate and trust worthy method. Less accurate and trustworthy method.
Results
Real Time Learning method takes place offline. Learning method takes place in real time.
Learning

10.Compare b/n structured and unstructured data ?

Properties Structured data Unstructured data

Technology It is based on Relational database table It is based on character and binary data
Transaction management Matured transaction and various No transaction management and no
concurrency techniques concurrency
Version management Versioning over tuples, row, tables Versioned as a whole
Flexibility It is schema dependent and less flexible It is more flexible and there is absence
of schema
Scalability It is very difficult to scale DB schema It is more scalable.
Robustness Very robust Less robust
Query performance Structured query allow complex joining Only textual queries are possible
Format It has a predefined format. It has a variety of formats, i.e., it comes
in a variety of shapes and sizes.

Analysis It is easy to search Searching for unstructured data is more

difficult.
UNIT 5: IoT Physical Devices and Endpoints

1. Explain the smart city IoT architecture.

 A smart city IoT infrastructure is a four-layered architecture,
 Data flows from devices at the street layer to the city network layer and connect to the
 data center layer, where the data is aggregated, normalized, and virtualized.
 The data center layer provides information to the services layer, which consists of the
 applications that provide services to the city

1) Street Layer
• The street layer is composed of devices and sensors that collect data and take action based on instructions
from the overall solution, as well as the networking components needed to aggregate and collect data.
• A sensor is a data source that generates data required to understand the physical world. Sensor devices are
able to detect and measure events in the physical world.
• ICT (information and communication technology ) connectivity solutions rely on sensors to collect the
data from the world around them so that it can be analyzed and used to operationalize use cases for cities.
• A variety of sensors are used at the street layer for a variety of smart city use cases
• A magnetic sensor can detect a parking event by analyzing changes in the surrounding magnetic field
when a heavy metal object, such as a car or a truck, comes close to it (or on top of it).
• A lighting controller can dim and brighten a light based on a combination of time based and ambient
conditions. Video cameras combined with video analytics can detect vehicles, faces, and traffic conditions
for various traffic and security use cases.
• An air quality sensor can detect and measure gas and particulate matter concentrations to give a hyper
localized perspective on pollution in a given area.
• Device counters give an estimate of the number of devices in the area, which provides a rough idea of the
number of vehicles moving or parked in a street or a public parking area, of pedestrians on a sidewalk, or
even of birds in public parks or on public monuments—for cities where bird control has become an issue.
2) City Layer
 At the city layer, which is above the street layer, network routers and switches must be deployed to match
the size of city data that needs to be transported.
 This layer aggregates all data collected by sensors and the end-node network into a single transport
network.
 The city layer may appear to be a simple transport layer between the edge devices and the data center or
the Internet.
 However, one key consideration of the city layer is that it needs to transport multiple types of protocols,
for multiple types of IoT applications.
 Some applications are delay- and jitter-sensitive, and some other applications require a deterministic
approach to frame delivery.
 As a result, the city layer must be built around resiliency, to ensure that a packet coming from a sensor or
a gateway will always be forwarded successfully to the headend station.

igure 5.13 shows a common way of achieving this goal.

3) Data Center Layer
 Ultimately, data collected from the sensors is sent to a data center, where it can be processed and
correlated.
 Based on this processing of data, meaningful information and trends can be derived, and information can
be provided back.
 For example, an application in a data center can provide a global view of the city traffic and help
authorities decide on the need for more or less common transport vehicles.
 At the same time, an automated response can be generated.
 For example, the same traffic information can be processed to automatically regulate and coordinate the
street light durations at the scale of the entire city to limit traffic congestion.
 The key technology in creating any comprehensive smart solution with services is the cloud.
 With a cloud infrastructure, data is not stored in a data center owned directly or indirectly by city
authorities.
 Instead, data is stored in rented logical containers accessed through the Internet.
 Because the containers can be extended or reduced based on needs, the storage size and computing power
are flexible and can adapt to changing requirements or budget conditions.
 In addition, multiple contractors can store and process data at the same time, without the complexity of
exclusively owned space.
  This proximity and flexibility also facilitate the exchange of information between smart systems and
allow for the deployment of new applications that can leverage information from several IoT systems.
 Fig shows the vision of utilizing the cloud in smart solutions for cities.

Figure 5.14 The Role of the Cloud for Smart City Application

 The cloud provides a scalable, secure, and reliable data processing engine that can handle the immense
amount of data passing through it.
 Smart city issues require not just efficient use of infrastructure, which the cloud helps enable, they also
require new data processing and management models.
 For example, cloud services allow for Software as a Service (SaaS) models that create cyclical returns on
investment
 With the cloud approach shown in Figure 5.14 , smart cities can also take advantage of operating expense–
based consumption models to overcome any financial hurdles in adopting solutions to their most critical
issues.
 Critical data, such as air condition (humidity, temperature, pollution) levels monitoring, can be processed
initially.
 Then, as the efficiency of IoT is scaled up, richer data processing can be enabled in the cloud applications.
 For example, the humidity level can be used to regulate the color and luminosity of street lights. In times
when city budgets are strained, data processing can be scaled down to essential services.

4) Services Layer
 Ultimately, the true value of ICT connectivity comes from the services that the measured data can provide to
different users operating within a city.
 Smart city applications can provide value to and visibility for a variety of user types, including city
operators, citizens, and law enforcement.
 The collected data should be visualized according to the specific needs of each consumer of that data and
the particular user experience requirements and individual use cases.
 For example, parking data indicating which spots are and aren’t currently occupied can drive a citizen
parking app with a map of available spots, as well as an enforcement officer’s understanding of the state
(utilization and payment) of the public parking space,
 while at the same time helping the city operator’s perspective on parking problem areas in the city at any
given time.
 With different levels of granularity and scale, the same data performs three different functions for three
different users.
 Along the same lines, traffic information can be used by individual car drivers to find the least congested
route.
 A variation of the same information can be made available to public transportation users to estimate travel
times.
 Public transportation systems, such as buses, can be rerouted around known congestion points.
 The number of subway trains can be increased dynamically to respond to an increase in traffic congestion,
anticipating the decisions of thousands or even millions of commuters to take public transportation instead
of cars on days when roads are very congested.
 Here again, the same type of data is utilized by different types of users in different ways based on their
specific use cases.

2. Explain the features of Arduino?

 Arduino is a prototype platform (open-source) based on an easy-to-use hardware and software.
 It consists of a circuit board, which can be programed (referred to as a microcontroller) and a ready-made
software called Arduino IDE (Integrated Development Environment), which is used to write and upload the
computer code to the physical board.
The key features are −
 Arduino boards are able to read analog or digital input signals from different sensors and turn it into an
output such as activating a motor, turning LED on/off, connect to the cloud and many other actions.
 You can control your board functions by sending a set of instructions to the microcontroller on the board
via Arduino IDE (referred to as uploading software).
 Unlike most previous programmable circuit boards, Arduino does not need an extra piece of hardware
(called a programmer) in order to load a new code onto the board. You can simply use a USB cable.
 Additionally, the Arduino IDE uses a simplified version of C++, making it easier to learn to program.
 Finally, Arduino provides a standard form factor that breaks the functions of the micro-controller into a
more accessible package.