Cloud Computing for IoT

Bibhas Ghoshal
IIIT Allahabad
 Traditional Computing
Traditional Computing Components:
Data Storage Core Logic User Interface
1. Data Base 1. Application Specific 1. Command Line
2. File Based 2. Algorithm Implementation 2. GUI

All components sit on the same computer

Core Logic
Distributed Computing : 1. Application Specific
2. Algorithm Implementation

Data Storage
User Interface
1. Data Base
1. Command Line
2. File Based
2. GUI

Internet of Things
2
Instructor : Dr. Bibhas Ghoshal
Distributed Computing

Distributed System : A collection of independent computers

that appears to its users as a single coherent system
Ex : Internet, computing cluster, LAN, Data centre, web
Distributed Application :
One single system or one of several subsystems
Collection of processors – parallel processing ( increased performance,
reliability and fault tolerance)
Partitioned or replicated data
Why Distribution?
Sharing of information and services
Increase in reliability, fault tolerance and improved performance

Internet of Things
3
Instructor : Dr. Bibhas Ghoshal
Goals and Challenges for Distributed Systems

Goals :
Making Resources Available
Distribution Transparency
Openness
Scalability
Security
System Design requirements
Challenges :
Naming and Access control, Security, Performance, Mutual exclusion
of users, replication and migration, failure modes, concurrency,
heterogeneity

Internet of Things
4
Instructor : Dr. Bibhas Ghoshal
Client Server Architectures in Distributed
Systems

Internet of Things
5
Instructor : Dr. Bibhas Ghoshal
Trends in Distributed Computing

Cluster Computing : Homogeneous computing nodes

( connected loosely or tightly) working together

Internet of Things
6
Instructor : Dr. Bibhas Ghoshal
Trends in Distributed Computing

Grid Computing : Heterogeneous computing nodes

distributed over a wide area to perform very large tasks

Internet of Things
7
Instructor : Dr. Bibhas Ghoshal
 Trends in Distributed Computing
Utility Computing : Packaged resources available for
computing and storage ( on rent )

Q: What do you look out for ?

Ans :
1. Can I afford the rent?
2. Is there enough computing and power
And memory?
3. Is it safe from other users - memory / code leaks
4. will the application use affect me?
5. Can I pay for what I use?

Internet of Things
8
Instructor : Dr. Bibhas Ghoshal
Trends in Distributed Computing

Cloud Computing : Model for enabling convenient, on-

demand network access to shared pool of configurable
computing resources.
Ex: network infrastructures, servers, storage, application

Step from utility computing

High level generalization of computing and storage model
It can be rapidly allocated and released with low management effort
Some essential characteristics, service models and deployment
models

Internet of Things
9
Instructor : Dr. Bibhas Ghoshal
Trends in Distributed Computing

Internet of Things
10
Instructor : Dr. Bibhas Ghoshal
 Virtualization
- Share same hardware among independent users
- Increases degree of hardware parallelism
- Easier management and energy usage
- Flexible allocation and utilization
- Decouple apps from underlying hardware
H/W upgrades without impact on OS

Internet of Things
11
Instructor : Dr. Bibhas Ghoshal
 Virtualization Raises Abstraction
- Virtual Memory to access large address space ( Physical memory
mapping is hidden by OS using paging)
- Allow code on one architecture to run on another
- Physical devices to Virtual Devices

Internet of Things
12
Instructor : Dr. Bibhas Ghoshal
Virtualization in Cloud

Internet of Things
13
Instructor : Dr. Bibhas Ghoshal
Cloud Computing Framework

Internet of Things
14
Instructor : Dr. Bibhas Ghoshal
Cloud Computing : Characteristics

Improved agility in resource provisioning

Ubiquitous – independent of device and location
Multi tenancy – sharing of resources and costs accross a
large pool of users
Dynamic load balancing
Highly reliable and scalable
Low cost and low maintenance
Improve security and access control

Internet of Things
15
Instructor : Dr. Bibhas Ghoshal
Cloud Computing : Service Models

Internet of Things
16
Instructor : Dr. Bibhas Ghoshal
Software-as-a Service (SaaS)

●
Facility to execute service provider’s application at users end
●
Services can be accessed via different types of client devices such as web browser or app
●
End users do not have control of cloud platform
●
Example : Google Apps

Internet of Things
17
Instructor : Dr. Bibhas Ghoshal
Platform-as-a-Service(PaaS)
●
Facilitates consumer created or acquired applications onto cloud infrastructure
●
Support for deployment of such applications
●
Users do not control cloud infrastructure
●
Users can control deployed apps using given configuration
●
Ex : Google App Engine , Windows Azure

Internet of Things
18
Instructor : Dr. Bibhas Ghoshal
 Infrastructure-as-a Service(Iaas)
●
Access of computing resources such as Operating System, Storage and Network
●
Users can deploy and execute any software
●
In some cases, users can control selected network components
●
Ex : Amazon EC2, GoGrid

Internet of Things
19
Instructor : Dr. Bibhas Ghoshal
Cloud Computing Services

Internet of Things
20
Instructor : Dr. Bibhas Ghoshal
Cloud Computing : Deployment Models

Internet of Things
21
Instructor : Dr. Bibhas Ghoshal
 Cloud Computing for IoT
Cloud is an IoT enabler :
Huge amount of IoT Data needs storage, retrieval and management ( sustained services)
Fast Analytics for Prediction and Critical Decision making
Benefits of Cloud in IoT :
Scalability ; Data Mobility; Time to Market; Security; Cost effectiveness

Dr. Internet
Bibhas Ghoshal,
of ThingsIIIT Allahabad
24/02/2021
Instructor
Recent Trends
: Dr. Bibhas
in SIoTGhoshal
( RTSIoT - 2021 )
IoT Cloud Architecture

Internet of Things
23
Instructor : Dr. Bibhas Ghoshal
 IoT Management
●
IoT-A Project [1] : physical entities represented as virtual entities access point to
the real world by IoT apps through well defined and standardized interfaces
●
European FP7 iCore project [2] : virtual object as alter ego of real world object,
dynamically created and destroyed. VOs interact giving them some Cognitive
capabilities

[1] [Link]
[2] Giaffreda : iCore: A Cognitive Management Framework for the Internet of Things, In: Galis A., Gavras A. (eds)
The Future Internet. FIA 2013. Lecture Notes in Computer Science, vol 7858.

Internet of Things
24
Instructor : Dr. Bibhas Ghoshal
Cloud Based IoT Platforms - Characteristics

●
Objects use HTTP protocol to send and receive data; allows interoperability among platforms
●
Objects do no directly communicate with each other. An intermediate server is used
●
Every object has a data point associated with it on the server side to keep track of the data sent
●
The methods POST and GET are are used to send and request data
●
Tag is assigned to every data point
●
Data point discovery is performed through tags through an internal search engine
●
System identifies every object with its API key
RESTful Architecture:
Lightweight, scalable and then fits perfectly with the principles and the current protocols of the Internet.
Resources as representation of the objects, that are uniquely identified through Uniform Resource
Identifiers (URIs).
Object information can be obtained, deleted or posted through the HTTP protocol using a given method
(GET, DELETE, POST, PUT).
The payload of the message can be incapsulated in a negotiated format such as XML or JSON.

Internet of Things
25
Instructor : Dr. Bibhas Ghoshal
 IoT Platform Implementation - 1
Cosm ( formerly Pachube) [6] : platform to store and redistribute real-time data, freely usable, which manages millions
of devices per day. Following the nuclear accidents in Japan in 2011, Cosm was used by volunteers to interlink Geiger
counters across the country to monitor the fallout.
Xively (formerly Cosm) [7] : commercial IoT platform to transmit, store and access to data generated by objects.
It exploits PaaS partially ( data stored in provider DB only). Now owned by Google.

[3] [Link]
[7] [Link]

Internet of Things
26
Instructor : Dr. Bibhas Ghoshal
 IoT Platform Implementation - 2
Nimbits [8] :
●
open source web application built on Google App engine
●
Provides email alert, math calculations
●
Storing and processing data
●
User can define data points and use them to share several kinds of
data
●
Integrated with Twitter, Facebook
●
Allows to manage data points, share sensor diagrams

[8] [Link]

Internet of Things
27
Instructor : Dr. Bibhas Ghoshal
 IoT Platform Implementation - 3
MATLAB ThinkSpeak

[Link]

Internet of Things
28
Instructor : Dr. Bibhas Ghoshal