write a review for this IEEE paper

2017 IEEE International Conference on Technological Innovations in ICT For Agriculture and Rural Development

(TIAR 2017)

978-1- 5090-4437- 5/17/$31.00 ©2017 IEEE

IoT Application Development: Home Security System

Raj G Anvekar School of Electronics and Communication

Engineering, KLE Technological University Hubli, India

[email protected]

Dr. Rajeshwari M Banakar Professor, Member IEEE

School of Electronics and Communication Engineering, KLE Technological University Hubli, India

[email protected]

Abstract—One of the prominent application segment of Internet of Things framework is in the Security Sector. It is important to arrive at a unique low cost solution to prevent theft and ensure security to members of the home. The Internet of Things(IoT) Layered Architecture based design approach assists the system designer to conveniently differentiate the system component requirements distinctly at various layers. This papers highlights the model driven development process for Home Security System. It remarks the uses of customers end application such as Telegram to securely transmit information through layers of IoT architecture. This paper aims at providing a low-power, cost effective and unobtrusive IoT based home security system which assist in presence detection, identification and authentication of stranger. The proposed solution makes use of USB Webcam as an image capturing unit, Electric Door Strike as an actuator and Telegram which has an amazing feature as Telegram Bot which provide APIs to build solutions which is compatible with Raspberry Pi IoT infrastructure.

Keywords—Internet of Things, Telegram, Raspberry Pi, Home Security, Electric Door Strike, Doorbell.

I. INTRODUCTION

Internet of Things (IoT) is an interconnection of physical objects and other devices embedded with electronics, software, network connectivity and application specific sensors and actuators that enables the objects to collect and exchange data. It is an important topic in technology industry and its applications has gained popularity in recent years and is predicted to have billions of smart objects connected together by 2020. IoT has become a major platform in the field of Home Automation and Security, and foresee large number of application possibilities in near future. Smart connectivity [1] with existing networks and context-aware computation using network resources is an indispensable part of IoT. In today's world, Wi-Fi and 4G-LTE wireless Internet access is predominantly used, which justifies the evolution of ubiquitous information and communication networks. However, for the Internet of Things vision to successfully emerge, the computing standard will need to go beyond conventional mobile computing scenarios that use smart phones and portables, and evolve into connecting everyday

existing physical objects and embedding intelligence into our environment.

Security has become a major issue everywhere and its in this crime ridden society. The

role of IoT infrastructure in the development of home security system is significant. The security of home members acts as a fore runner in the development of home security systems. The primary aim is to develop feasible solution to transform a traditional doorbell into an intelligent bell which provides information of the stranger to home owner thereby enabling him to answer the door through smartphone with easy user interface. The design solution in IoT system is multidisciplinary and is scattered through various domain specific challenges.

In [2], Ovidiu and Peter discuss the growth and application of IoT based design systems in the near future. The authors give an IoT layered architecture illustration which is useful for the IoT system designer. The networking infrastructure and protocols requires special attention during system level integration of IoT. The network infrastructure demands [3] need of high quality data and failure free wireless communication. There are many researches that anticipate the failure of wireless network which is growing as the use of technology evolves. The main factors are network topology, failure of network measurement and cost minimization which decides the efficiency of the network and data delivery. The data delivered or exchanged over internet can be audio, video or an image file. Jayashri and Arvind [4] proposes the use GSM based security system to enhance the security. It illustrates two methods of home security system, first one has a webcam which captures the motion and sends a mail to the owner of house. Second uses GSM-GPS module to send SMS to the owner with the assistance of sensor and micro-controller. In [5] Arun and Doreen uses Bluetooth as a wireless standard to detect, identify and locate entities inside home. WLAN is setup to collate a server data at a centralized server. Uses 3 Bluetooth beacon which provide the RSSI value and supports 3 decision model, they are peak RSSI decision model, supervised and unsupervised learning model for detection, identification and localization of the entities. Exchange of information in the form of video or image is the requirement

2017 IEEE International Conference on Technological Innovations in ICT For Agriculture and Rural Development

(TIAR 2017)

978-1- 5090-4437- 5/17/$31.00 ©2017 IEEE

of varied application, including smart transportation, smart health care and smart structures. Umar Albalawi and Prabha Sundaravadivel [6] presents a modular and extensible quadrotor architecture and its specific prototyping for automatic tracking applications. To demonstrate the power and flexibility of proposed design, a target tracking and acquisition applications is presented here. The designed module implements the basic proportional integral derivative control and a custom target acquisition algorithm. Importance of the sliding-window-based algorithm is also presented.

II. IOT LAYERED ARCHITECTURE: HOME SECURITY SYSTEM

The IoT layered architecture consist of four major layers. They are Application Layer, Information Service Layer, Networking Infrastructure Layer and Smart Object Layer. The Smart Object Layer mainly consist of sensors for data acquisition and actuators for data deployment. It is the base layer of IoT Layered Architecture.

Fig. 1. IoT Layered Architecture

The sensors of Smart Object Layer interact with real time parameter and enables data collection which can be used by the upper layers. Fig. 1 illustrates the IoT Layered Architecture. Choice of the sensors and actuators in an IoT based system design makes them Smart Objects. These Smart Objects can be used in distinct application as they can be programmed using appropriate Internet modules. As they are programmable they can be used to provide user services. In the IoT layered architecture, Information Services Layer mainly deals with end user applications. It has Software as a Service (SaaS) as its important design attribute. SaaS allows user to connect and use cloud based application over Internet. The customer specification governs the designer's model approach. The salient features of SaaS is that user can gain access to sophisticated applications, access applications data from anywhere and its ability to automatically update to the latest version. The Networking infrastructure Layer acts as a bridge between Smart Object and Information Service Layer. For the data transmission between the door bell and

the Raspberry Pi, Webcam acts as an intermittent node. The APIs of webcam are installed in order to take a picture of

as a source of communication between source node and target node. It includes Ethernet LAN cable, USB Wi-Fi Adapter, which connects Raspberry Pi to Internet. The Information Service Layer uses SaaS to reach customer end using Internet. It sends and receives the information in the form of data packets. The social networking application such as WhatsApp, Gmail and Telegram acts as SaaS for customer end applications. The Application Layer is customer specific design feature. It supports various application in field of security, home automation, energy, transportation and automobiles. It can receive and retransmit information through the layers on IoT architecture. Fig. 2 illustrates IoT Layered Architecture in Home Security System.

Fig. 2. IoT Layered Architecture for Home Security System

III. IOT APPLICATION DEVELOPMENT: CONCEPTUAL MODEL

With the increased availability of consumer sensing devices including modern smartphones, the domain of networked sensing has evolved into the Internet of Things, where application exhibit traditional Wireless Sensor and Actuator Network (WSAN) features and such large number of sensing and actuation device are coupled with increased device heterogeneity and integration with database server, etc. As IoT has gained wide acceptance the need for sound software engineering approaches to adequately manage the development of complex application arises. Based on roles

Fig. 3. Conceptual Model of Home Security System

2017 IEEE International Conference on Technological Innovations in ICT For Agriculture and Rural Development

(TIAR 2017)

978-1- 5090-4437- 5/17/$31.00 ©2017 IEEE

defined in Fig. 2 we propose the following model driven

development process that consists of following steps as shown in Fig. 3.

A. Domain Specific Concept The concepts that fall into this category are specific to a

target application domain. For example, the Security System domain is reasoned in terms of rooms and doors. An entity of interest (EoI) is an object (e.g., room, book, door), including attributes that describe it, and its state that is relevant from a user or an application perspective. The entity of interest has an observable property called phenomenon. Typical examples are the sensing range of motion detector of a door and a tag ID. The conceptual representation of an actuator, a sensor, a storage or a kind of user interface is a resource. These are the following types of resources:

A sensor has the ability to detect changes in the environment. PIR and Webcam acts as a sensing device. The sensor observes a phenomenon of an EoI. For instance, the PIR sensor detects the presence of a stranger in front of a door.

An actuator makes changes in the environment through an action. Electric door strike and alarming systems are examples of actuators. The actuator affects a phenomenon of an EoI by performing actions. For instance, the electric door strike controlled by smartphone allows the owner to enter the home.

A storage has the ability of storing data in a persistent manner. The storage stores information about a phenomenon of an EoI. For instance, a database server stores information in the form of picture captured using webcam about the people

A user interface represents tasks available to users to interact with entities of interest. For home security system, it sends a notification to home owner indicating the presence of stranger at his door. On authenticating the identity, the owner unlocks the door using his smartphone.

B. Functionality Specific Concept The computational element of an application mainly the

software component and any interaction among them falls in this category. A computational element is basically a type of software component that encloses a part systems functionality and data, and restricts access to that part via explicitly defined interfaces. We use the term application logic to refer a functionality of a software component. The application logic to send information of visitor to owner on pressing the doorbell, to unlock the door using smartphone etc. This concept includes a software component with specific functionality, a computational service related to it, which is a type of software component that consumes one or more parameter information as inputs, processes it, and

produces an output. An output could be command or a data message that is consumed by others or data message that triggers an action of an actuator. A computational service is a representation of the processing element in an application.

A software component communicates with other software components to exchange data or control. These interactions occurs based on different modes such as request- response and command. This is in principle as an instance of the component-port-connector architecture used in software engineering.

C. Deployment Specific Concept All the devices employed for an application falls in this

category. Each device hosts zero or more resources. For example, a device could host resources such as a PIR sensor

input data and give output data to actuators, a smartphone to display information of stranger, a door strike to unlock the door and a storage to store the information of stranger.

D. Platform Specific Concept The computer programs mainly the device drivers which

acts as a translator between a hardware and an application

falls in this category. We identify the following platform

specific concepts:

For efficient functioning of a sensor, sensor driver

is required. It accesses data observed by the sensor

and generates the meaningful data that can be used

by other software components. For instance, PIR

sensor generates a value on detecting the presence

of visitor and precedes this value to Raspberry Pi

for further calculation.

An actuator driver is required to allow the actuator

to function appropriately. It translates a command

from other software components and actuates the

actuator appropriately. For instance,

locking/unlocking mechanism of electric door

strike when triggered by a user application.

In order to provide read and write access to storage

device, a storage service is required. A storage

service provides access to the storage. Other

software components uses this storage service to

access data from the storage. For instance, MySQL

storage service provides access to a database

server.

An end-user application is a basically a type of

software component that is required to help a user

to perform tasks (e.g., receiving notifications,

submitting information). It provides access to

available tasks. For instance, in the home security

application a user can authenticate the identity of

visitor and allow access to home using an

application installed on his phone.

2017 IEEE International Conference on Technological Innovations in ICT For Agriculture and Rural Development

(TIAR 2017)

978-1- 5090-4437- 5/17/$31.00 ©2017 IEEE

Fig. 4. Flowchart of Home Security System

The flowchart in Fig. 4, illustrates the initialization

process to invoke the webcam and sets it to active mode.

The steps followed by IoT infrastructure to reflect the output

response are also shown. When a visitor presses the

doorbell, it sends an analog signal to Raspberry Pi which

triggers the Webcam to take the picture of visitor. Pi

connects to Internet and sends the data to Telegram server.

The server responds to the request and sends the data to

Telegram user. The user authenticates the identity and

allows access to visitor by unlocking the door remotely

using his smartphone. The above algorithm is repeated

every time the bell is pressed or motion is detected.

IV. DESIGN METHODOLOGY

The methodology used for making the system operate is divided into two important phases.

A. Human Detection and Authentication When a visitor approaches a home and presses the

doorbell, an analog signal is fed to Raspberry Pi. If the visitor does not presses the doorbell, the PIR sensor detects his presence and sends signal to Pi which triggers the webcam to take the picture of visitor. The algorithm is deployed on Pi to process the data. Pi connects to internet and sends the data to Telegram server. Telegram is highly secure as it supports AES (Advanced Encryption Standard) and Triple Data Encryption standards.

In this part we have installed Telegram libraries in the Raspberry Pi to make use of Telegram application and created a new Telegram account for the Raspberry Pi by registering a mobile sim card number. Once we register the mobile number, a Telegram code will be sent to that respective mobile number for activation. The captured image as shown in Fig. 5 is sent using Send_Photo command of Telegram. Along with the photo, a notification message is also sent using command Send_Message. Thus, at the sender end, it is the Raspberry Pi which does the main work, but data has been sent through Telegram. The user authenticates the identity and take necessary actions.

Fig. 5. Picture of Visitor

B. Restrict/Allow access to home To take control action we have used an amazing feature

of Telegram which is known as Telegram bot. Bots are answer machines that takes the proper action for questions asked or commands sent.

On the Raspberry Pi Telegram bot libraries must be installed and upgraded to the new version. The controlling action is implemented by writing python script. In this experiment, when user sends commands to control the Raspberry Pi from his Telegram account to the bot, python script will get called. Function bot.notifyOnMessage(handle) will intimate the bot that some command has been received. To check which command has been sent, a series of else-if conditions are written inside the handle(msg) function. Hence the Telegram bot calls this function and when the command sent by the user matches with the options written inside the handle(msg) function it will execute the commands which are under the else-if keywords. Inserting the token of Telegram Bot in python script is very important. In Telegram bot, messages can be sent by using function bot.sendMessage(chat id, msg). To execute the OS commands we can use sub process python module or OS system module. In order to control the door lock, home

/unlock /lock commands to the bot. The IoT device and smartphone need not be connected to the same network.

V. EXPERIMENTAL RESULTS

We have performed the experiment by interfacing Webcam and Electric Door Strike with the Raspberry Pi. The use of Telegram application to receive the messages and send the commands makes the whole system user friendly and effective. When the home owner receives the pictorial information of stranger/visitor, he authenticates the identity and takes action by sending commands to Raspberry Pi from his Telegram account using his smartphone connected to Internet. The door locking mechanism is implemented using Electric door strike. When the Raspberry Pi receives

2017 IEEE International Conference on Technological Innovations in ICT For Agriculture and Rural Development

(TIAR 2017)

978-1- 5090-4437- 5/17/$31.00 ©2017 IEEE

/unlock command from user/home owner, it runs python code which makes respective GPIO pins high, triggers the relay and unlocks the door.

Fig. 6. System Implementation

In the same way the door will get locked by making respective GPIO pins low. Fig. 6 represents the system implementation in real time. Thus by sending commands to the bot we can control any actuators that are connected to IoT device. Also this action can be taken from any remote place. Once the door gets unlocked a notification The door has been

system is highly responsive and robust.

VI. CONCLUSION

An IoT enabled design framework is presented for

security sector. The conceptual model developed for the IoT

based Home Security System has been built using the

fundamentals of IoT Layered Architecture feature. The

successful implementation of the proposed design solution

for IoT Home Security System, with the projection of 21

billion Smart Objects by 2020 advances the knowledge

during system integration. In this paper, we have

conceptualized and implemented a low-power, cheap and

unobtrusive IoT based Home Security System which assist

in presence detection, identification and authentication of

stranger. The system requires extremely low maintenance

and can be employed in smart homes of tomorrow. Future

work is on, a similar approach in developing a prototype for

face detection and identification using the concept of

machine learning and artificial intelligence.

REFERENCES

[1] Jayavardhana Gubbia, Rajkumar Buyyab, Slaven Marusic,

journal on Future Generation Computer Systems 29 (2013)

1645 1660.

[2]

Converging Technology for Smart Environments and Integrated

Eco-

Aalborg 2013, Denmark.

[3] Elisha Colmenar, Fadi Al-Turjman, Mohammad Biglarbegian,

Networks, WLN 2014, Edmonton, Canada.

[4]

Implementation of Security Systems for Smart Home based on

No.6 (2013), pp.201-208.

[5] Alaa Alhamoud, Arun Asokan Nairy, Christian Gottron, Doreen

Identification and Tracking in Smart Homes Utilizing Bluetooth

Wireless Local Networks, WLN 2014, Edmonton, Canada.

[6] ELIAS KOUGIANOS, SARAJU P. MOHANTY, GAVIN

COELHO, UMAR ALBALAWI AND PRABHA

SUNDARAVADIVEL, " Design of a High-Performance

System for Secure Image Communication in the Internet of

Things", SPECIAL SECTION ON SECURITY AND

RELIABILITY AWARE SYSTEM DESIGN FOR MOBILE

COMPUTING DEVICES, Date of current version April 6,

2016.