Improving Smart Home Security;

Integrating Logical Sensing into

Smart Home
 ABSTRACT

 This paper explains various security issues in the existing home

automation systems and proposes the use of logic based security

algorithms to improve home security.

 The work classifies natural access points to a home as primary and

secondary access points depending on their use.

 Logic based sensing is implemented by identifying normal user behavior at

these access points and requesting user verification when necessary.

 User position is also considered when various access points changed states.
CONT….

 Moreover, the algorithm also verifies the legitimacy of a fire alarm

by measuring the change in temperature, humidity and carbon

monoxide levels, thus defending against manipulative attackers.

 The experiment conducted in this paper uses a combination of

sensors, microcontrollers and ZigBee communication to identify

user behavior at various access points and implement the logical
sensing algorithm.
 INTRODUCTION

 As technology advanced with time, electronic devices and internet

became more popular and affordable.

 So the concept of home automation and people’s expectation from a

smart home has changed dramatically.

 Modern smart home is a sophisticated combination of various

Ubiquitous Computing Devices and Wireless Sensor/Actor Networks.

 The concept of home automation security has also evolved with time,

sensors and actuators were integrated into the home to detect, alert and
prevent intrusions.
CONT…

 Thus we improve the Smart Home Security by

Integrating Logical Sensing into Smart Home

 The proposed work uses microcontrollers and sensors to

detect and distinguish between normal and attack user

behaviors at various access points.
 EXISTING SYSTEM

 The previous system uses Infrared (IR) grids and wearable

identification (ID) badges to identify the location of a user

at home and predict the context of user actions.
 DISADVANTAGES

 IR grid proved to be difficult to implement in a home

environment while the wearable ID tags proved to be

inconvenient and provided misleading information for
inexperienced and careless users.

 Static object checking severely limits the flexibility of the home

environment and when these static objects are shifted the

proposed system is easily fooled and is unable to adapt.
 PROPOSED SYSTEM

 In this paper, we analyzed various access points in a home to identify

different improbable scenarios within a smart home during its operation.

 Access points are inherent in the structure of a home, which can be used

for entering and exiting a home.

 In a typical home these natural access points are front door, back door,

balcony doors and windows.

 Thus here we proposed a smart home security by means of Integrating

Logical Sensing.
Cont….

 The proposed work uses Raspberry Pi, microcontrollers

and sensors to detect and distinguish between normal and

attack user behaviors at various access points.
Cont….

 Distinguish between primary and secondary access points in a home

based on how they are used. Detect all user actions at these access
points.

 Understand user behavior after change in state of an access point.

 Identify and isolate attack behavior by analyzing the user behavior at

various access points using our logical sensing algorithm. Trigger

warning or raise alarms depending on the situation.
Cont….

 If Door State is changed, Get multiple Proximity and

Motion Sensor values before door state change.
 Otherwise, Keep Monitoring the Primary Access Points.

 Then, Start Door Observation Timer.

 If Door Observation Timer Expired, Identify intermediate

and final states of the door along with sensor values after
the door’s final state change.
Cont….

 Else, Keep observing primary access point state changes

to identify intermediate state AND Collect Motion and
Proximity Sensor values after the door state has changed.
 Then increment Door Observation timer, and again
check the Door Observation Timer Expired.
 Then apply the Logical Sensing algorithm to the gathered
sensor values and identified door states to determine if
state of the home has changed.
Cont….

 Then if the Home was Empty AND User Entered the

Empty Home, Call the subroutine for user verification.
 Otherwise, Wait for state change timer to expire (if any)
while observing the door status.
 If Home State Changed from Occupied to Empty, Call the
subroutine for secondary access point check.
 Finally the process ends.
Cont….

 To Identity Verification Subroutine, Start the Activation

Timer, then Activate the Identity Verification Mechanism.
 If user Passed Identity Verification, User Identity
Confirmed. Reset the Activation Timer.
 Change Home Status to, “Occupied”. Reset all previous
states.
 Finally ends the process.
Cont….

 Or else Increment activation timer.

 If Activation Timer > Time allowed for Identity

Verification.
 So, User Identification Failed, Intrusion Detected.

 Then, Activate Defensive Measures like audible alarms,

silent alarms etc.
 Finally the process ends.
Cont….

 The proposed work observes primary and secondary

access points to identify logical sensing parameters and

detect intrusion and does not cause inconvenience to the
user with wearable tags or laser grids.

 It offers implementation ease and flexibility compared to

the previously proposed security system

Cont….

 The system requires minimal user input to identify when

the home becomes empty or occupied, it was able to

observe various access points in the home and deduce the
change of state of the home.

 The algorithm was able to successfully predict home state

changes and activate identity verification mechanisms

when necessary.
Cont….

 All the ZigBee wireless communication used in the work is

encrypted using 128 bit AES encryption and the encryption key is
never exchanged in clear text over the air, so an eavesdropping
attacker will not be able to gain access to the system and manipulate
it.
 We utilized context aware computing to improve home security, user
context is identified by saving, analyzing and sharing data regarding
user behavior and context which raised security and privacy
concerns.
Cont….

 In the proposed work, access point data is stored in a data

base on Raspberry Pi which is kept inside the home and

secured using physical locks with access limited to
authorized personals.

 Moreover, the stored data is never shared and it can be

encrypted to further improve security.

Cont….

 Previously, utilized NFC tags to predict user location in a

home and the user had to verify his fingerprint and enter
the password each time when the user wants to access
their home.

 The security algorithm proposed in the paper, did not

utilize NFC communication tags which reduces the

complexity of the system and improves user convenience.
Cont….

 Moreover, the IVM is only triggered when someone

enters an empty home.

 The algorithm was also able to identify secondary access

point actions initiated by the user and was able to

distinguish them from intruder actions.
Cont….

 This paper detects user actions at primary and

secondary access points in a home using different

sensors.

 These detected user actions and behaviors are

compared with normal user behavior at various

access points to identify intrusions or intrusion
attempts.
 ADVANTAGES

 Successful smart home security system which has

high accurate sensing.

BLOCK DIAGRAM
 EXPLANATION

 Based on the purpose of the access points, the paper classifies access points into

primary and secondary.

 In a home, when an access point is used by its inhabitants as a primary means to enter

and exit from their home, it is categorized as primary access point like the front door,
back door etc.

 On the other hand, secondary access points like the window, balcony door etc. also

provide entry/exit to a home but they are rarely used for that purpose because there are
other convenient ways in and out of a home for a legitimate user.

 Four boards are designed to obtain logical sensing parameters.

 Board I is deployed near the primary access point.

CONT….

 It consist of wires to proximity, motion and contact sensors to obtain various

logical sensing parameters.

 The board is connected to the Arduino Uno module which supplies the power

to the sensors.

 The Arduino Uno microcontroller is then connected to a ZigBee

communication module.

 Board II is deployed near the secondary access point. It is connected to two

contact sensors, a motion and proximity sensors.

 Each of the contact sensors are connected to window and balcony door.
CONT….

 The motion and proximity sensors are deployed near the window to identify user

movements before the window is opened.

 Board II also has provisions to be connected to Board IV which is connected to the

force sensors in the bed.

 Similar to board I, board II is also connected to Arduino Uno microcontroller

which powers the board and communicates the sensor values to the Pi using
ZigBee module

 Board III is deployed in the kitchen, it is connected to MQ 9 carbon monoxide

sensor, DHT 11 temperature and humidity sensor. Board III is designed to measure
the carbon monoxide, temperature and humidity levels in the area to detect fire.
CONT….

 Board IV is deployed near the bed it is connected to two circular 0.5

inch diameter force sensing resistors.

 Both force sensors are deployed underneath the mattress.

 One in a region where the user places his shoulder and another close to

the middle where the user’s abdomen and pelvic region usually rests.

 Both force sensors has a full scale measurement accuracy of ±5%.

Moreover, force sensors used in the experiment are readily available,

cost effective, flexible and provides un-obstructive force
measurements.
 HARDWARE REQUIREMENTS

 Arduino UNO

 Passive Infrared (PIR) Motion Sensors

 HC-SR04 ultrasonic range sensors

 SOFTWARE REQUIREMENTS

 Php

 MySQL database

 Xampp

 Arduino Interactive Development Environment (IDE).

 CONCLUSION

 The paper detects user actions at primary and secondary access points in a

home using different sensors. These detected user actions and behaviors are

compared with normal user behavior at various access points to identify

intrusions or intrusion attempts.

 The alarm was triggered five times when the user failed to confirm his identity.

 In addition to identifying intrusions in home, the algorithm also warns user

about imminent and live potential security vulnerabilities by identifying the

status of various access points, user position and behaviors.

 Thank You….