Date of Publication

8-2-2023

Document Type

Bachelor's Thesis

Degree Name

Bachelor of Science in Computer Science major in Network and Information Security

Subject Categories

Information Security

College

College of Computer Studies

Department/Unit

Computer Technology

Thesis Advisor

Arlyn Verina O. Tiu
Fritz Kevin S. Flores

Defense Panel Chair

Fritz Kevin S. Flores

Defense Panel Member

Marnel S. Peradilla
Geanne Ross L. Franco

Abstract/Summary

The explosive growth of the field of IoT. The smart home continues to rise in popularity, with a number of nodes that outnumber the population of the cities. This brings a rising interest in smart home security. While smart home security systems have been made before, a true context-aware home security system that can provide a new level of access control faces several issues stemming from the chaos of the home environment. The paper observed and collected data from a home environment using a network of ESP32 sensors. The data was preprocessed and run through several machine learning algorithms to find a series of attributes that were most helpful in access control for a smart home. Among the observed attributes, attributes that help identify appliances with lots of limits to who can access them such as Object IDs, Object Type, and Location have been observed to be integral to access control. Adjacent attributes such as Participant IDs with low privilege have also been observed to be useful in access control by various machine learning algorithms. Performing a Principal Component Analysis shows that these attributes are also important in describing the data collected. The study achieved its objectives where it identified Object IDs, Object Type, and Location as attributes that are useful in access control in a context-aware environment. Nonetheless, it's crucial to acknowledge that the attributes identified may differ based on the particular context and environment. As such, while these findings hold true within the scope of this study, varying contexts could result in a different set of significant attributes.

Abstract Format

html

Language

English

Format

Electronic

Physical Description

40 leaves

Keywords

Context-aware computing; Home automation; Home computer networks

Upload Full Text

wf_yes

Embargo Period

8-14-2023

Download

Share

COinS