Date of Publication

3-30-2023

Document Type

Bachelor's Thesis

Degree Name

Bachelor of Science in Electronics and Communications Engineering

Subject Categories

Electrical and Computer Engineering | Engineering | Systems and Communications

College

Gokongwei College of Engineering

Department/Unit

Electronics And Communications Engg

Thesis Advisor

Elmer R. Magsino

Defense Panel Chair

Ann E. Dulay

Defense Panel Member

Jose Antonio M. Catalan
Cesar A. Llorente

Abstract/Summary

In this study, a smart parking system based on multiple Wi-Fi Received Signal Strength Indicator (RSSI) fingerprinting methods was developed on scaled-down models of an indoor parking system by employing multiple Wi-Fi access points (APs). The development of a smart parking system is essential for smart cities, as it minimizes time consumption, CO2 emissions, and the stress caused by searching for parking spaces. Thus, the main objective of the proponents in this project is to effectively and efficiently create a system capable of detecting slot availability. In doing so, the presence of Wi-Fi routers were exploited – the Wi-Fi’s RSSI signals were collected in order to establish the parking fingerprint settings and then later used to predict the number of occupied/vacant slots. The study was conducted in two phases, namely offline training and online matching phases. In the offline stage, the system gathered Wi-Fi RSSI readings to determine the parking lot's fingerprint and stored it in a fingerprint database that can be updated periodically. During the online stage, the number of available parking slots was predicted based on the actual scenario compared to the stored database. In coming up with predictions, two methods were used, the KNN algorithm and cross-correlation. During their experimentation with the KNN Algorithm, the proponents used a downsized model of an indoor parking facility to achieve a realistic representation, within a 8.60 square meter bedroom in a larger 60-square-meter house. The model measured 1.1 meters in width and 0.4 meters in height, which only catered 5 cars and a maximum of 6 routers. Meanwhile, when experimenting with cross-correlation technique, they have utilized multiple router setups in generating Wi-Fi signals and exhaustively considered all possible parking scenarios given the combination of 10 maximum access points and 10 cars, in two separate locations with a parking area dimension of 13.40 m2 and 6.30 m2 . Using the KNN algorithm, the researchers acquired the best accuracy of 71.86% when six routers were used, meanwhile, a perfect prediction rate of 100% was achieved using cross-correlation and 8 APs. Moreover, the developed system serves as experiential evidence on how to exploit the available Wi-Fi RSSI readings towards the realization of a smart parking system.

Abstract Format

html

Language

English

Format

Electronic

Keywords

Automobile parking; Parking facilities

Upload Full Text

wf_yes

Embargo Period

3-29-2023

Download

Share

COinS