Department - Author 1

Computer Engineering Department

Degree Name - Author 1

BS in Computer Engineering

Date

2010

Primary Advisor

Christopher Clark

Abstract/Summary

Wireless Local Area Networks using Wi-Fi is becoming more and more ubiquitous. As such, they provide a potential pre-built infrastructure for small area localization. This project serves as a proof of concept for a playground child tracking system to be deployed at Cal Poly's Child Development Playground Lab. The two main options for doing Wi-Fi localization are triangulation and fingerprinting. Triangulation involves mapping signal strength as a function of distance while fingerprinting creates a probability distribution of signal strengths at a given location and uses a map of these distributions to predict a location given signal strength samples. The triangulation method did not show promising results and the fingerprinting method had promising results with various ways of making predictions.

Fingerprinter.zip (85 kB)
C# Fingerprinting Tool

analysis of design V2.docx (22 kB)
Analysis of Design

FingerprintParser.cpp (4 kB)
C++ Fingerprint Parser

Markov.cpp (21 kB)
C++ Markov Localization

avg_10f_tr.csv (4 kB)
Average RSSI Values

stdev_10f_tr.csv (3 kB)
Standard Deviation of RSSI Values

COinS