Localizacao em ambientes internos utilizando redes Wi-Fi

TL;DR

This paper introduces an indoor localization method using Wi-Fi RSSI data, kNN, and quartiles analysis, achieving high accuracy with minimal APs and rapid response time.

Contribution

It presents a novel indoor localization approach combining kNN and quartiles analysis that significantly improves accuracy with fewer access points.

Findings

Null error with four APs and 10 readings per sample

Location achieved in 0.69 seconds

Method shows promising accuracy in indoor environments

Abstract

This paper presents a localization method for indoor environments capable of improving the location accuracy that is hampered by instability in RSSI of the IEEE 802.11 networks. The method employs the k-Nearest Neighbors (kNN) algorithm and quartiles analysis in the data representation. The proposal had null error with only four APs and 10 readings per sample of each AP with just 0.69 second to locate. These values are important contributions, confirming that the method is promising to locate objects in indoor environments.