Identifying Indoor Points of Interest via Mobile Crowdsensing: An Experimental Study

TL;DR

This study introduces a mobile crowdsensing method using Wi-Fi signal similarity to accurately identify indoor points of interest, overcoming GPS limitations indoors.

Contribution

It proposes a smartphone-based system leveraging Wi-Fi RSS cosine similarity measurements for indoor POI detection, validated through experimental results.

Findings

Wi-Fi similarity can distinguish different indoor POIs.

The system accurately identifies user-specific and common POIs.

Experimental validation confirms effectiveness in indoor environments.

Abstract

This paper presents a mobile crowdsensing approach to identify the indoor points of interest (POI) by exploiting Wi-Fi similarity measurements. Since indoor environments are lacking the GPS positioning accuracy when compared to outdoors, we rely on widely available Wi-Fi access points (AP) in contemporary urban indoor environments, to accurately identify user POI. We propose a smartphone application based system architecture to scan the surrounding Wi-Fi AP and measure the cosine similarity of received signal strengths (RSS), and demonstrate through the experimental results that it is possible to identify the distinct POI of users, and the common POI among users of a given indoor environment.