Analysis and use of Wi-Fi data for signal state identification

Abstract

Efficient planning, operation and management of transportation facilities require extensive data regarding the traffic demand, patterns and conditions prevalent in the transportation network. Conventional data collection techniques such as loop detectors bear practical limitations such as limited accuracy and applicability, especially in mixed traffic conditions. Since use of smart phones has gained prominence in the recent times, crowd sourced data using Bluetooth and Wi-Fi technologies is perceived to be a reliable alternative for traffic data collection. This eases the rigorous data collection process by considerably reducing the investments on labor, time and other resources. Significant research has been carried out in the extraction and analysis of traffic data from Bluetooth sensors. Changes in privacy settings of smartphones has necessitated the devices to be put on “Discoverable” mode for passive data collection thereby resulting in drastic drops in market penetration rates. Unlike their Bluetooth counterparts, Wi-Fi protocol just requires the Wi-Fi to be switched on for passive data collection, thus resulting in higher penetration rates. This paper presents a preliminary analysis of data extracted from Wi-Fi sensors and the use of it for extracting the signal state information.