A novel framework to avoid traffic congestion and air pollution for sustainable development of smart cities

Abstract

Traffic management is crucial for the sustainable development of smart cities. There has been a continuous emphasis from the research community to predict air quality and manage traffic for congestion-free roads. This paper proposes a novel framework to identify traffic congestion and estimate air quality using crowdsourced information. Navigation Reference Spatial Data (NRSD) was created using the spatial likelihood method using GPS trajectories and OpenStreetMap. This NRSD was also used to design the traffic environment using distance, speed, and time interval mapping. After the NRSD creation, the framework used the real-time data of the registered user to compute the traffic density using the Graham scan with the k means algorithm. Eight standard air quality parameters were used to identify the air quality index of a particular node using a Bayesian Classifier. A case analysis was performed on the road network of three major cities of north India lying within a bounding box of (lat:30.518065, lon:76.659055) to (lat: 30.512501, lon:76.658808) was performed to validate the proposed framework. Based on the performed experiment, the proposed framework achieved maximum accuracy of 98 percent. Similarly, NSRD improves the framework's accuracy by 3.8% in comparison to the baseline dataset. In 87 percent of scenarios, this framework can also determine the shortest path by avoiding traffic.