RMS: A Delay Sensitive Road Monitoring System Using Edge Intelligence

Abstract

Smart vehicles equipped with onboard sensors can detect the surrounding cars, cycles, pedestrians, and other objects in real time and can make control decisions to achieve road safety. The high dynamism of vehicular mobility switches base stations in their trajectory, and at an instant, one edge server services one base station only. So, making control decisions by processing and transmitting vehicular data in real time is challenging. In this article, we propose a road monitoring system (RMS) in which moving vehicles collect road data and offload collected data to the paired edge server. The edge devices and connected vehicular networks receive road-hindering alerts. It comes up with Road Ambient Intelligence that supports ubiquitous computing and generates context-aware alerts with low latency. It uses a deep learning model for the decision-making about road-hindering alerts and a <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">${K}$ </tex-math></inline-formula> -nearest neighbor algorithm to form clusters for the distribution of alert messages. For intense occurrences, the data are sent to the cloud server for further notifications to remote-control offices. The proposed study is compared with social vehicle route selection (SVRS) and delay tolerant processing (DTP) schemes. The performance analysis shows that rms presents 12%–44% better performance compared to the SVRS scheme and 15%–53% better compared to the DTP scheme in terms of delay. The energy requirement analysis shows that compared to SVRS, RMS involves 50% less energy when processed at edge clouds, and DTP involves 51%–65% more energy requirements compared to the RMS scheme.