EVAC-AV: The Live Road Surveillance Control Scheme Using an Effective-Vision-Area-Based Clustering Algorithm With the Adaptive Video-Streaming Technique

Abstract

A new type of safety service in vehicular ad hoc networks (VANETs) is live road surveillance (LRS) video streaming. However, video streaming is a high-bandwidth-demanded service and causes network congestion if there are many vehicles that stream video in the wireless network at the same time. Clustering is an efficient technique to relieve network congestion in VANET because it can reduce the number of vehicles to transmit data. The adaptive video-streaming technology can also relieve the aforementioned problem based on monitoring the wireless network's utilization situation. This paper proposes to provide the LRS service using an effective-vision-area-based clustering algorithm with the adaptive video-streaming technique (EVAC-AV) based on the effective vision of the cameras in vehicles. EVAC-AV can construct clusters such that a driver X can join a suitable cluster to get the LRS service from the appropriate vehicle ahead of X to have additional vision. Additionally, the adaptive video-streaming technique is adopted to have an adaptive LRS video quality in EVAC-AV based on the IEEE 802.11p wireless network's utilization condition. In this paper, we compare the proposed EVAC-AV control scheme for LRS service with the following: 1) other LRS service schemes that did not adopt the clustering mechanism and 2) other clustering algorithms in terms of different performance metrics to prove EVAC-AV's effectiveness.