OPTIMIZING REMOTE VEHICULAR CONTROL THROUGH LOW-LATENCY VIDEO STREAMING OVER CELLULAR NETWORKS

Abstract

Ground vehicles and autonomous systems that can be controlled remotely are now common in transportation, security, farming, and industry because live video streaming helps achieve higher control and knowledge of events. When communication between parts of the system is reliable and very quick, operation is much safer. Regular vehicular networks based on Wi-Fi can be challenging because they typically run into limited coverage, many handoffs, and occasional delays when being used in the outdoors or while a vehicle is moving. They cause the video to freeze and make it take longer for the robot to respond to commands. This paper explores how video streaming can be made low-latency using cellular networks 4G LTE and 5G New Radio (NR) with Multi-access Edge Computing (MEC) for remote vehicle control. A testbed and a way to simulate various elements are created to study latency linked to video encoding, network transfer, decoding, and edge processing. Proper adjustments are investigated for each codec and network setting to decrease delays. The experiments in this paper confirmed that cellular systems with MEC are capable of low end-to-end latency, below 50 milliseconds, as opposed to Wi-Fi systems that, on average, have latency above 90 milliseconds. Handoff within cellular networks is seamless, while priority is set for important traffic, controlling major changes in system delay. By having MEC, devices can process data at the edges, which lowers delays. Using this procedure, remote control of ground vehicles can be secure, responsive, and of high quality even in changing conditions.