Congestion-Aware Channel Selection for ETSI ITS-G5
TL;DR
Two congestion-aware mechanisms for ALI selection are introduced that consider not only channel occupancy, which is measured through the interval between transmissions allowed by Decentralized Congestion Control (DCC), but also local queue conditions, enabling more accurate and responsive ALI selection.
Abstract
Access Layer Instance (ALI) selection is essential for ensuring reliable performance in vehicular networks under multi-channel operation. Standardized mechanisms rely on limits to decide whether an ALI can handle traffic coming from a service with a set of requirements. These limits may refer to, e.g., channel utilization, data rate or transmit power. Following a policy, a node selects one of the available ALIs to send traffic through it. This selection can be performed based on using all suitable available ALIs equally (i.e., load balancing) or by turning them on in a sequence of prioritized ALIs that are to be filled sequentially. This letter introduces two congestion-aware mechanisms for ALI selection that consider not only channel occupancy, which is measured through the interval between transmissions allowed by Decentralized Congestion Control (DCC), but also local queue conditions. These measurements are reflected in a metric based on the waiting time before a message can get access to the medium, enabling more accurate and responsive ALI selection. Our results demonstrate that one of our proposed mechanisms, CALIS, significantly improves radio resource usage and application performance; and the other, CASF, improves the performance of the sequential filling mechanism while maintaining the ability to disable radios when traffic load allows it.