Abstract
The channel width for Intelligent Transportation System (ITS) band for vehicle-to-everything (V2X) system is expected to be relatively narrow. It implies that effective congestion control mechanisms will be essential. A well-established standard mechanism is the Cooperative Awareness Message (CAM) generation rule that drops unnecessary periodic beacons. Dropping packets from a periodic packet stream produces a non-periodic traffic. It renders the distributed Semi-Persistent Scheduling (SPS) algorithm exposed to resource waste and packet collision problems. The problems can be tackled by supplementing SPS with machine learning (ML) that predicts the transmission times of non-dropped CAM packets and reserves a transmit resource at the predicted time. The proposed approach is shown to significantly extend the distance up to which the target packet reception ratio (PRR) is met, compared to the current SPS algorithm.
| Original language | English |
|---|---|
| Title of host publication | 2023 IEEE 97th Vehicular Technology Conference, VTC 2023-Spring - Proceedings |
| Publisher | Institute of Electrical and Electronics Engineers Inc. |
| ISBN (Electronic) | 9798350311143 |
| DOIs | |
| Publication status | Published - 2023 |
| Event | 97th IEEE Vehicular Technology Conference, VTC 2023-Spring - Florence, Italy Duration: 2023 Jun 20 → 2023 Jun 23 |
Publication series
| Name | IEEE Vehicular Technology Conference |
|---|---|
| Volume | 2023-June |
| ISSN (Print) | 1550-2252 |
Conference
| Conference | 97th IEEE Vehicular Technology Conference, VTC 2023-Spring |
|---|---|
| Country/Territory | Italy |
| City | Florence |
| Period | 23/6/20 → 23/6/23 |
Bibliographical note
Publisher Copyright:© 2023 IEEE.
Keywords
- V2X
- congestion control
- deep learning
- packet reception ratio (PRR)
- prediction
- quasi-periodic traffic
ASJC Scopus subject areas
- Computer Science Applications
- Electrical and Electronic Engineering
- Applied Mathematics