Non-Negotiated Implicit ETSI VAM Clustering

TL;DR

The paper tackles the problem of VRU awareness in C-ITS under medium-capacity constraints by proposing a non-negotiated implicit VAM clustering scheme. It shows that clustering can be achieved purely through modifications to regular VAMs, avoiding additional negotiation messages and maintaining resilience in lossy channels. The evaluation demonstrates that implicit clustering preserves VRU awareness while reducing message generations and stabilizing reliability metrics (IPG/IGG) compared with ETSI clustering and standalone operation, even under vehicle traffic. This approach offers a scalable, low-overhead alternative for VRU clustering with practical implications for real-world V2X deployments. Future work will explore integrating VRU intention sharing and evaluating more VRU types and densities.

Abstract

Including Vulnerable Road User (VRU) in Cooperative Intelligent Transport Systems (C-ITS) framework aims to increase road safety. However, this approach implies a massive increase of network nodes and thus is vulnerable to medium capacity issues, e.g., contention, congestion, resource scheduling. Implementing cluster schemes -- to reduce the number of nodes but represent the same number of VRUs -- is a direct way to address the issue. One of them is suggested by European Telecommunications Standards Institute (ETSI) and consists of nodes (connected pedestrians and cyclists) sending vicarious messages to enable a leader node to cover for a cluster of VRUs. However, the proposed scheme includes negotiation to establish a cluster, and in-cluster communication to maintain it, requiring extra messages of variable sizes and thus does not fully resolve the original medium capacity issues. Furthermore, these exchanges assume network reliability (i.e. a lossless channel and low latency to meet time constraints). We propose a method for VRU Awareness Message (VAM) clustering where 1) all cluster operations are performed without negotiation, 2) cluster leaders do not require sending additional messages or meet deadlines, and 3) assumes a lossy communication channel and offers a mechanism for cluster resilience. Our results show the feasibility of the concept by halving message generations compared to individual messages while keeping the awareness levels (i.e., that VRUs are accounted for).

Figures (6)

• Figure 1: VAM generation conditions
• Figure 2: State diagram for ETSI VAM cluster operations
• Figure 3: State diagram for Implicit Clustering
• Figure 4: Generated VAMs per station
• Figure 5: Inter-packet Gaps with VRU and vehicular traffic