Let's talk in groups: A distributed bursting scheme for cluster-based vehicular applications

Clustering has been an important concept in the scope of vehicular networks. The idea is to reduce channel contention, enable building backbones, and improve spatial reuse. We propose a seemingly simple, yet unexplored idea: extending IEEE 802.11 frame bursting MAC access to multiple stations aggregated into a cluster. We call this approach Distributed EDCA Bursting (DEB). The focus of our work is not on building clusters, but exploring what is the gain that can be achieved by the standard IEEE 802.11p channel access if we introduce the principle of frame bursting (presently not allowed in IEEE 802.11p standard, but the key factor for the efficiency of IEEE 802.11n/ac WiFi channel access). The fundamental idea is to extend the standard frame bursting mechanism so that only the cluster head contends for the channel. Upon winning the contention, the reserved transmission opportunity is shared among the vehicles in the cluster, which transmit a coordinated burst of frames. We describe the idea in detail, particularly analyzing theoretical benefits and limitations of the approach. We implement DEB in the platooning simulation framework Plexe-Veins and analyze its behavior under different loads (i.e., number of vehicles sharing the channel), transmit power policies, and different MAC and physical layer parameters. We present benefits and drawbacks of our approach, showing that it can overcome channel limits in saturation conditions, reduce channel usage at moderate loads, and improve spatial reuse. Moreover, by coordinating and distributing channel access, DEB reduces the number of collisions at the expenses of a slightly increased average beacon inter-arrival time.