Fair Decentralized Congestion and Awareness Control for Vehicular Networks

Abstract

Challenges surrounding 802.11p vehicular networks have become more apparent in recent years within the research community. The exchange of periodical beacons on the control channel is mandatory to support higher layer applications such as active safety applications. However, the nature of such wireless vehicular networks imposes a lot of serious problems, especially the risk that the radio channel becomes quickly saturated. Thus, congestion control and awareness control are crucial mechanisms to control the load and ensure that each vehicle is able to communicate with its most relevant neighbours respectively. In this paper, we tackle jointly congestion and awareness control by taking into consideration 2-hops influence, hidden-terminals, SINR, received power, and a fair bandwidth indicator measuring the performance distribution among a coalition of vehicles in a highly volatile vehicular network. A two-steps proactive distributed algorithm called TPApIAB is proposed to improve the fair bandwidth utilization. Each step of the algorithm achieves a distinct goal: Intelligent Adaptive Beaconing (IAB) is in charge of controlling the congestion while the Transmit Power Adaptation (TPA) is in charge of controlling which neighbour should be part of a coalition with respect to down-link channel qualities. Different from awareness control algorithms which assume an optimal channel model, here a more realistic model is used to estimate dynamically the best communication range based on a non-deterministic propagation model. Realistic simulations with NS-3 show that the solution presented outperforms the well-known D-FPAV algorithm in non-homogeneous graphs.