Cooperative perception has been shown an effective approach to address the perception limitation problem faced by the individual perception in the era of autonomous driving. Considering a two-source two-hop cooperative perception vehicular network, we investigate the average age of information (AoI) with different automatic repeat-request (ARQ) strategies in this paper. We formulate the packet transmission process as a Markov chain, which allows us to establish linear equations about the first and second moments of all states’ residual waiting time for the destination to receive a new update packet. Then, we derive the average AoI based on the analytical solution to the equations. Furthermore, the analysis is applied to a special case of the single-source two-hop cooperative perception system. Simulation results show that the optimal strategy is neither non-ARQ nor infinite-ARQ, but truncated ARQ, which means that there exists a finite retransmission limit to minimize the average AoI, implying that better cooperative perception can be achieved by optimizing the retransmission limit, especially in the single-source two-hop situation.
AoI Analysis for Automatic Repeat-Request in Vehicular Cooperative Perception Networks
07.10.2024
620992 byte
Aufsatz (Konferenz)
Elektronische Ressource
Englisch
Cooperative perception in vehicular networks using multi-agent reinforcement learning
BASE | 2021
|Perception-Connection Tradeoff for Radar Cooperative Sensing in Multi-hop Vehicular Networks
Springer Verlag | 2022
|