Decentralized Cooperative Resource Allocation with Reliability at Four Nines

Abstract

Decentralized cooperative resource allocation schemes for robotic swarms represents an alternative to infrastructure-based communications across different commercial, industrial and environmental protection use cases. The cooperative communication schemes, device sequential and group scheduling in [1], have shown superior performance in comparison to 5G NR sidelink mode 2, but have also shown performance issues due to signaling overhead and signaling induced failures. In this paper we introduce different techniques that reduce the failure probability of data packet transmissions and the packet inter-reception (PIR) time. We evaluate two techniques, respectively, of incremental redundancy using hybrid automatic repeat request and link adaptation by aggregation, as well as their combination for our decentralized cooperative resource allocation schemes and sidelink mode 2. Our results show that the introduced enhancements, allow to double the amount of supported swarm members while achieving four nines reliability when compared to the case where the same enhancements are applied to the sidelink mode 2.