Cooperative Resource Allocation for Proximity Communication in Robotic Swarms in an Indoor Factory

Abstract

Robotic swarms are becoming relevant across different industries. In an indoor factory, collective perception of the environment can be used for increased factory automatization. It requires reliable, high throughput and low latency communication of broadcasted video data among robots within proximity.We introduce two new decentralized resource allocation schemes that meet these stringent requirements. The two proposed decentralized schemes are denoted as: (i) device sequential, where robots take turns to allocate resources, and (ii) group scheduling, where robots select local group leaders who perform the resource allocation. A comparative evaluation is performed by simulation against a centralized resource allocation scheme and the current 3GPP release 16 NR sidelink mode 2 scheme.Our results show that the two proposed decentralized resource allocation schemes outperform sidelink mode 2 due to the mitigation of the half-duplex problem. The proposed schemes reach the throughput target of 10 Mbps with a reliability of 99.99% for a swarm size of 50 robots.

Evaluation of NR-Sidelink for Cooperative Industrial AGVs

“…Additionally, authors in [9] and [10] address features such as decentralized (aperiodic) traffic support over NR-sidelink and the introduction of reevaluation strategies for V2X communication. [11] Presents an approach for resource allocation in randomly distributed robotic swarms for proximity communication in an indoor factory environment. The authors propose a cooperative approach to exchanging inter-UE coordination (IUC) messages to maintain an efficient resource allocation algorithm and reliable communication.…”