Variable Cluster Routing for Large-Scale Human-Robot Collaboration Systems — A Top-Down Cluster Routing

Abstract

With the growing development of human-robot collaboration systems, large-scale network topology and node heterogeneity lead to new communication problems. Focusing on these actual requirements, we propose variable cluster routing (VCR) to ensure the communication stability of large-scale human-robot collaboration systems (called large-scale heterogeneous self-organizing networks in the communication field). VCR is characterized by variable window (VW) and a top-down cluster head selection. Each VW includes one or more rounds. In the first round of each VW, base station adjusts the number of rounds (also called the window length) and broadcasts the window length. Then, each node randomly generates its round ID from 1 to the window length and only participates in its round-ID clustering. Therefore, nodes are uniformly assigned to different rounds for transmitting data packets. By adjusting the window length, VCR avoids network congestion and instability effectively. Besides, according to the information of last-round nodes, base station uses a top-down approach to select the present-round cluster heads and then broadcasts cluster head list to accomplish the assignment of cluster heads. This approach ensures network connectivity. We compared other protocols with VCR. VCR outperforms others. Moreover, these results indicate that VCR can extend to large-scale wireless sensor networks, high-density UAV swarms in the military, large-scale human-machine heterogeneous systems, smart open-pit mines, and so on.