Congestion in a network is determined by the resource constraints and the number of deployed sensor nodes. Congestion can significantly degrade the quality of services (QoS) regarding in terms of throughput and end-to-end delay. In this paper, a hybrid bio-inspired algorithm is proposed for congestion control in large-scale Industrial industrial IoTinternet of things (IIoT). First, a competitive Lotka–-Volterra (C-LV) model to avoid congestion is employed, while fairness among sensor nodes is maintained. Second, the dragonfly algorithm (DA) is employed to enhance C-LV by optimizing the parameter for minimizing to minimize end-to-end delay. DA makes this scheme adaptive to change. Simulation The simulation results verify that the proposed scheme improves the QoS in IIoT environments.