Resource utilization and cost optimization oriented container placement for edge computing in industrial internet

With the continuous evolution of the modern industrial internet in an intelligent direction, intelligent devices generate many delay-sensitive task requests. Container-based edge computing service deployment can save the bandwidth resources of the core network and reduce service delay. However, an unreasonable container deployment leads to the waste of edge server resources and fails to meet the requirements of real-time processing of services. In this paper, we establish a basic container deployment model to optimize the resource utilization and deployment cost. On this basis, we additionally establish a fault-tolerant deployment model for containers, which enables the edge computing system still provide services and improves the fault-tolerant ability of the factory when the container deployment fails or deployment speed drops caused by hardware failures. To solve the optimal deployment strategy, we propose an improved genetic-simulated annealing algorithm (IGSAA). The proposed algorithm can achieve optimal container deployment by improving initialization, crossover and mutation operations of the genetic algorithm. The simulation results show that the established model has remarkable effect in resource utilization and cost optimization. Compared with the existing deployment algorithms, IGSAA outperforms them by at least 22% in optimizing resource utilization and deployment costs.