A Hybrid Heuristic Algorithm for Disaster Backup in Data Center Networks

Backup path selection optimization plays an important role in disaster backup problems, which can improve bandwidth utilization and load balancing in data center networks. However, during the optimization process, most path selection algorithms either fail to fully utilize the network transmission capability, or tend to trap into the local optima when the number of data center nodes and the amount of backup data increase. In this paper, we propose a hybrid heuristic algorithm that combines particle swarm optimization (PSO) algorithm and ant colony optimization (ACO) algorithm for backup path selection in rapid disaster backup problems. We introduce a sigmoid function to integrate the pheromone information of ACO and the velocity information of PSO, and the resulting algorithm is capable of avoiding premature convergence and making full use of network transmission capability. Simulation results show that our proposed algorithm efficiently reduces backup completion time and improves load balancing while considering multi-constraint in data center networks.