Abstract: Aiming at the problem that conventional cloud computing resource prediction algorithm could not predict accurately in the abnormal network environment, the author proposed a scheduling strategy based on improved ant colony algorithm. The strategy integrated the concept of the number of information, which could not only balance load quickly, but also ensure users’ needs under multiple conditions, reduce energy consumption and improve the performance of cloud computing. The experimental results show that the improved ant colony scheduling algorithm improves the utilization of cloud computing resources, reduces energy consumption and increases task capacity for every single node, and greatly improves the performance of cloud computing and service quality.

Key words: cloud computing, ant colony algorithm, resource scheduling, abnormal network