An Energy-aware Virtual Machine Placement Algorithm in Cloud Data Center

In order to reduce the energy waste in the data center, while taking into account the violation rate of user service level and the resource utilization rate, this paper studies how to adopt effective strategy to get the reasonable suboptimal solution by combining above three indexes. This paper designs a nonlinear energy consumption model based on polynomials and exponents to measure the energy consumption of different deployment schemes. It is the basis of the deployment strategy. At the heart of this paper, the probability transfer function and the fitness function are designed to optimize the ant colony algorithm. Finally a multi-objective optimization ant colony algorithm based on threshold-dependent pheromone updating was proposed. The algorithm is a kind of distributed optimization method, which is beneficial to parallel computation and has a positive feedback mechanism. The optimal solution can be efficiently obtained by continuous updating of pheromone. The experimental results show that the ant colony algorithm of multi-objective virtual machine placement can achieve the optimal trade-off between multiple conflicting targets, so that the system's energy consumption is less, the violation rate of user service level and the resource utilization rate is also small.