基于麻雀搜索算法优化分层粒子群的虚拟机放置

Optimizing Virtual Machine Placement of Hierarchical Particle Swarm Based on the Sparrow Search Algorithm

随着用户和应用程序数量的不断增长,云数据中心对虚拟机的需求也日益增加。虚拟机放置(VMP)作为实现高效资源管理的关键问题,备受关注。本文针对VMP问题提出了一种新的优化模型,综合考虑了放置时间、功率消耗和资源浪费三个目标的最小化。为了优化VMP方案,我们采用了基于麻雀优化分层粒子群算法(SSA-HPSO)。该算法通过对粒子进行层次划分,使得粒子的搜索策略和更新规则针对不同层次和能量水平进行优化。同时,结合麻雀搜索算法,进一步提高了搜索效率和全局搜索能力。这种混合优化策略充分利用了分层粒子群算法的全局搜索和麻雀搜索算法个体之间的协同搜索能力,从而有效地解决了VMP问题。实验结果表明,所提出的基于麻雀搜索算法优化分层粒子群的虚拟机放置算法要优于传统的方法,显著提升了虚拟机放置性能。

As the number of users and applications continues to grow, so does the demand for virtual machines in cloud data centers. Virtual machine placement (VMP), as a key issue to achieve efficient resource management, has attracted much attention. In this paper, we propose a new optimization model for the VMP problem, considering the minimization of three objectives: placement time, power consumption and resource waste. To optimize the VMP scheme, we used a sparrow-based optimization algorithm (SSA-HPSO). The algorithm optimizes the search strategy and updates rules for different levels and energy levels. At the same time, combined with the sparrow search algorithm, further improves search efficiency and global search ability. This hybrid optimization strategy fully utilizes the global search ability of the hierarchical particle swarm algorithm and individual sparrow search algorithm, thus effectively solving the VMP problem. The experimental results show that the proposed algorithm for optimizing hierarchical particle swarm based on the spar-row search algorithm is better than the traditional method and significantly improves the VVC placement performance.