大规模混合数据库分层高效访问仿真研究

Research on Hierarchical and Efficient Access Simulation of Large-Scale Hybrid Database

针对传统的大规模混合数据库分层访问方法,普遍存在数据访问丢包率较高、内存耗费较大、访问完成时间较长等问题。提出一种基于遗传算法的混合数据库分层高效访问方法。通过对大规模混合数据库进行分析,利用非对称粒子滤波方法对混合数据库中分层数据进行目标位置估计,实现对混合数据布局,降低混合数据个体对适应度函数的灵敏度,使寻优曲线开始逐渐变平缓,当适应度值进化到最优状态时成为优秀基因,完成混合数据库分层高效访问。实验结果表明,所提出方法数据分层访问丢包率较低、内存耗费较小、完成时间较短。

Traditionally, packet loss rate of data access is high and the memory consumption is large. This paper presented a hierarchical efficient access method for hybrid database based on genetic algorithm. By analyzing the large-scale hybrid database, the asymmetric particle filter method was used to estimate the target position of hierarchical data in hybrid database, so that the hybrid data layout was achieved. Meanwhile, the sensitivity of the mixed data individual to the fitness function was reduced, so that the optimization curve began to flatten gradually. When the fitness value evolved to the optimal state, it became an excellent gene. Thus, the hierarchical and efficient access to the hybrid database was achieved. Simulation results prove that the proposed method has lower packet loss rate, less memory consumption and shorter completion time during the hierarchical access.