基于并行的SDD-1算法的改进  

Improvement of SDD-1 Algorithm Based on Parallel

作　　者：谢旭升[1] 陈复兴[1] 

机构地区：[1]江西师范大学计算机信息工程学院,江西南昌330022

出　　处：《山西大学学报（自然科学版）》2013年第3期338-343,共6页Journal of Shanxi University(Natural Science Edition)

摘　　要：传统分布式数据库查询SDD-1算法是基于爬山(Hill Climbing)算法而形成的,在执行时没有考虑可并行选择,在元组数据剧增的情况下,该算法的复杂度呈指数增长,这将无意义于分布式数据库系统.针对这一问题提出了一种基于并行的改进的SDD-1算法.The traditional SDD-1 query algorithm is based on the mountain climbing algorithm（Hill Climbing）.As it did not consider parallel selection while execution,cost of query will increase rapidly when the tuple data is huge,and the complexity of the algorithm is very high.We proposed a kind of advanced SDD-1 algorithm based on parallel to solve this problem.

关 键 词：分布式数据库 SDD-1查询算法 并行 

分 类 号：TP311.13[自动化与计算机技术—计算机软件与理论]