Distributed query optimization research is under way in universities, research laboratories and similar establishments. It is a very active research area in distributed database system. There is difference between centralized database and distributed database, which distributed database needs data transportation. So most research of distributed query optimization aim at reducing communication cost. The distributed query optimization problem is the key point in distributed database system, and it is an NP hard one.The paper focuses primarily on distributed query optimization. In distributed query optimization, this paper aims at how to reduce communication cost, adopting semijoin program. This paper gives emphasis to the research SDD-1 optimization algorithm applying in the distributed query, and brings forward two improved algorithms on the basis of the research on the SDD-1 algorithm. One is based on the multiple relations semijoin, which can be applied to the circumstance in which the cache of DDBS is the final assembly site of temporary query result; the other is based on virtual join, which reduce the cost of both communication and local processing, and it is adaptive to different values of selectivity. Experiment results showed the two improved algorithms are efficient. |