Font Size: a A A

Research On Join Query Optimization Algorithm In Distributed Database

Posted on:2011-04-28Degree:MasterType:Thesis
Country:ChinaCandidate:Y Y NiuFull Text:PDF
GTID:2178330332462629Subject:Computer applications
Abstract/Summary:PDF Full Text Request
In the distributed database system,because the data is distributed and redundancy, which make the network transmission high cost, query processing increases the difficulty and complexity.Therefore, now the efficient join query optimization algorithms become a hotspot problem of distributed database.Based on the research of traditional distributed query processing, take aim at some drawbacks in traditional query optimization,such as optimization standard is single,not flexible,ignored the use of the common query,a new method for distributed query processing is proposed in this paper.Optimized the model of query processing and designed relevant data dictionary. A new function is added into the user module,so that according to users'different demands,they can choose different criterion of query optimization which was set up before.The utilization of commonly used statements is improved by the high utilization rate results recorded in the data query table.This method can reduce the quantity of data transmission.Take aim at the hierarchy,function,cost estimates of query in the distributed database,we studied the semi-join optimization algorithm and algebraic signatures.This paper proposed a semi-join query optimization algorithm with using algebraic signatures.The algorithm queries fast,and dosen't have any data decoding step in the site which received these data nor any additionally caculation,the process is simple.Experimental performance analysis proved that the semi-join query optimization algorithm with using algebraic signatures can reduce the amount of the data transfer between sites and the web communication cost significantly, improved the benefit of optimization.For the premature problem,combine the genetic algorithm with the characteristics of immune system and apply the improved immune genetic algorithm to the database multi-join query optimization.A multi-join query optimization of database based on immunity genetic algorithm is put forward in this paper.This algorithm uses the vector distance between antibodies to measure affinity and concentration.Using elitist and roulette strategy to ensure algorithm can accelerate convergence to the global optimum The use of the immune system, fully guarantee the diversity of population, to avoid the population tends to unity since the late operation, and premature to be trapped in local optimum problem.Simulation results show that,in deal with multi-join query of database,the enchanced algorithm is of better global search ability and higher search speed than simple genetic algorithm and simple immunity genetic algorithm.
Keywords/Search Tags:Distributed DataBase, Query optimization, Data dictionary, Semi-join, Algebraic Signatures, Multi-join, Immunity genetic algorithm
PDF Full Text Request
Related items