An Optimization for Distributed Database Multi-join Query Based on Improved Genetic Algorithm

The paper considers the optimization of the multi-join query process of distributed database, which is mainly based on the computer time or resource cost and search strategy of querying database. To model the cost, the multi-join query is divided from a full-join query into some semi-join queries steps to reduce the communication cost and then a new cost model is presented. As for the search strategy, by comparing the advantages and disadvantages of dynamic programming algorithm (DPA), simulated annealing algorithm (SAA) and iterative correction algorithm (ICA), genetic algorithm (GA) is improved to optimize the query process because of its better search performance. To implement such an optimization, the paper deals with GA’s process of coding, selection, crossover and mutation carefully in novel way so as to improve the accuracy and speed of searching for optimal values of querying a distributed data base. Finally, the simulation results show that the optimization of cost model and search strategy achieves the expected effect and it can improve the accuracy and speed of the distributed database query.