The estimation of Direction of Arrival (DOA) is an important area in the signal processing of hydroacoustical array. A sensor network consisting of multiple vector hydrophones is able to make DOA estimation by using the sound pressure and particle velocity of the location for each node, and the coordination cooperation from every node could determine the location of the target sound source. Based on vector hydrophone characteristics, correlation, and DOA algorithm theory, the influence of correlation on the DOA algorithm is analyzed through the simulation. It is able to improve the frequency spectrum of the nodes data and the cross correlation by analyzing and processing the data of actual lake trial. Since each node data contains the serious noise, as a result, it causes the frequency spectrum situation to become very bad. After data signal processing, the requirements of DOA algorithm on data correlation are met, and the broadband noise interference is able to be effectively inhibited. It is indicated that after the actual data processing, the line spectrum signal of node stands out and the cross correlation of various nodes has been improved. The practicality of the study is proved.