A new two dimensional direction of arrival estimation algorithm based on multiple symmetric virtual transformation is proposed. Multiple symmetric virtual transformation is used to get several virtual arrays which are symmetric to the original array

and further construct translation invariant virtual-real sub-arrays. The rotation invariant factor can be used to get the pitch angle of signal source

from which the corresponding azimuth angle can be obtained by MUSIC one dimensional search

and the azimuth and pitch angle are paired automatically. By using the cumulative data of each virtual transformation array

the positive and negative error of each virtual array data can be eliminated

which can significantly reduce the data error between the virtual array and the real array

hence the estimated precision is obviously improved. Simulation experiments analyze on root mean square errors（ RMSE）and performance of success rate of the proposed algorithm in case of variation of signal to noise ratio（ SNR） and snapshot

the results show the effectiveness of the proposed algorithm.