Combined real-valued subspace based two dimensional angle estimation using L-shaped array

Two dimensional (2D) angle estimation using L-shaped array is studied, and a combined real-valued subspace based method is proposed. Firstly, the structure of L-shaped array is modified through some translations of the two subarrays, where the subarray along Z-axis can be extended based on the cross correlation matrix. By constructing the relationship between the two subarrays from two real-valued correlation matrices, angle estimation function based on combined subspace can be established. Finally, two dimensional angle can be estimated via one-dimensional polynomial root finding and the constructed relationship, respectively. The proposed method has low complexity due to the real-valued decompositions and low-dimensional correlation matrices, and it achieves better angle estimation results and higher angular resolutions than other state-of-the-art methods using L-shaped array. Simulation results verify the effectiveness of the algorithm.