This paper generalizes a new algorithm for Direction of Arrival (DoA) estimation in one and two dimensional (1D and 2D) underwater non-uniform sensor array networks (UWSANs). The extension of augmented matrix approach from the minimum redundancy array (MRA) to the nested array and co-prime array is the main issue addressed in this work. We elaborate numerical examples of how to construct these new structures of the non-uniform array. The performances of classic MUSIC method and the augmented matrix MUSIC approach are also compared. We also analyze the Cramér-Rao bound (CRB) in the non-uniform sensor deployment and it decreases about 10dB comparing to the uniform linear array (ULA). Additionally, the sparse array can exploit more virtual sensors which are able to detect more sources and have a large dynamic range of the signal peaks to the background. Finally, we provide the detailed performance simulations of 1D and 2D nested array and co-prime array respectively.