Abstract: A computerized ionospheric tomography algorithm based on truncated singular value decomposition (TSVD) regularization is put forward for minimizing the impact of the background ionosphere model. Spherical harmonic functions and empirical orthogonal functions are utilized as the basic function to represent the spatial variation of the ionosphere. The generalized cross validation (GCV) method is used to determine the proper truncated parameter, which helps to enhance the stability and reliability of the TSVD regularization method. The comparison between multiplicative algebraic reconstruction technique (MART) and the TSVD regularization method is also made in the ionospheric tomography simulation experiment in China with 23 observations. Results verify the validity of the proposed algorithm without the background ionospheric electron density as the prior condition.