This chapter describes the proposed methods to extract parametric features at sub-segmental, segmental and supra-segmental levels to capture the language-specific excitation source information. In this work, glottal pulse, spectral and epoch parameters are used for representing sub-segmental, segmental and supra-segmental information present in excitation source signal. Further, these individual features are combined at score level to enhance the accuracy of LID systems by exploiting the non-overlapping information present among the features.