It is demonstrated in this paper that the searching step τ in the TV-Tikhonov algorithm plays an important role for determining the original layer structure directly from measured high-resolution depth profiling data. The optimization of τ is very necessary for obtaining a reliable deconvoluted result. Considering a large increase of simulated data points by decreasing the depth interval for quantification of high-resolution depth profiling data, the optimization of τ is performed simply by one-dimensional search algorithm. Furthermore, the noise effects from “measured” depth profiling data on the deconvoluted profile with different optimal τ values are also quantitatively evaluated. Finally, as examples, the measured SIMS depth profiling data of the two In nano-layers embedded in InGaAs thin films and the 28Si/30Si isotope nano-superlattices are deconvoluted directly and the respective original layer structures are obtained accordingly.