Snow attenuation depends on many factors which are hard to observe and identify or classify. Modeling of snow attenuation is relatively complex. There are two main classes of methods used in snow attenuation prediction: the empirical method and the physical method. Physical method which we used in this work focuses on reproducing the physical behavior of factors involved in the process. The attenuation in the frequencies of mobile communication due to snow is simulated using Discrete Propagation Model. For this modeling, certain ice-crystal categories are chosen to be investigated. Needles, plates and branches are the main 3 groups which are focused, and 13 different models of snow particles in total are chosen to represent snow. The element in each group is chosen according to similar physical characteristics of ice crystal. It was found that attenuation due to snow is higher than rain attenuation specifically due to differences in particle size. In our simulations, frequencies of GSM communication, 900MHz, 1800MHz and 2270MHz, are used for calculation of attenuation.