This paper proposes LTE radio scheduler analytical models. The motivation behind the models is to evaluate the performance of the complex LTE system analytically, rather than using simulations or real life experiments (which are not always possible to perform) that are too exhaustive and consumes large computation power and time. The models are developed to approximate the system behavior mathematically; it enhances the performance evaluation's processing effort and time significantly compared to traditional simulations. The proposed analytical models are based on the Continuous Time Markov Chain (CTMC) and represent different LTE Time Domain Schedulers (TDS) as well as two different model categories: a single class model and a two class model. In the single class model only a single traffic class is considered with no QoS differentiation. Whereas, the two classes model supports QoS differentiation between the two classes. An extensive analysis have been performed using the aforementioned analytical models and is compared with a well developed simulation model.