Due to the inherent complexities of navigation, localization is one of the key fundamental elements required for robotic navigation. In this thesis, three different methods of localization including relative methods, absolute methods and probabilistic methods are represented in the first section. These methods are discussed in detail in the second section. Afterwards, in the third section, Bayes rule, as the basis of probabilistic methods, and Markov model will be introduced including their mathematical equations, then we demonstrate Particle Filter as the subject of this thesis. We prove that using odometry data alone to localize a robot is unsatisfactory due to the accumulation of errors, and therefore does not produce the required data input on its own. In the fourth section algorithm that's used to implement this particle filter is introduced. Finally, in the fifth section we will compare Kamlan Filter output with Particle Filter output for mobile robot localization and we will demonstrate that Particle Filter produces more accurate localization. Also, changes in Maximum Particle and Minimum Particle parameters will be evaluated in detail and will demonstrate how the Particle Filter improves mobile robot localization accuracy.