This research delves into an AI-based approach to enhance path loss (PL) modeling in wireless communication systems, particularly vital for advanced technologies like 4G/5G and TETRA (Terrestrial Trunked Radio). Path loss is the reduction in power density (attenuation) of a wireless transmission and wave as it propagates through a medium over a long distance. Conventional methods, such as the Okumura-Hata, Extended Sakagami, Egli, and Free Space models, face limitations, necessitating the exploration of accurate models. Through deep learning and regression analysis, the study aims to develop a comprehensive PL model, considering factors like frequency, antenna height, and terrain. The research addresses the fundamental challenge of characterizing PL in 4G/5G and TETRA systems, highlighting the need for AI strategies to enhance precision in modeling signal attenuation. Overall, this thesis makes significant contributions by developing AI-Based Pathloss Characterization Models, providing a historical overview of Artificial Neural Networks (ANN) and Support Vector Regression (SVR). The study explores typical structures and applications of both ANN and SVR, utilizing custom-created MATLAB codes for accurate pathloss characterization in future mobile network links. Integration of machine learning algorithms, particularly using MATLAB's learner regression application, enhances prediction accuracy for robust and efficient characterization in diverse scenarios. Additionally, empirical path loss models are implemented using MATLAB code, offering a comprehensive understanding of path loss behavior in varying environments. Result visualization and comparative analysis in Excel improve the interpretability of path loss predictions. The research contributes to future mobile network planning, offering insights for optimizing signal propagation estimations, aiding in efficient network design and performance across diverse operational scenarios. This work recognizes the pivotal role of accurate path loss models in shaping the performance, capacity, and efficiency of wireless communication systems, aligning with the evolving demands of present and future wireless communication technologies. This thesis accentuates the significance of fine-tuning path loss prediction models for TETRA and GSM systems, recognizing their sensitivity to dataset variations. TETRA's AI model demonstrates commendable performance (RMSE: 0.00 to 10.898 dB), alongside both AI and modified empirical models (RMSE: 7.386 to 16.120). GSM exhibits superior predictive ability in all models, leveraging AI (RMSE: 0.004 to 2.572 dB) and modified empirical models (RMSE: 2.544 to 19.799 dB). The study underscores the location-specific nature of real-world data models, showcasing TETRA's need for adaptations and GSM's requirement for adjustments in diverse scenarios. Emphasizing the critical role of model fine-tuning in wireless communication systems, the results provide a comprehensive understanding utilizing RMSE, MAE, and MSE metrics.