The increasing need, for internet connection speed and reliable communication with delays and a high number of connected devices, has led to the advancement of networks beyond fourth and third generation (Next Generation Cellular Technology). It is essential to distribute resources in these networks to make sure that the available spectrum is used optimally and that service quality is improved. This study introduces an approach using reinforcement learning, for adapting resource distribution in linear multiple access technology enabled B4th/B3rd generation networks. We’ve developed a cutting edge deep Q-learning algorithm that adjusts resource allocation, on the fly based on channel conditions and user needs. Through tuning power control and spectrum allocation strategies our innovative method reduces delays boosts data transfer speeds and enhances efficiency. The results, from our analysis showcase the effectiveness of our approach when compared to resource allocation methods.