One of the key aspects of 5G networks is the implementation of massive MIMO (Multiple Input Multiple Output) technology combined with adaptive beamforming. This study explores the use of a linear array antenna to manage and reduce unwanted signals such as jamming, interference, and noise, while also boosting the signal strength towards the intended user or device. The main challenge lay in optimizing the weights of the antenna elements, which was tackled by employing adaptive algorithms like LCMV (Linearly Constrained Minimum Variance) and RLS (Recursive Least Squares). To simplify the optimization process, two soft computing techniques—Particle Swarm Optimization (PSO) and Genetic Algorithm (GA)—were utilized. The performance of the beamforming weights and radiation patterns was assessed in terms of minimizing unwanted signals and maximizing the desired signal. To check how well the proposed methods work, some commonly used algorithms like MVDR (Minimum Variance Distortionless Response) and LCMV are also applied. The outcomes were compared to those from other algorithms. A Differential Beamforming method is applied to examine how effectively the system can focus the signal in the target direction while minimizing unwanted interference from other directions. Additionally, the fminsearch algorithm, which is a basic local search method, is used to compare how well it can adjust the beamforming weights compared to the more advanced global optimization techniques. The results indicate that PSO and GA produce highly similar performance levels.