Optimized beamforming and network slicing for dense urban 5G deployments

Abstract

Optimizing beamforming and network slicing is critical for enhancing spectral efficiency, energy efficiency, and resource distribution fairness in dense urban 5G networks. This paper proposes a hybrid genetic algorithm particle swarm optimization (GA-PSO) method to jointly optimize beamforming weights, bandwidth allocation, and power distribution, balancing computational efficiency with near optimal performance. The hybrid approach uses GA for global exploration and PSO for fast convergence, overcoming the limitations of standalone heuristic and exact optimization methods. Simulation experiments in a dense urban 5G network with massive MIMO base stations show that proposed method achieves up to 15% higher spectral efficiency and 18% better energy efficiency compared to existing integer linear programming (ILP) method, while significantly reducing computational complexity. Convergence analysis further confirms that the hybrid method requires fewer iterations to reach near-optimal solutions, making it suitable for real-time 5G resource management. Additionally, fairness evaluation using Jain’s Index shows that proposed method ensures more equitable resource distribution than conventional methods. These results establish hybrid GA-PSO method as an effective and scalable solution for next- generation wireless networks.