MATLAB-Based Simulation Performance Evaluation of Network Slicing in the Fifth Generation (5G)

This study outlines a simulation model developed using MATLAB that assesses the performance of 5G network slicing for static and adaptive resource allocation.The model simulates a standalone 5G base station working within a 100-MHzwide mid-band channel while supporting three types of services, including enhanced mobile broadband, ultra-reliable low-latency communication (URLLC), and massive machine-type communication.A backlog-dependent adaptive bandwidth allocation policy using stochastic Dirichlet weighting is proposed and compared to a classical static slicing algorithm in terms of performance measured through throughput, delay, jitter, packet loss rate, reliability, and spectral efficiency.The performance of the algorithms is analyzed under varying loads.The accuracy of the simulation is validated through comparison with 3GPP Release 17 target performance indicators, communication theoretic capacity assessment, and queuing theory analysis, as well as through comparing the results with other research.User load scalability experiments are conducted.It can be concluded from the results that the method of dynamic slicing allows increasing the efficiency of the use of the spectrum and obtaining an additional 10-12% of throughput during moderate loads, while reaching a spectral efficiency of up to 2.6 bps/Hz.However, an increase in the variability of latency and packet loss under high-load conditions is observed.Static slicing ensures more predictable behavior with low latency and minimal packet losses for URLLC services.The described approach is a reliable and computationally inexpensive solution to analyze the trade-off between efficiency and reliability of 5G network slicing.