METHOD OF OPTIMIZING PLANNING AND CONTROL PROCESSES IN MOBILE COMMUNICATION SYSTEMS

DOI: 10.31673/2412-4338.2025.048921

Abstract

The paper addresses the problem of optimizing planning and control processes of radio resources in fifthgeneration mobile communication systems and prospective 6G networks under conditions of increasing traffic load, service diversity, and stringent quality requirements. It is shown that classical scheduling algorithms mainly focused on QoS metrics and throughput maximization are unable to ensure stable performance of real-time services under high load.

A QoE-oriented method for radio resource scheduling optimization is proposed, based on an integrated QoS/QoE criterion that combines heuristic resource allocation with subsequent gradient-based refinement. A mathematical model of radio resources and traffic load is developed, taking into account physical limitations of the radio channel, traffic statistics, and service requirements of different user classes. For the first time at the scheduling algorithm level, the utility function is directly associated with Quality of Experience indicators such as delay, jitter, and packet loss.

The potential of integrating machine learning methods for traffic load and QoS/QoE parameter prediction is investigated to enable predictive radio resource management. Simulation results demonstrate that the proposed method provides higher QoE levels, better delay control, and improved fairness compared to Max C/I and Proportional Fair algorithms, especially under high-load conditions. The obtained results confirm the feasibility of QoE-oriented approaches for modern and future mobile networks and their applicability to network slicing and Self-Optimizing Networks.

Keywords: 5G/6G mobile networks, radio resource scheduling, optimization, Quality of Service (QoS), Quality of Experience (QoE), proportional fair scheduler, delay, jitter, packet loss, machine learning, traffic load prediction, selfoptimizing networks, network slicing