Optimization of resource allocations in 5G mobile network using Active Reward Learning

This paper introduces a real-time dynamic resource allocation method for 5G mobile network slicing, leveraging active reward learning to enhance network performance and Quality of Experience (QoE). Unlike traditional static or reactive approaches, our method proactively predicts future resource availability and intelligently prioritizes requests based on urgency and importance. We employ a sophisticated active reward function that incorporates key network parameters, including Probability of Connectivity, Spectrum Efficiency, Sub-channel Occupancy Ratio, Packet Loss Ratio, and Packet Delay. This function dynamically adjusts parameter weights based on network conditions and real-time traffic patterns, ensuring efficient resource utilization. Furthermore, we extend this approach to Intelligent Transportation Systems (ITS) for traffic light control. Simulation results demonstrate that our proposed method achieves a 15% reduction in average packet delay and a 10% improvement in spectrum efficiency in 5G network slicing compared to traditional methods. In the ITS application, we observe a 20% decrease in average vehicle waiting time and a 5% increase in traffic throughput. These results highlight the effectiveness of our approach in enhancing network performance and responsiveness in dynamic environments.