Development of Hybrid Deep Learning Frameworks for Dynamic Resource Allocation in mmWave 6G Networks to Optimize Quality of Service Parameters

The research had developed an LSTM-DQN hybrid Deep Learning model with the intention of improving Quality of Service (QoS) through increased Throughput, decreased Latency, and Energy Savings in the Resource Allocation for Dynamic Networks utilizing mmWave technology in 6G. The Hybrid model had been trained utilizing synthesized data to mimic different Network Scenarios such as Urban/Rural Environments. Results from the Hybrid model showed better performance than traditional Rule-Based models and Other Models, along with the ability to Learn/Adapt Over Time. Confidence Intervals and Feature Analysis were utilized to help determine how the Hybrid Model functioned, and its Reliability. In Conclusion, the Research demonstrated the Power of Deep Learning to effectively Manage Network Resources, thus enabling communications to be Faster, More Reliable, and capable of supporting the High Demand of Future 6G Networks.