Design Load-Balanced Energy-Efficient Network Algorithms for 5th Generation Ultra-Dense Heterogeneous Networks

5G allows the rapid growth in User Equipment (UEs) various applications, which lead to an exponential increase in data traffic and result in massive overloading and huge imbalance across co-tier and inter-tier networks in Ultra-Dense Heterogeneous Networks (Ud-HetNet). In previous cellular networks, the power saving mechanism, Discontinuous Reception/Transmission (DRX/DTX), improves the UE’s battery efficiency. However, the power saving mechanism shows inefficiency with 5G advanced technologies such as Beam alignment, which is needed to be addressed. In this project, we address the load balancing and energy efficiency of Ud-HetNet and UEs, respectively, while guaranteeing the Quality of Service (QoS) of UEs. In order to address the high diversity, uncertain environment, real-time dynamic decision, and complexity of the network architecture, our designed solutions are based on Reinforcement Learning (RL). The preliminary results show that these problems deserve to be studied further.

Project ID:PB10907-3848
External Project ID:MOST109-2221-E182-038-MY3