Abstract
With the explosive growth in demand for mobile traffic, one of the promising solutions is to offload cellular traffic to small base stations for better system efficiency. Due to increasing system complexity, network operators are facing severe challenges and looking for machine learning-based solutions. In this work, we propose an energy-aware mobile traffic offloading scheme in the heterogeneous network jointly apply deep Q network (DQN) decision making and advanced traffic demand forecasting. The base station control model is trained and verified on an open dataset from a major telecom operator. The performance evaluation shows that DQN-based methods outperform others at all levels of mobile traffic demand. Also, the advantage of accurate traffic prediction is more significant under higher traffic demand.
Highlights
The demand for mobile data traffic has been rapidly growing over recent years and continuing at an even faster pace
Due to the nature of discrete actions in this work, we explore the potential of deep Q network (DQN) to enhance network management tasks and find that it is capable of providing satisfactory results
The cloud controller dynamically adjusts the number of active small base stations (BSs) in a macrocell every M data records, so the discrete decision time index t is defined with decision period Td = M · Tr
Summary
The demand for mobile data traffic has been rapidly growing over recent years and continuing at an even faster pace. Both optimization theory and reinforcement learning (RL) based solutions have been actively proposed [10]–[15]. We propose an energy-aware mobile traffic offloading scheme in HetNet jointly apply deep Q network (DQN) based decision and traffic demand forecasting. The proposed network management scheme offloads the transmitting data from macro to small BSs, considering traffic and performance provisioning for improved energy efficiency. SYSTEM MODEL This section describes the environment as well as the system architecture of mobile traffic offloading in HetNet. Figure 1 presents the environment where a cloud controller connects to multiple macro and low-power small BSs forming macro and small cells. The proposed architecture predicts the mobile data traffic in advance and decides the number of active small BSs considering energy efficiency
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
