Abstract
In this paper, a deep learning approach is used to conduct an in‐depth study and analysis of intelligent resource allocation in wireless communication networks. Firstly, the concepts related to CSCN architecture are discussed and the throughput of small base stations (SBS) in CSCN architecture is analyzed; then, the long short‐term memory network (LSTM) model is used to predict the mobile location of users, and the transmission conditions of users are scored based on two conditions, namely, the mobile location of users and whether the small base stations to which users are connected have their desired cache states, and the small base stations select the transmission. The small base station selects several users with optimal transmission conditions based on the scores; then, the concept of game theory is introduced to model the problem of maximizing network throughput as a multi‐intelligent noncooperative game problem; finally, a deep augmented learning‐based wireless resource allocation algorithm is proposed to enable the small base station to learn autonomously and select channel resources based on the network environment to maximize the network throughput. Simulation results show that the algorithm proposed in this paper leads to a significant improvement in network throughput compared to the traditional random‐access algorithm and the algorithm proposed in the literature. In this paper, we apply it to the fine‐grained resource control problem of user traffic allocation and find that the resource control technique based on the AC framework can obtain a performance very close to the local optimal solution of a matching‐based proportional fair user dual connection algorithm with polynomial‐level computational complexity. The resource allocation and task unloading decision policy optimization is implemented, and at the end of the training process, each intelligent body independently performs resource allocation and task unloading according to the current system state and policy. Finally, the simulation results show that the algorithm can effectively improve the quality of user experience and reduce latency and energy consumption.
Highlights
Wireless communication technology has developed from the first generation of mobile communication technology, which emerged in the 1980s, to the fifth generation of mobile communication technology, from the beginning of satellite communication, radio transmission, and developed into intelligent terminal devices, which makes wireless communication technology able to provide general voice communication or simple data services and is fully integrated into people’s daily life, becoming an indispensable part of today’s society [1]
It is worth noting that the energy efficiency of both the RL-long short-term memory network (LSTM) algorithm and the RL algorithm gradually decreases as time increases
This is because as the RL-LSTM algorithm serves increased users, the energy efficiency of the system decreases for all users connected to the small base station within the coverage area of the small base station for which the transmission condition is best has been completed, and the small base station may have to serve users with longer transmission distances
Summary
Wireless communication technology has developed from the first generation of mobile communication technology, which emerged in the 1980s, to the fifth generation of mobile communication technology, from the beginning of satellite communication, radio transmission, and developed into intelligent terminal devices, which makes wireless communication technology able to provide general voice communication or simple data services and is fully integrated into people’s daily life, becoming an indispensable part of today’s society [1]. With the increasing number of smart terminal devices, the exponential increase in wireless data demand and usage, and the introduction of emerging multimedia applications, it is very difficult to support the rapidly growing data rates and connected devices in the current 4GLTE cellular system. The variable density feature of the network helps to adapt the network to user flows at different densities, ensuring that the throughput of the network is maintained in the right range when the density of user flows increases sharply
Sign-in/Register to view highlights & summary 
Published Version (
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