Research Article
Machine Learning Enhanced CPU-GPU Simulation Platform for 5G System
@INPROCEEDINGS{10.1007/978-3-030-94763-7_3, author={Yuling Ouyang and Caiyuan Yin and Ting Zhou and Yan Jin}, title={Machine Learning Enhanced CPU-GPU Simulation Platform for 5G System}, proceedings={Mobile Networks and Management. 11th EAI International Conference, MONAMI 2021, Virtual Event, October 27-29, 2021, Proceedings}, proceedings_a={MONAMI}, year={2022}, month={1}, keywords={CPU GPU System-level simulation Logistic regression enhanced Mobile Broadband (eMBB)}, doi={10.1007/978-3-030-94763-7_3} }- Yuling Ouyang
Caiyuan Yin
Ting Zhou
Yan Jin
Year: 2022
Machine Learning Enhanced CPU-GPU Simulation Platform for 5G System
MONAMI
Springer
DOI: 10.1007/978-3-030-94763-7_3
Abstract
The exponential growth of mobile terminals and the explosion of data volume are promoting the continuous evolution of mobile communication network and also increasing the complexity of the system. Meanwhile, 5G system-level simulation also requires more complex operations and more data processing. Conventional system simulation platform based on CPU can not satisfy the computing power requirement of system-level simulation of 5G. For tremendously shorten the execution time, we proposed to develop the CPU-GPU based parallelization platform, which adopts Logistic Regression algorithm to optimizing the use of computational resources. Numerical results demonstrate the effectiveness in terms of reducing execution time and guaranteeing reliability of system-level simulation result in 5G scenarios.
