A Novel Link-to-System Mapping Technique Based on Machine Learning for 5G/IoT Wireless Networks

Area of Science:

• Telecommunications Engineering
• Computer Science
• Machine Learning

Background:

• Accurate simulation of wireless systems requires effective inter-connection between link-level simulators (LLS) and system-level simulators (SLS).
• Existing link-to-system (L2S) mapping methods may lack prediction accuracy or efficiency for complex 5G New Radio (NR) systems.

Purpose of the Study:

• To propose and validate a novel machine learning (ML) based link-to-system (L2S) mapping technique for 5G NR systems.
• To enhance the prediction accuracy and reduce the processing time of L2S mapping.

Main Methods:

• Developed a machine learning (ML) based exponential effective signal-to-noise ratio (SNR) mapping (EESM) method utilizing a deep neural network (DNN) regression algorithm.
• Integrated and validated the proposed technique within the 5G K-Simulator, which includes LLS, SLS, and network-level simulator (NS).
• Compared the ML-based EESM method against conventional L2S mapping techniques.

Main Results:

• The proposed ML-based L2S mapping technique demonstrated superior prediction accuracy concerning block error rate (BLER).
• Significant reduction in processing time was observed compared to conventional L2S mapping methods.
• The ML-based EESM method effectively bridges the gap between LLS and SLS for 5G NR simulations.

Conclusions:

• The novel ML-based EESM technique offers a more accurate and efficient approach for L2S mapping in 5G NR systems.
• This method can significantly improve the performance of system-level simulations by leveraging ML for accurate link-level predictions.
• The findings contribute to advancing simulation methodologies in wireless communication research.