Comparative analysis of machine learning-based link adaptation schemes for downlink 5G communications system.
Saved in:
| Title: | Comparative analysis of machine learning-based link adaptation schemes for downlink 5G communications system. |
|---|---|
| Authors: | Sheik Mamode, Maryam Imran1 (AUTHOR) maryamsheikmamode@hotmail.com, Fowdur, Tulsi Pawan2 (AUTHOR) |
| Source: | Wireless Networks (10220038). Apr2026, Vol. 32 Issue 2, p767-795. 29p. |
| Subjects: | Machine learning, Modulation coding, Channel coding, Modulation theory, 5G networks, Telecommunication systems, MatLab (Computer software), Computer performance |
| Abstract: | 5G is the fifth generation of mobile communications, designed to deliver higher data rates, improved spectral efficiency, and enhanced user experience. In this context, link adaptation plays a critical role by dynamically adjusting transmission parameters such as modulation order and coderate based on prevailing channel conditions. Advancing Machine Learning (ML)-driven link adaptation strategies that jointly optimize modulation and coding decisions under realistic 5G downlink conditions remains a key research priority. This study develops an adaptive modulation and coderate selection framework utilizing four ML techniques—K-Nearest Neighbour (KNN), Decision Tree, Random Forest, and Multi-Layer Perceptron—to enhance link adaptation performance in 5G downlink systems. Unlike conventional schemes, the proposed approach integrates additional 5G-relevant modulation formats, specifically 32QAM, 128QAM, and 512QAM, and evaluates the performance across three coderate configurations (340/1024, 490/1024, and 772/1024). Simulation results for the coderate 772/1024 scenario demonstrate significant throughput gains over traditional mapping strategies for KNN, Decision Tree and Random Forest, achieving improvements of 46.9%, 32.6%, 21.9%, 24.1%, 20.5%, 28.2% and 26.4% compared to 16QAM, 32QAM, 64QAM, 128QAM, 256QAM, 512QAM and 1024QAM respectively. The simulation findings demonstrate the efficacy of incorporating incremental modulation schemes and jointly optimizing coderate and modulation selection for enhanced link adaptation in 5G downlink systems. Furthermore, the proposed work provides detailed methodological documentation and a reproducible MATLAB-based implementation, enabling independent validation and supporting future advancements in ML-based link adaptation research. [ABSTRACT FROM AUTHOR] |
| Copyright of Wireless Networks (10220038) is the property of Springer Nature and its content may not be copied or emailed to multiple sites without the copyright holder's express written permission. Additionally, content may not be used with any artificial intelligence tools or machine learning technologies. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.) | |
| Database: | Engineering Source |
| Abstract: | 5G is the fifth generation of mobile communications, designed to deliver higher data rates, improved spectral efficiency, and enhanced user experience. In this context, link adaptation plays a critical role by dynamically adjusting transmission parameters such as modulation order and coderate based on prevailing channel conditions. Advancing Machine Learning (ML)-driven link adaptation strategies that jointly optimize modulation and coding decisions under realistic 5G downlink conditions remains a key research priority. This study develops an adaptive modulation and coderate selection framework utilizing four ML techniques—K-Nearest Neighbour (KNN), Decision Tree, Random Forest, and Multi-Layer Perceptron—to enhance link adaptation performance in 5G downlink systems. Unlike conventional schemes, the proposed approach integrates additional 5G-relevant modulation formats, specifically 32QAM, 128QAM, and 512QAM, and evaluates the performance across three coderate configurations (340/1024, 490/1024, and 772/1024). Simulation results for the coderate 772/1024 scenario demonstrate significant throughput gains over traditional mapping strategies for KNN, Decision Tree and Random Forest, achieving improvements of 46.9%, 32.6%, 21.9%, 24.1%, 20.5%, 28.2% and 26.4% compared to 16QAM, 32QAM, 64QAM, 128QAM, 256QAM, 512QAM and 1024QAM respectively. The simulation findings demonstrate the efficacy of incorporating incremental modulation schemes and jointly optimizing coderate and modulation selection for enhanced link adaptation in 5G downlink systems. Furthermore, the proposed work provides detailed methodological documentation and a reproducible MATLAB-based implementation, enabling independent validation and supporting future advancements in ML-based link adaptation research. [ABSTRACT FROM AUTHOR] |
|---|---|
| ISSN: | 10220038 |
| DOI: | 10.1007/s11276-025-04081-4 |
