,
Professor, Department of CSE-AIML, Sapthagiri NPS University, Bengaluru, Karnataka, India
,
Professor & Director, Department of Electrical and Electronics Engineering, Sapthagiri NPS University, Bengaluru, Karnataka, India
,
Assistant Professor, Department of CSE-DS, Sapthagiri NPS University, Bengaluru, Karnataka, India
,
Professor, Department of Electronics and Communication Engineering, Sapthagiri NPS University, Bengaluru, Karnataka, India
,
Assistant Professor, Department of AI &DS, Koneru Lakshmaiah Education foundation (KLEF), Greenfields, Vaddeswaram, Guntur, Andhra Pradesh, India
,
Professor, Department of CSE-DS, Sapthagiri NPS University, Bengaluru, Karnataka, India
Associate Professor, Department of School of Applied Science, Sapthagiri NPS University, Bengaluru, Karnataka, India
Adaptive Modulation and Coding (AMC) is a key principle that increases spectral efficiency, reliability, and capacity in high-data-rate mmWave wireless communication systems. But AMC strategies depend upon static threshold-based rules that generally struggle to perform well in varying channel conditions. In this study, an innovative HO-DNN architecture is introduced to choose an appropriate AMC strategy in real-world scenarios, using the DeepMIMO dataset for channel emulation. This architecture leverages channel quality features and propagation properties like SNR, SINR, Path Loss, Delay Spread, Angle of Arrival, Angle of Departure, Communication Distance, and Doppler Frequency. A combination of Bayesian Optimization and Genetic Algorithm (BO-GA) is used for tuning of learning rate, batch size, number of layers, and dropout probability of the neural network model. The experiments were carried out using around one million DeepMIMO channel samples in the scenario of 28 GHz millimeter wave communications with massive 64×64 MIMO antennas. The classification accuracy of the proposed architecture was obtained as 99.08%, surpassing existing DNN (97.08%), CNN (95.26%), and LSTM (96.41%). Moreover, the designed network achieved a throughput of 11.84 Gbps and reduced the bit error rate to 2.1×10-5. This proves that hyperparameters play an important role in improving AMC predictions' accuracy and enhancing the overall communication process for advanced 5G and 6G applications.
This is an open access article distributed under the Creative Commons Attribution Non-Commercial License (CC BY-NC) License which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
The statements, opinions and data contained in the journal are solely those of the individual authors and contributors and not of the publisher and the editor(s). We stay neutral with regard to jurisdictional claims in published maps and institutional affiliations.
