Enhancement of Spectral and Energy Performance for mmWave MIMO Communication System for Green Communication
Loading...
Date
item.page.authors
Journal Title
Journal ISSN
Volume Title
Publisher
Abstract
The 5G network aggregates a number of efficient technologies, supporting green communication as it significantly reduces CO2 emissions. Numerous new 5G enabled sustainability offerings that individuals and different organizations can enforce to lessen the carbon footprints and to gain different sustainability goals is possible using advanced 5G connectivity. The recent years have witnessed the utilisation of machine learning architecture adopted in the recent frames of mm-wave based MIMO communication system. In a similar context, the study proposes a load utilisation-oriented Swarm-based Artificial Bee Colony algorithm for better utilisation of the resources. The proposed work uses the Levenberg principle for layer propagation which is also a Machine Learning mechanism that is utilised for better allocation of channels to deliver enhanced quality in the communication process. The improved strength of the proposed allocation mechanism is evaluated in terms of throughput and Bit Error Rate (BER) by comparing against different scenarios. Further, comparative analysis against existing swarm-based optimization architectures is also conducted to demonstrate and#8764; 2% reduction in the BER using proposed allocation work. The second phase of study presents an improved mm-wave communication system by incorporating Artificial Bee Colony (ABC) algorithm for feature vector selection and using the firefly algorithm for hyperparameter tuning. Experiment results demonstrate the superiority of the proposed system in terms of both qualitative and quantitative parameters. For qualitative parameters analysis, the proposed work has been evaluated in terms of throughput and BER, where the proposed system outperforms by increment in throughput with 10.31% and lower BER over existing system. In terms of quantitative analysis, the proposed algorithm outperforms with the accuracy of 92.03 %, precision of 96.30%, Recall of 98.96 % and F-measure of 97.61 %.
newline