Proactive Prediction of Dynamic Link Blockages in mmWave Optimized Using Global Awareness Siamese Network

In Millimetre-wave (mmWave) communications, dynamic link blockages can cause major signal propagation disturbances that affect network performance as a whole. The increased susceptibility to obstacles brought on by mmWave signals' great sensitivity to environmental influences and physical barriers is a significant disadvantage. In this Manuscript, Proactive Prediction of Dynamic Link Blockages in mmWave Optimized Using Global Awareness Siamese Network (PPDLB-MMW-GASN) is proposed. At first, the data are collected from the Large-Scale actual-world mmWave Datase, which t is provided as input. Afterward, the data are fed to pre-processing. In pre-processing, Subaperture Keystone Transform Matched Filtering (SKTMF) is employed for Data standardization. Then the pre-processed output is fed to the Global Aware Siamese Network (GASN) for predicting the future dynamic line of sight (LOS) link blockages. In general, GASN does not express adaptive optimization strategies to determine optimal parameters to predict the future dynamic LOS link blockages. The Coronavirus Mask Protection Algorithm (CMPA) is utilized to optimize the weight parameters of GASN. The effectiveness of the proposed PPDLB-MMW-GASN method is implemented in Python and evaluated through performance measures like Accuracy, Mean Square Error (MSE), Coefficient of Determination (R2), Mean Opinion Score (MOS), Residual Sum of Squares (RSS) and Throughput. The proposed PPDLB-MMW-GASN approach attains 26.42%, 23.36% and 29.27% higher accuracy, 15.36%, 14.42% and 15.13% lower Mean Square Error, 28.36%, 26.42% and 28.27% higher Mean Opinion Score, 23.40%, 27.33% and 29.37% lower Residual Sum of Squares and 27.44%, 17.31% and 34.36% higher Throughput when compared with existing approaches, such as LiDAR-Aided Future Beam Prediction in Real-World Millimeter Wave V2I Communications (LAFB-V2IC-RNN), Proactively Predicting Dynamic 6G Link Blockages Using LiDAR and In-Band Signature (6GLB-LiD-IBS-MMNN), and Blockage Prediction Using Wireless Signatures: Deep Learning Enables Real-World Demonstration (BPWS-RWD-CNN).