Proactively Predicting Dynamic 6G Link Blockages Using LiDAR and In-Band Signatures

Line-of-sight link blockages represent a key challenge for the reliability and latency of millimeter wave (mmWave) terahertz (THz) communication networks. To address this challenge, paper leverages mmWave LiDAR sensory data to provide awareness about environment proactively predict dynamic before they occur. This allows network make proactive decisions hand-off/beam switching, enhancing latency. More specifically, addresses following questions: (i) Can we line-of-sight blockage, it happens, using in-band mmWave/THz signal sensing data? (ii) also when blockage will occur? (iii) duration? And (iv) can direction moving blockage? For that, develop machine learning solutions that learn special patterns received data, which call pre-blockage signatures, infer future blockages. evaluate proposed approaches, build large-scale real-world dataset comprises co-existing measurements in outdoor vehicular scenarios. Then, an efficient denoising algorithm applies some pre-processing data. Based on dataset, developed approaches are shown achieve above 95% accuracy predicting occurring within 100 ms more than 80% prediction one second. Given capability, shows order magnitude saving latency, further highlights potential wireless