| Abstract |
Massive Internet of Things (IoT) with millions of devices communicating at high data rates and low latency will be made possible by 6G. Advanced sensing is a major application area for 6G. Higher speeds, however, entail shifting to higher frequencies and typically call for greater transmission power. IoT sensors, however, need to be aware of their positions in many applications. Promising localization options have resulted from recent advancements in the drone industry. This study looks into the “dead zone” and coverage region for indoor flying receiver localisation (such as drones). Weighted graph structures are used to perform the drones’ sensor visiting order. Additionally, Using a non-Hinge sensor removal method, the transmission’s path length is shortened. Next, a machine learning model known as Random Forest is used to segment the photos that the drones took. DL DenseNet is a deep-learning network used for drone monitoring after the segmentation process. Drone energy consumption can be decreased by optimizing drone location and sensor visitation sequence. The suggested model is superior, as demonstrated by the simulation results. ML classifiers became a more sensible option for specialised applications like wetland mapping, where networks must be trained for every unique site, topography, season, and other meteorological factors. The top pixel-based machine learning classifier was Random Forest (RF), which offered good picture segmentation accuracy (91%) during testing. The enhanced accuracy of 98.3% in detecting DL-based dead zones helps rescue personnel identify dangerous areas more rapidly. Following contemporary smart city ambitions, disaster management systems will be transformed using technologies like real-time dead zone monitoring systems. © The Author(s), under exclusive licence to Springer Science+Business Media, LLC, part of Springer Nature 2024. |
