On vehicular communications in 5G dynamic environments

The goal of autonomous driving systems is to reach the highest possible level of autonomy, where no human intervention would be required. Achieving this level will rely heavily on Vehicle-to-Everything (V2X) communications. Focusing on the physical (PHY) layer of 5G-based communications, Line-of-Sight (LOS) identification is essential for V2X communications. In V2X systems, LOS paths typically offer the most robust and reliable signals, so as recognizing LOS conditions may lead to improved communication quality, with less signal fading and fewer errors, especially if compared to Non LOS (NLOS) scenarios, where buildings, vehicles, or terrain may obstruct signals. In this Thesis dissertation, a Matlab-based 5G NR DL simulator is used in order to test a developed approach for automatically identifying the channel status conditions of several gNBs-UEs links, by exploiting the collected CSI-RSRP measurements. In dense networks requiring reliable and low latency communications, such as the vehicular networks, managing HandOvers (HOs) becomes of paramount importance. In particular, a phenomenon that deteriorates the QoS and Quality of Experience (QoE) perceived by the UEs is represented by the Unnecessary HOs (UHOs). In this Thesis dissertation, an approach for UHOs mitigation in a 5G NR dynamic environment is proposed. Finally, even focusing on physical aspects of vehicular communication-oriented environments, nowadays, Integrated Sensing and Communication (ISAC) is one of the most exploited technologies for wireless communications thanks to its potential to reduce hardware costs and improve spectral efficiency. In this Thesis dissertation, an approach for the minimization of the resource allocated to the vehicular UEs for the UpLink (UL) transmission of the data extracted from a target is proposed. The goal is to allow the serving Base Station (BS) to receive the necessary information to manage the vehicular network in time to take the required actions.