Reconfigurable Intelligent Surfaces (RISs) are being considered as a potential technology for upcoming 6G cellular communication systems. RISs offer the capability to steer a wireless signal in a desired direction, thereby enhancing the overall performance of wireless systems in terms of coverage and spectral efficiency. However, the conventional methods of RIS beamforming result in significant inter-beam interference primarily caused by relatively high Side Lobe Levels (SLLs). In antenna beamforming, amplitude tapering has shown promising results in Multi-User Multiple-Input Multiple-Output (MU-MIMO) systems. Amplitude tapering allows for a notable reduction in inter-beam interference by decreasing SLLs, albeit at the expense of a reduction in the Main Lobe Level (MLL) and an increase in its width. In this paper we investigate the feasibility of utilizing this technique for RIS beamforming. In addition, we demonstrate the effectiveness of this technique through link-level simulations of a RIS-aided multi-user downlink system.

In 4G/5G networks, the usual role of the Random Access Channel (RACH) signal is to establish uplink synchronization by means of Timing Advance (TA) estimation. In this study, we explore the upper limits of precision for this TA parameter, as well as for the Carrier Frequency Offset (CFO), by means of an original Cramer-Rao Lower Bound (CRLB) calculation. Remarkably, this performance can be encapsulated within a straightforward formula, uncovering a potential for precision that surpasses the Nyquist period by one to two orders of magnitude. This enhanced precision holds promise for geolocation application. Furthermore, we demonstrate that the Maximum Likelihood (ML) estimator closely approximates this theoretical boundary, and we propose a novel implementation method employing a compact Neural Network. A series of numerical evaluations affirm the potential of this innovative approach, ensuring both statistical optimality and the feasibility of real-time implementation.