US20250365056
2025-11-27
Electricity
H04B7/06952
The patent application focuses on enhancing beam alignment in communication systems by utilizing artificial intelligence and machine learning. In the process of beam alignment, a client device identifies a set of potential transmit beams from a network access node based on reference signals. The client device then calculates beam measurement information to optimize future reference signal transmissions. This information is communicated back to the network access node to improve subsequent interactions.
Beam management, particularly in high-frequency bands like 5G NR frequency range 2, involves selecting suitable transmission and reception beams for effective communication. The complexity of this process arises from the narrow beam widths required for adequate coverage and the dynamic nature of user equipment (UE) movement. Traditional approaches struggle with these challenges, prompting the integration of AI/ML to predict spatial and temporal beam characteristics more efficiently.
The proposed solution addresses limitations of conventional methods by improving beam management between client devices and network nodes. The client device is configured to determine candidate transmit beams based on measured reference signals, calculate beam measurement information, and communicate this data to the network node. This process optimizes subsequent reference signal transmissions, reducing the number of necessary measurements and reporting overhead.
The client device's beam measurement information can indicate specific transmit or receive beams for future measurements, reducing unnecessary data collection and latency. This information can be formatted as a bitmap, which simplifies the identification of optimal beam pairs. Additionally, the client device selects beam pairs based on these measurements and transmits indicators to the network node, further optimizing communication efficiency.
By leveraging existing 5G NR beam measurement metrics like signal-to-noise ratios and received signal power, the implementation remains practical while enhancing performance. The proposed method reduces complexity and latency in beam alignment processes, ensuring more efficient use of uplink control resources and minimizing overhead during reference signal measurements.