US20260045982
2026-02-12
Electricity
H04B7/0608
The patent application focuses on optimizing antenna selection in multi-antenna user equipment (UE) using artificial intelligence (AI). It introduces a method where grip data indicating how a user holds the UE is generated. This grip data supports a lookup process to determine antenna selection and tuner state from multiple options. Reference signal measurements from each antenna are also used as inputs to an AI model, which outputs the optimal antenna selection, tuner state, or a predicted angle of arrival for signal optimization.
Cellular networks facilitate communication between a UE and the network, with data transferred via downlink and uplink. Techniques such as frequency division duplexing (FDD) and time division duplexing (TDD) are employed for signal transmission and reception, managed by the radio frequency (RF) front end of the UE. Selecting the right antenna is crucial for efficient data transmission, especially when the signal's angle of arrival (AoA) is unknown.
The application outlines methods for selecting the best antenna for transmission and reception based on user grip and other factors like UE type and frequency band. A predefined lookup process, informed by simulations and grip data, identifies an antenna subset likely to offer superior performance. If multiple antennas are viable, further measurements determine the optimal choice. Alternatively, an AI model can predict the best antenna or tuner state based on reference signal measurements and grip data.
The described techniques enhance data throughput and reduce power consumption by ensuring optimal antenna selection. These improvements are applicable across various network types, including 4G, 5G, and emerging 6G networks, as well as WiFi and Bluetooth systems. The AI-driven approach allows for dynamic adaptation to changing conditions, ensuring consistent performance.
The application is demonstrated within the context of a 5G network but is adaptable to other wireless technologies. It covers scenarios where different antennas might be used for uplink and downlink, depending on factors like signal strength and user grip. The use of AI models allows for real-time adjustments, providing a robust solution for modern wireless communication challenges.