MIMO Beam Identification via Training Signal Feedback

The patent applies preliminary action by introducing a beam training signal sent by the Node B before actual data transmission. This training signal allows the terminal to detect and measure beam characteristics in advance, enabling the terminal to feed back beam indication information to the Node B. Through this preliminary beam training process, the Node B can identify the optimal beam for coverage before obtaining the beamforming weight, resolving the contradiction between improving coverage performance and the difficulty of beam identification.

The patent applies segmentation by dividing the beam training process into discrete beam directions, where each beam training signal corresponds to a specific beam direction. The Node B sends multiple beam training signals covering different directions, and the terminal detects these segmented signals to identify the optimal beam. This segmentation approach allows systematic beam identification while managing time and frequency resource consumption through structured training sequences.

The patent applies periodic action by implementing time division multiplexing where beam training signals are sent periodically in different time slots. Each beam training signal is transmitted in a specific time period, allowing the terminal to detect and measure each beam's characteristics sequentially. This periodic transmission pattern improves beam separation and detection precision while managing training duration through efficient time slot allocation and periodic repetition of training sequences.