US 12,088,369 B2
Method and apparatus for learning-based channel matrix prediction
Hyukjoon Kwon, San Diego, CA (US); Mohamed Mahmoud, Richardson, TX (US); Federico Penna, San Diego, CA (US); Yanlin Zhou, San Diego, CA (US); and Ramy E. Ali, San Diego, CA (US)
Assigned to Samsung Electronics Co., Ltd, (KR)
Filed by Samsung Electronics Co., Ltd., Gyeonggi-do (KR)
Filed on Dec. 30, 2022, as Appl. No. 18/091,641.
Claims priority of provisional application 63/399,856, filed on Aug. 22, 2022.
Prior Publication US 2024/0063852 A1, Feb. 22, 2024
Int. Cl. H04B 7/0456 (2017.01); H04L 5/00 (2006.01); H04W 72/0446 (2023.01); H04W 72/0453 (2023.01)
CPC H04B 7/0456 (2013.01) [H04L 5/0023 (2013.01); H04W 72/0446 (2013.01); H04W 72/0453 (2013.01)] 20 Claims
OG exemplary drawing
 
1. A method comprising:
estimating, by a user equipment (UE), channel matrices at time slots using reference signals received from a base station (BS) at the time slots;
predicting, by the UE, a sequence of channel matrices at future time slots using the estimated channel matrices as input to a neural network (NN) trained based on known sets of past and future channel matrices; and
determining, by the UE, a parameter using at least one channel matrix from the sequence of channel matrices.