	US 11,991,646 B2
	Power management apparatus based on user pattern and method
Kang Yoon Lee, Seoul (KR); Jong Wan Jo, Suwon-si (KR); Young Gun Pu, Suwon-si (KR); Dong Soo Park, Suwon-si (KR); Joon Hong Park, Suwon-si (KR); Jae Bin Kim, Suwon-si (KR); and Yun Gwan Kim, Suwon-si (KR)
Assigned to SKAICHIPS CO., LTD., Suwon-si (KR); and Research & Business Foundation Sungkyunkwan University, Suwon-si (KR)
Filed by SKAICHIPS CO., LTD., Suwon-si (KR); and Research & Business Foundation Sungkyunkwan University, Suwon-si (KR)
Filed on Oct. 22, 2021, as Appl. No. 17/508,095.
Claims priority of application No. 10-2020-0137742 (KR), filed on Oct. 22, 2020.
Prior Publication US 2022/0132439 A1, Apr. 28, 2022
Int. Cl. H04W 52/28 (2009.01); H04W 24/08 (2009.01)

CPC H04W 52/28 (2013.01) [H04W 24/08 (2013.01)]

8 Claims

1. A power management apparatus, comprising:

an artificial intelligence (AI) controller configured to monitor a user pattern, based on frequency band selection information of all users using a base station, to predict the user pattern; and

a DC-DC converter configured to output a supply voltage based on the predicted user pattern,

wherein the AI controller is further configured to detect a load of a low-dropout (LDO) to control a core size, a soft-start time, a discontinuous Pulse-width modulation/digital power system management (DPWM/DPSM) mode, a gate driver size, and a switching frequency to track a minimum supply voltage satisfying a predetermined minimum noise performance based on an AI algorithm.