公开(公告)号：US20230216760A1

公开(公告)日：2023-07-06

申请号：US18047241

申请日：2022-10-17

Applicant: Samsung Electronics Co., Ltd.

Inventor： Khuong N. Nguyen ,  Guanbo Chen ,  Hao Chen ,  Abhishek Sehgal ,  Rebal Al Jurdi

IPC: H04L43/062 ,  H04L43/028 ,  H04L41/16

CPC classification number: H04L43/062 ,  H04L43/028 ,  H04L41/16

Abstract: A method includes obtaining input features based on network traffic received during a time window. The method also includes generating multiple network service type predictions about the network traffic during the time window using a machine learning (ML) classification system operating on the input features. The method also includes storing the multiple network service type predictions in different time steps in a first-in first-out (FIFO) buffer and generating decisions about a presence of each of multiple service types in the network traffic using a voting algorithm. The method also includes reducing fluctuations in the generated decisions using a logic-based stabilizer module to generate a final network service type decision.

公开(公告)号：US20220377664A1

公开(公告)日：2022-11-24

申请号：US17656589

申请日：2022-03-25

Applicant: Samsung Electronics Co., Ltd.

Inventor： Khuong N. Nguyen ,  Wenxun Qiu ,  Guanbo Chen ,  Peshal Nayak

IPC: H04W52/02 ,  G06V10/764 ,  G06K9/62

Abstract: A method includes obtaining network traffic information based on network traffic received during a time window. The method also includes determining a network service type using a machine learning classification system operating on the network traffic information. The method also includes determining a latency requirement based on the network service type. The method also includes adjusting one or more Target Wakeup Time (TWT) intervals and a wake duration based on the latency requirement to optimize power consumption of a Wi-Fi station.

公开(公告)号：US11467254B2

公开(公告)日：2022-10-11

申请号：US17184450

申请日：2021-02-24

Applicant: Samsung Electronics Co., Ltd.

Inventor： Guanbo Chen ,  Wenxun Qiu ,  Boon Loong Ng

IPC: G01S7/41 ,  G06F3/01

Abstract: A method includes receiving from a transceiver, at a processor of an electronic device, a radar signal, the radar signal including a set of signal strength values across a range of Doppler frequency shift values and a range of time values. The method further includes extracting a time series of frequency features from the radar signal, wherein frequency features of the time series of frequency features include a determined Doppler frequency shift value for each time value of the range of time values, identifying segments within the time series of frequency features, analyzing the segments to determine a start time for a classification engine, at the start time, analyzing, by the classification engine, at least one of a subset of the time series of frequency features or a subset of the radar signal to identify a control gesture, and triggering a control input associated with the control gesture.

公开(公告)号：US20220217630A1

公开(公告)日：2022-07-07

申请号：US17645953

申请日：2021-12-23

Applicant: Samsung Electronics Co., Ltd.

Inventor： Guanbo Chen ,  Abhishek Sehgal ,  Wenxun Qiu ,  Hao Chen

IPC: H04W52/02 ,  H04W72/12 ,  H04W74/08

Abstract: Embodiments of the present disclosure provide methods and apparatuses for updating a target wake time (TWT) service period and interval. The apparatuses include a communication device comprising a transceiver and a processor. The transceiver is configured to transmit and receive higher layer data packets in a TWT operation during a time period. The processor is configured to determine, based on PHY data rates during the time period, effective higher layer data rates, estimate an initial data transceiving time based on the effective higher layer data rates and total lengths of the data packets, adjust the initial data transceiving time to obtain a higher layer data transceiving time based on an estimated network congestion level, an estimated re-transmission rate, and a total amount of TWT overhead during the time period, and determine a new TWT service period and interval based on the higher layer data transceiving time.