Telefonaktiebolaget Lm Ericsson (publ)

Samsung Electronics Co., Ltd.

Utopus Insights, Inc.

Jiangsu University

Gyrk International Technology Co., Ltd.

Lenovo (beijing) Limited

International Business Machines Corporation

Hewlett Packard Enterprise Development LP

Institute of Pharmacology and Toxicology Academy of Military Medical Sciences P.l.a.

Other

Beijing Didi Infinity Technology and Development Co., Ltd.

New H3c Technologies Co., Ltd.

Beijing Samsung Telecom R&d Center

Beijing Jayyalife Biotechnology Co, Ltd.

Nanjing Sanhome Pharmaceutical Co., Ltd.

Hai Feng Wang

Jinhua Liu

Jianfeng Wang

Rui Fan

Qingyu Miao

Gen Li

Qianxi Lu

Fredrik Gunnarsson

Erik A Eriksson

Pål Frenger

Martin Hessler

Claes Tidestav

Zhiheng Guo

Jonas Fröberg Olsson

Niklas Johansson

Wen Jun Yin

Jin Dong

Robert Baldemair

Stefan Parkvall

Bengt Lindoff

Gunnar Mildh

Icaro L J Da Silva

Johan Rune

Yi-Pin Eric Wang

Erik Bengt Dahlman

Andres Reial

Jari Tapio Vikberg

Henrik Sahlin

Stefan Wager

Dennis Hui

Magnus Stattin

Sven Oscar Petersson

Joakim Axmon

Karl Norrman

Johan Torsner

Pradeepa Ramachandra

Gabor Fodor

Fredrik Athley

Janne Peisa

Joachim Sachs

Kumar Balachandran

Ali S Khayrallah

Thorsten Lohmar

Zhang Zhang

Bo Hagerman

Lars Lindbom

Osman Nuri Can Yilmaz

Niclas Wiberg

Johan Furuskog

Torgny Palenius

Lars Sundström

Peter Alriksson

Zhan Zhang

Jose Luis Pradas

Andreas Bergström

Pontus Wallentin

Patrik Rugeland

Henrik Asplund

Martin Johansson

Karl Werner

Mikael Coldrey

Patrik Torgny Karlsson

Johan Axnäs

Mårten Ericson

Jonas Kronander

Shehzad Ali Ashraf

Reza Moosavi

Fredrik Lindqvist

Dennis Sundman

Ning He

Walter Müller

Anders Furuskär

Ruiqi Zhang

Long Chen

Yang Hu

Meng Zhang

Liang Liu

Li Li

Johan Nilsson

Muhammad Ali Kazmi

Wei Wang

David Astely

Dong-seek Park

Xinghua Song

Shaohua Li

Anders Wallén

Wei Liu

Feng Wu

Zhouyue Pi

Dirk Gerstenberger

Cong Shi

Jing Li

Ho-Kyu Choi

Peng Li

Chang-Hoi Koo

Huaisong Zhu

Imadur Rahman

Pan-Yuh Joo

Zhenshan Zhao

The provided are a federated reinforcement learning (FRL) end-to-end autonomous driving control system and method, as well as vehicular equipment, based on complex network cognition. An FRL algorithm framework is provided, designated as FLDPPO, for dense urban traffic. This framework combines rule-based complex network cognition with end-to-end FRL through the design of a loss function. FLDPPO employs a dynamic driving guidance system to assist agents in learning rules, thereby enabling them to navigate complex urban driving environments and dense traffic scenarios. Moreover, the provided framework utilizes a multi-agent FRL architecture, whereby models are trained through parameter aggregation to safeguard vehicle-side privacy, accelerate network convergence, reduce communication consumption, and achieve a balance between sampling efficiency and high robustness of the model.

Complex network cognition-based federated reinforcement learning end-to-end autonomous driving control system, method, and vehicular device

Methods and apparatus in a fifth-generation wireless communications, including an example method, in a wireless device, that includes receiving a downlink signal comprising an uplink access configuration index, using the uplink access configuration index to identify an uplink access configuration from among a predetermined plurality of uplink access configurations, and transmitting to the wireless communications network according to the identified uplink access configuration. The example method further includes, in the same wireless device, receiving, in a first subframe, a first Orthogonal Frequency-Division Multiplexing (OFDM) transmission formatted according to a first numerology and receiving, in a second subframe, a second OFDM transmission formatted according to a second numerology, the second numerology differing from the first numerology. Variants of this method, corresponding apparatuses, and corresponding network-side methods and apparatuses are also disclosed.

Network architecture, methods, and devices for a wireless communications network

The provided is a method and system for trajectory prediction based on Cro-IntentFormer. The method starts by preprocessing vehicle trajectory data collected by sensors to produce raw data suitable for model input. Vehicles are treated as nodes, and the distance between vehicles serves as the basis for determining whether there is an edge between two vehicle nodes. A physical relationship graph is constructed and, along with the raw data, input into a spatio-temporal feature extraction module to obtain the spatio-temporal features of the trajectory. The spatio-temporal feature matrix is then input into an intent prediction module to determine the predicted intentions of the vehicles. Based on the intent information output by the intent prediction module, a semantic relationship graph is reconstructed and input, along with the raw data, into the spatio-temporal feature extraction module to derive the semantic features of the trajectory.

Cro-IntentFormer-based method and system for predicting surrounding vehicle trajectories by integrating driving intentions

A multi-task joint perception network model and detection method for traffic road surface information can simultaneously detect a lane line and a drivable area. A coordinate attention mechanism is integrated into a traditional feature extraction network to ensure that a feature extraction effect is enhanced while a calculated amount is not increased. In a neck network, a dilated convolution residual module is proposed to enhance performance of prediction of details by the network, and a decoder part shares features of the drivable area into lane line detection to enhance a lane line detection effect under complex road conditions. In a training stage, there is provided a alternating optimization training method to improve integral segmentation performance of the model. The multi-task joint perception network model and detection method realizes quite high accuracy and excellent speed performance in a challenging BDD100K dataset.

Multi-task joint perception network model and detection method for traffic road surface information

Methods and apparatus in a fifth-generation wireless communications network, including an example method, in a wireless device, that includes determining a reporting quality threshold for a parameter related to channel state information (CSI); performing a measurement for each of a plurality of beams from a first predetermined set of beams for evaluation; evaluating the measurement for each of the plurality of beams against the reporting quality threshold; discontinuing the performing and evaluating of measurements in response to determining that the reporting quality threshold is met for one of the beams, such that one or more beams in the first predetermined set of beams are not measured and evaluated; and reporting, to the wireless communications network, CSI for the one of the beams.

Embodiments of the present disclosure relate to methods and devices for broadcasting system information on demand. In example embodiments, the system information is classified as essential and optional system information. The essential information is broadcasted periodically, and the optional system information is broadcasted in response to the request by UEs. In this way, the system information can be in an effective and efficient way, thereby saving significant resource and energy can be saved especially in the case of no load or low load.

Methods and devices for broadcasting system information on demand

The present disclosure provides a method () in a network device for channel estimation. The method () includes: transmitting () to a terminal device an instruction to precode each of a number, L, of DeModulation Reference Signals, DMRSs, using a number, N, of linearly independent precoders, respectively; receiving () from the terminal device L*N precoded DMRSs; estimating () an equivalent channel associated with an uplink channel from the terminal device to the network device based on one or more of the L*N precoded DMRSs; and determining () the uplink channel from the equivalent channel based on the N precoders.

Methods and devices for channel estimation

The present invention relates to a built-in bio-sleeve and preparation method and use thereof. In order to provide a suitable built-in bio-sleeve material for dorsal nerve isolation surgery and penis enlargement surgery, a novel acellular allogeneic dermal matrix material is prepared in the present invention, which is manufactured into a built-in bio-sleeve for the treatment of premature ejaculation and short and small penis. The acellular allogeneic dermal matrix or the built-in bio-sleeve of the present invention is used for dorsal nerve isolation surgery, penis enlargement or lengthening surgery without cutting off the dorsal nerve, thereby avoiding the probability of psychological erectile dysfunction caused by the patient's worry about cutting off the dorsal nerve.

Built-in bio-sleeve and preparation method and use thereof

A multi-intelligence federated reinforcement learning (FRL)-based vehicle-road cooperative control system and method at the complex intersection use a vehicle-road cooperative control framework based on the Road Side Unit (RSU) static processing module and the vehicle-based dynamic processing module. The historical road information is supplied by the proposed RSU module. The Federated Twin Delayed Deep Deterministic policy gradient (FTD3) algorithm is proposed to connect the federated learning (FL) module and the reinforcement learning (RL) module. The FTD3 algorithm transmits only neural network parameters instead of vehicle samples to protect privacy. Firstly, FTD3 selects only specific networks for aggregation to reduce the communication cost. Secondly, FTD3 realizes the deep combination of FL and RL by aggregating target critic networks with smaller Q-values. Thirdly, RSU neural network participates in aggregation rather than training, and only shared global model parameters are used.

Multi-intelligence federal reinforcement learning-based vehicle-road cooperative control system and method at complex intersection

The present invention relates to a method and apparatus for controlling a wind turbine. The method includes: dividing a plurality of wind turbines into at least one group based on a similarity in status information of the plurality of wind turbines; in response to having detected a fault in a first wind turbine of the plurality of wind turbines, searching a group to which the first wind turbine belongs for a second wind turbine matching status information of the first wind turbine; and controlling the first wind turbine based on parameters from the second wind turbine.

Controlling wind turbine

A method and device for determining power control configuration. it is determined whether a first transmit power control configuration based on a first uplink transmission is outdated; and a second transmit power control configuration for a second uplink transmission is determined based on a power control scheme. Therefore, transmit power control efficiency of sparse data transmissions may be improved.

Method and device for determining power control configuration

A method and apparatus for forecasting output power of wind turbine in a wind farm. The present invention provides a method for forecasting output power of a wind turbine in a wind farm, including: generating a corrected data set based on environmental data collected from at least one sensor in the wind farm; correcting a weather forecasting model by using the corrected data set; obtaining a forecast value of wind information at the wind turbine based on the corrected weather forecasting model; and forecasting the output power of the wind turbine based on the forecast value and a power forecasting model.

Forecasting output power of wind turbine in wind farm

Growing community of inventors

Beijing, China

Hai Feng Wang