Intel Corporation

Cisco Technology, Inc.

Google Inc.

Deepmind Technologies Limited

Xiaohong Gong

Jesse R Walker

Bretton Lee Douglas

Johannes Petrus Kruys

Brian Donald Hart

David Sheldon Stephenson

Robert B O'Hara, Jr

Minyoung Park

Amogh Sanjay Asgekar

Krishnamurthy Dvijotham

Anton Zhernov

Robert J Stacey

Emily H Qi

Adrian Paul Stephens

Tsung-Yuan Charles Tai

Wei Zhang

Jonathan Segev

Victor B Lortz

Anand P Rangarajan

Patrice R Calhoun

Lu Qian

Christopher Graham Meil

Jyoti Raju

Timothy Arthur Mann

Dan-Andrei Calian

Nima Asgharbeygi

Arturo Bajuelos Castillo

Anton Alexandrov

Xueliang Lu

Carsten Miklos Steinebach

Ivan Lobov

Sanjeev Jagannatha Rao

Douglas Bretton Lee

Methods and systems for low-latency multi-constraint ranking of content items. One of the methods includes receiving a request to rank a plurality of content items for presentation to a user to maximize a primary objective subject to a plurality of constraints; initializing a dual variable vector; updating the dual variable vector, comprising: determining an overall objective score for the dual variable vector; identifying a plurality of candidate dual variable vectors that includes one or more neighboring node dual variable vectors; determining respective overall objective scores for each of the one or more candidate dual variable vectors; identifying the candidate with the best overall objective score; and determining whether to update the dual variable vector based on whether the identified candidate has a better overall objective score than the dual variable vector; and determining a final ranking for the content items based on the dual variable vector.

Low latency multi-constraint ranking of content items

Methods, systems, and apparatus, including computer programs encoded on computer storage media, for training a neural network used to generate a ranking score for a network input. One of the methods includes generating training data and training the neural network on the training data. The training data includes a plurality of training pairs. The generating comprising: obtaining data indicating that a plurality of training network inputs were displayed in a user interface according to a presentation order, obtaining data indicating that a first training network input of the plurality of training network inputs has a positive label, determining that a second training network input of the plurality of training network inputs (i) has a negative label and (ii) is higher than the first training network input in the presentation order, and generating a training pair that includes the first training network input and the second training network input.

Pairwise ranking using neural networks

Embodiments provide techniques for device power management in wireless networks. For instance, an apparatus may include a power management module, and a transceiver module. The power management module determines a beacon interval and a wakeup interval. The transceiver module to send a transmission to one or more remote devices that includes the beacon interval and the wakeup interval. The beacon interval indicates a time interval between consecutive beacon transmissions of the apparatus, and the wakeup interval indicates a time interval between when the apparatus receives two consecutive beacons from a peer device.

Power management for wireless networks

Methods, systems, and apparatus, including computer programs encoded on computer storage media, for re-ranking a collection of documents according to a first metric and subject to a constraint on a function of one or more second metrics. One of the methods includes: obtaining, for each document in the first collection of documents, a respective first metric value corresponding to the first metric and respective one or more second metric values corresponding to the one or more second metrics; re-ranking the first collection of documents, comprising: determining the constraint on the function of one or more second metrics by computing a first threshold value using a variable threshold function that takes as input second metric values for the documents in the first collection of documents; and determining the re-ranking for the first collection of documents by solving a constrained optimization for the first metric constrained by the first threshold value.

Variable thresholds in constrained optimization

Systems, methods, and computer-readable storage media for distributed and automated prediction of future customer revenue are provided. One method involves accessing data structures, each representing a unique customer, storing a set of customer-specific characteristics, segregating the data structures into groups based on a target amount of data structures for each group, and inputting the customer-specific characteristics into a training model. The method includes generating a set of prediction model parameters for each group by applying the customer-specific characteristics to a training model. The method includes transforming the characteristics of each data structure in a first group into respective future revenue values using a first non-linear prediction model, and the characteristics of data structures in a second group into respective future revenue values using a second prediction model. A portion of the future revenue values for the groups is calculated in parallel, and the calculated values are stored in a memory.

Distributed and automated system for predicting customer lifetime value

System and methods of this disclosure are directed to recommending content in real-time or near real-time. The system comprises a number of pipelines updated a different frequencies that process temporally different sets of web property visit data. Within each pipeline, the system can employ different number of algorithms to process visit data to generate content recommendations. One algorithm is a content filter that filters from the visit data those determined to be unsuitable as recommendations. Another is a content analyzer that analyzes the content of each URL in the visit data by topic category and attribute. Another is an item-to-item collaborative filter that determines a correlation score for each URL based on those in the visit data in a single session Another is a device-to-item matrix factorization that determines an affinity score for each URL based on visit data, context information, and topic category.

Real-time content recommendation system

Embodiments of a system for station group management and a method for managing station-management groups in a wireless network are generally described herein. In some embodiments, a system element, such as an access point, may establish a station-management group including selected stations, transmit a unicast group allocation message to a station to indicate whether the station is part of the station management group, transmit data frames to the station in a downlink (DL) multi-user multiple-input multiple-output (MU-MIMO) transmission, and delete the station from the station-management group.

System for station group management and method for managing station-management groups

A method includes determining a total bandwidth for a downlink communication, segmenting the downlink communication into data units and assigning the data units to channels assigned to stations of a basic service set. The method also includes transmitting a channel announce frame to stations of the basic service set. The channel announce frame indicates each station assigned to receive a corresponding data unit, and one or more channels assigned to each station for transmitting the corresponding data unit. The method further includes transmitting the corresponding data units over the assigned channels to the assigned stations in a single downlink communication.

Multiple-channel downlink transmission for orthogonal frequency-division multiple access

A method includes receiving, by control circuitry, a received portion of a first data unit from a remote device and determining, by the control circuitry, a first duration of the first data unit based on the received portion. The method further includes generating, by the control circuitry, a second data unit with data to be transmitted and transmitting, by the control circuitry, the second data unit to the remote device while receiving the unreceived portion of the first data unit. The second data unit has a second duration equal to the remaining duration of an unreceived portion of the first data unit.

Method and control circuitry for performing full-duplex wireless communication

Embodiments of a MU-MIMO access point, user station and method for multi-user group management are generally described herein. In some embodiments, a MU-MIMO access point may assign a MU group identifier (GID) and a group member index (GMI) to each of a plurality of associated user stations. The MU-MIMO access point may transmit a MU-MIMO transmission that includes a plurality of spatial streams. The MU-MIMO transmission may be destined for the user stations of a single MU group indicated by a GID that is included within the MU-MIMO transmission. A subset of the spatial streams may be intended for each user station of the indicated MU group that is identified by the GID transmitted within the MU-MIMO transmission. A user station may use its assigned GMI along with an Nfield in the MU-MIMO transmission to determine which of the spatial streams to demodulate.

MU-MIMO access point and user station including methods for multi-user group management

Traffic load over a wireless medium due to wireless access point discovery is reduced. The wireless stations and wireless access points support wireless discovery using a reduced traffic load process. The reduced traffic load process includes providing aggregated short probe responses.

Wireless discovery using a reduced traffic load process

Growing community of inventors

Sunnyvale, CA, United States of America

Xiaohong Gong