Qualcomm Incorporated

Edward Harrison Teague

Avneesh Agrawal

Alexei Yurievitch Gorokhov

Aamod Dinkar Khandekar

David Jonathan Julian

Arak Sutivong

Dhananjay Ashok Gore

Fatih Ulupinar

Hemanth Sampath

Tingfang Ji

Rajat Prakash

Gavin Bernard Horn

Tamer Adel Kadous

Zhongze Wang

Rajiv Vijayan

Zhanfeng Jia

Leonard Henry Grokop

Max Welling

Shankar Sadasivam

Peter S Gaal

Juan Montojo

Alberto Rico Alvarino

Aleksandar Damnjanovic

Yongbin Wei

Ashok Mantravadi

Qingjiang Tian

Michael Franco Taveira

Paul Guckian

Nileshkumar Jayendrabhai Parekh

Charles Wheeler Sweet, Iii

Joseph Binamira Soriaga

Sanjiv Nanda

Yang Yang

Gwendolyn Denise Barriac

Xiaochun Zhu

Christopher G Lott

Ashwin Deepak Swaminathan

Da-shan Shiu

Messay Amerga

Min Dong

Somdeb Majumdar

Jin Won Lee

Charles Alfred Bergan

Jinwon Lee

Lu Xiao

Jeremy H Lin

Carlos Manuel Puig

Brian Michael Buesker

Prashanth Akula

Weiliang Zeng

David Merrill Jacobson

Rashmi Kulkarni

Vitor R Carvalho

Vito Remo Bica

Rinat Burdo

Disha Ahuja

Adrian J Prentice

Bhuwan Dhingra

Faith Ulupinar

Aamod Khandekar

Matthias Reisser

Hermanth Sampath

Siddika Parlak Polatkan

Saurabh Kedar Pitre

Alexei Gorokhov

Vitor Carvalho

Vito R Bica

Aamond Khandekar

Arak c/o Qualcomm Incorporated Sutivong

Gwendolyn D c/o Qualcomm Incorporated Barriac

Aamod c/o Qualcomm Incorporated Khandekar

Ji Tingfang

Rajat Qualcomm Incorporated Prakash

Aamodc/o Wagner & Geyer Khandekar

Avneesh Agraval

A method for optimizing the compilation of a machine learning model to be executed on target edge devices is provided. Compute nodes of a plurality of compute nodes are allocated to a compiler optimization process for a compiler of said machine learning model. The machine learning model has a compute graph representation having nodes that are kernel operators necessary to execute the machine learning model and edges that connect said kernel operators to define precedence constraints. A round of optimization is scheduled for the process amongst the allocated compute nodes. At each allocated compute node a sequencing and scheduling solution is applied per round to obtain a performance metric for the machine learning model. From each compute node the performance metric is received and a solution that has the best performance metric is identified and implemented for execution of the machine learning model on the target edge devices.

Distributed machine learning compiler optimization

In one embodiment, a method of simulating an operation of an artificial neural network on a binary neural network processor includes receiving a binary input vector for a layer including a probabilistic binary weight matrix and performing vector-matrix multiplication of the input vector with the probabilistic binary weight matrix, wherein the multiplication results are modified by simulated binary-neural-processing hardware noise, to generate a binary output vector, where the simulation is performed in the forward pass of a training algorithm for a neural network model for the binary-neural-processing hardware.

Systems and methods for modifying neural networks for binary processing applications

Systems, methods, and devices of the various embodiments enable local visual identification and verification of robotic vehicles. Various embodiments may enable disambiguation of a robotic vehicle from among a plurality of robotic vehicles.

Local drone identification verification

A method performs XNOR-equivalent operations by adjusting column thresholds of a compute-in-memory array of an artificial neural network. The method includes adjusting an activation threshold generated for each column of the compute-in-memory array based on a function of a weight value and an activation value. The method also includes calculating a conversion bias current reference based on an input value from an input vector to the compute-in-memory array, the compute-in-memory array being programmed with a set of weights. The adjusted activation threshold and the conversion bias current reference are used as a threshold for determining the output values of the compute-in-memory array.

Performing XNOR equivalent operations by adjusting column thresholds of a compute-in-memory array

In one embodiment, an electronic device includes a compute-in-memory (CIM) array that includes a plurality of columns. Each column includes a plurality of CIM cells connected to a corresponding read bitline, a plurality of offset cells configured to provide a programmable offset value for the column, and an analog-to-digital converter (ADC) having the corresponding bitline as a first input and configured to receive the programmable offset value. Each CIM cell is configured to store a corresponding weight.

Voltage offset for compute-in-memory architecture

Embodiments include devices and methods for managing network communication of an unmanned autonomous vehicle (UAV). A processor of the UAV may determine an altitude of the UAV. The processor may optionally also determine a speed or vector of the UAV. Based on the determined altitude and/or speed/vector of the UAV, the processor may adjust the communication parameter of the communication link between the UAV and a communication network. The processor may transmit signals based on the adjusted communication parameter, which may reduce radio frequency interference caused by the transmissions of the UAV with the communication network.

Managing network communication of an unmanned autonomous vehicle

An apparatus and method for transmitting an indicator of channel quality while minimizing the use of a broadcast channel is described. A metric of forward link geometry of observed transmission signals is determined. An indicator of channel quality value is determined as a function of the observed transmission signals. An access sequence is selected, randomly, from one group of a plurality of groups of access sequences, wherein each of the plurality of groups of access sequences correspond to different ranges of channel quality values.

Efficient signaling over access channel

Disclosed is an apparatus and method for classifying a motion state of a mobile device comprising: determining a first motion state associated with a highest probability value and with a first confidence level greater than a first threshold; entering the first motion state; while the first motion state is active, determining a second motion state associated with a highest probability value and with a second confidence level greater than the first threshold, the second motion state being different from the first motion state; determining whether the second motion state is to be entered; and in response to determining that the second motion state is to be entered, entering the second motion state.

Smartphone motion classifier

The disclosure relates to drone user equipment (UE) indications that may be conveyed to a wireless network. In particular, a UE that has flight capabilities (i.e., capabilities to operate as an unmanned aircraft system) and optional further capabilities to report a current height level may indicate such capabilities to the wireless network. As such, the wireless network may differentiate the drone UE from other UEs that only operate on the ground. Furthermore, the optional current height level may enable the wireless network to differentiate among drone UEs operating at different heights and/or from other UEs that are operating on the ground. The wireless network may further use the information indicating the flight capabilities either alone or in combination with the optional height information to configure power control parameters, manage interference, provide mobility management functions, generate neighbor lists, control beamforming, or implement a radio resource configuration or management procedure.

Drone user equipment indication

Various embodiments include methods for dynamic antenna steering on an unmanned aerial vehicle (UAV). The methods may include orienting an antenna on the UAV towards a serving ground station based on the UAV's position, orienting the antenna towards a neighboring ground station when it is time to conduct signal measurements of the neighboring ground station, conducting the signal measurements while orienting the antenna towards the neighboring ground station, and reorienting the antenna towards the serving ground station. Methods further include orienting a ground station antenna towards a UAV by obtaining a position of the UAV, calculating a vector between the position of the UAVs and the ground station, determining a direction to steer a beam based on the calculated vector, and steering the beam to the determined direction for the UAV.

Dynamic beam steering for unmanned aerial vehicles

Growing community of inventors

San Diego, CA, United States of America

Edward Harrison Teague