Mitsubishi Electric Research Laboratories, Inc.

Mitsubishi Electric Corporation

Renesas Electronics Corporation

Other

Samsung Electronics Co., Ltd.

Intel Corporation

Motorola Corporation

Nxp B.v.

Mitsubishi Electric Research Labs.

Mitsubishi Denki Kabushiki Kaisha

Philip Orlik

Jianlin Guo

Jinyun Zhang

Kieran Parsons

Pu Wang

Milutin Pajovic

Kyeong Jin Kim

Toshiaki Koike-Akino

Andreas F Molisch

Zafer Sahinoglu

Raymond Yim

Yukimasa Nagai

Petros T Boufounos

Chunjie Duan

Zhifeng Tao

Takenori Sumi

Karl Berntorp

Yebin Wang

Hassan Mansour

Ramesh Annavajjala

Toshiaki Koike Akino

Stefano Di Cairano

Ghulam Bhatti

Kyeong-Jin Kim

Koon Hoo Teo

Dehong Liu

David Millar

Heejin Ahn

Wataru Matsumoto

Bingnan Wang

Wataru Tsujita

Rui Ma

Ye Wang

Zhenzhen Ye

Man-On Pun

Lu Zhang

Retsu Sugawara

Frederick J Igo, Jr

Kota Sadamoto

Toshiyuki Kuze

Qinghe Du

Kyeongjin Kim

Fei Ye

Yehan Ma

Haoyu Fu

Evripidis Paraskevas

Jun Ma

Gang Yao

Jianyuan Yu

Ron Porat

Vahid Tarokh

Edgar Herbert Callaway, Jr

Lakshman Krishnamurthy

Ralph Michael Kling

Ye Li

Neelesh B Mehta

Jasmeet Chhabra

Yong Liu

Jonathan S Yedidia

Avishai Weiss

Chia Shen

Ankush Chakrabarty

Vernon Anthony Allen

Jay Bao

Chunhui Zhu

Daqing Gu

Tommy C Poon

Chungwei Lin

Chiori Hori

Lance Eric Hester

Dirk Brinkman

Jyhchau Horng

Marcel Menner

Marcus Greiff

Yanting Ma

Shigeru Uchida

Myung Jong Lee

Daniel Antonio Da Costa Dinis

Yacong Ding

Huai-Rong Shao

Xuhui Hu

Gal Shulkind

Amitav Mukherjee

Masashi Saito

Kaikai Liu

Atulya Yellepeddi

Koike-Akino Toshiaki

Phil A Jamieson

Kentarou Sawa

Talha Ahmed Khan

Raj Tejas Suryaprakash

Toshiaki Akino

Xin Yang

Vikrant Shah

Arindam Bose

Jesper H Soerensen

Zhiyuan Weng

Ke Xu

Koike Akino Toshiaki

Liangbin Li

A method for training a neural network for extracting domain-invariant features suitable for image registration comprises collecting a first set of multi-modal images including at least a first image of a first modality and at least one corresponding second image of a second modality. Using feature extraction subnets of the neural network, first features are extracted from the at least one first image while second features are extracted from the at least one second image. A domain invariant loss and homography loss is estimated for the images and the neural network is trained to minimize a multi-objective loss function including the domain-invariant embedding loss and the homography loss.

Systems and methods for processing multi-modal images

A tracking system for tracking an expanded state of an object is provided. The tracking system executes, for a predetermined time period, a probabilistic filter that iteratively tracks a belief on the expanded state of the object, wherein the belief is predicted using a motion model of the object and is further updated using a compound measurement model of the object. After the predetermined time period, the updated beliefs are smoothed to generate a state-decoupled online batch of training data. The compound measurement model includes multiple probabilistic distributions constrained to lie around a contour of the object with a predetermined relative geometrical mapping to the center of the object. The compound measurement model is updated using the online batch of training data. Further, the tracking system tracks the expanded state of the object based on the updated compound measurement model.

System and method for tracking an expanded state of a moving object using an online adapted compound measurement model

A frequency modulation continuous wave (FMCW)-based system configured to convert measurements of a linearly modulated wave from a time-domain into a frequency-domain to produce a non-linear frequency signal, where the non-linear frequency signal comprises a known linear component representing the desired linear modulation and an unknown non-linear component representing the non-linearity of the modulation. The FMCW-based system is further configured to determine coefficients of a basis function approximating a difference between the non-linear frequency signal and the linear frequency component in the frequency domain. The FMCW-based system is further configured to detect one or multiple spectrum peaks in the distorted beat signal with the distortion compensated according to the basis function with the determined coefficients to determine one or multiple distances to the one or multiple objects in the scene.

Frequency modulation continuous wave (FMCW)-based system for estimation and compensation of nonlinearity in linearly swept sources

A computer-implemented method is provided for training a global machine learning model using a learning server and a set of vehicle agents connected to roadside units (RSUs). It involves selecting agents, associating them with RSUs based on proximity, and transmitting training data and deadlines. Agents train models locally, which are then aggregated through RSUs to refine the global model to desired precision. The method includes steps of selecting vehicle agents from a pool of the vehicle agents connected to the RSUs, associating the selected vehicle agents and the RSUs respectively based on distances from the selected vehicle agents to the RSUs configured to provide measurements of the distances to the learning server, and transmitting a global model, a selected agent set and deadline thresholds in each global training round to the RSUs configured to transmit the global model and training deadlines to the selected vehicle agents.

Communication and computation aware distributed machine learning for vehicular networks

A controller is provided for generating carrier signals controlling a pulse width modulation (PWM) inverter driving an electric actuator. The controller includes an interface configured to connect to a victim circuit via sensors, wherein the victim circuit includes the power system or the electric actuator or a combination of the power system and the electric actuator, wherein the sensors are configured to measure an electromagnetic-interference (EMI) spectrum, a memory configured to store modulation frequency ranges, the measured electromagnetic-interference (EMI) spectral data of different frequency carriers, a desired EMI spectrum and a learning-based carrier design program, and a processor, in connection with the memory, configured to perform generating a frequency modulation (FM) carrier signal by solving an optimization problem, wherein the optimization problem is generated regarding a sweep time for a predetermined frequency by the learning-based carrier design program, and a PWM generator configured to generate PWM signals based on the FM carrier signal.

Electromagnetic interference (EMI) mitigation in pulse width modulation (PWM) inverters using learning-based frequency modulated carriers

An optical coherence tomography (OCT) system comprises an interferometer configured to split incident light into a reference beam and a test beam, and to interfere the test beam reflected from the specimen with the reference beam reflected from a reference mirror to produce an interference pattern. The OCT system also comprises a spectrometer configured to analyze spectral components of the interference pattern at non-uniformly sampled wavenumbers. A computer-readable memory of the OCT system is configured to store a measurement model with elements connecting different depth values with different non-uniformly sampled wavenumbers and weighted with weights derived from a power spectral density (PSD) of the incident light for corresponding wavenumbers. The OCT system further comprises a processor configured to determine the profilometry measurements of the specimen as a maximum likelihood estimate of the specimen surface depth by back-projection of the measured intensities with the measurement model.

Systems and methods for surface profile estimation via optical coherence tomography

Embodiments of the present disclosure disclose a method and a system for tracking the position of the one or more moving objects. The method includes collecting GNSS measurement data of satellite signals transmitted from multiple satellites. The method further includes extracting values of a plurality of features from the GNSS measurement data. The method includes mapping the extracted values of the plurality of features to a source domain. The method includes classifying the mapped transformed plurality of features using a neural network. The neural network is trained over simulated data sampled from a source domain. The method includes identifying multipath measurements of the GNSS measurement data based on classification of the corresponding mapped transformed plurality of features. The method includes tracking the position of the one or more moving objects by processing identification of GNSS measurement data affected by multipath.

Method and system for position estimation using domain adaptation

Provided herein is a method for controlling a manipulator, comprising accepting an initial pose of a load and a task for moving the load and retrieving using a mapping function, an identification trajectory corresponding to the initial pose of the load and controlling a plurality of actuators of the manipulator to move the load according to the retrieved identification trajectory and obtaining measured motion data and estimated motion data of the load each corresponding to motion of the load. The method further comprises estimating parameters of the load based on the measured motion data and the estimated motion data, obtaining a model of the manipulator having the load with the estimated parameters, and determining a performance trajectory to move the load according to the task based on the obtained model of the manipulator. The method further comprises controlling the actuators to move the load according to the performance trajectory.

Systems and methods for flexible robotic manipulation by fast online load estimation

A computer-executed method is provided for IEEE 802.15.4 devices based on a suspendable carrier-sense multiple access with collision avoidance (CSMA/CA) control program and standard CSMA/CA control program for an IEEE 802.15.4 network composing of IEEE 802.15.4 devices. The computer-executed method is provided on an IEEE 802.15.4 device, and causes a processor of the IEEE 802.15.4 device to perform steps that include determining the permission of backoff suspension and the intention of IEEE 802.15.4 device to perform backoff suspension, selecting the suspendable CSMA/CA control program if the backoff suspension is permitted and IEEE 802.15.4 device intends to perform backoff suspension. The suspendable CSMA/CA control program is configured to perform active CCA within each unit backoff period and suspend backoff if channel is detected to be busy, performing a CCA when backoff completes, transmitting frame when the detected channel status is an idle state or incrementing a number of backoff (NB) when the detected channel status is an busy state, determining if a NB exceeds the macMaxCSMABackoffs, incrementing a number of retransmissions (NR) when a NB exceeds the macMaxCSMABackoffs, and discarding frame when a NR exceeds macMaxFrameRetries.

Suspendable CSMA/CA for IEEE 802.15.4 system to reduce packet discard caused by backoff failure

The present disclosure provides a multiple-input multiple-output (MIMO) radar system and a method for detecting an object in a scene. The method comprises transmitting frequency modulated continuous wave (FMCW) in a radio frequency (RF) band, and collecting radar measurements of the scene sampled in a time-frequency domain within an intermediate frequency (IF) bandwidth. The method further comprises transforming the radar measurements into range-doppler space to produce measurements of different segments of the scene for different range-doppler bins formed by intersections of different range bins with different Doppler bins, classifying a presence of the hypothetical transmitter at different segments of the scene according to a signal model with an internal classification, combining the results of the classification to produce parameters of the object, and outputting the parameters of the object.

System and method for detecting an object in a scene

A wireless smart utility network (Wi-SUN) device is provided for coexistence with a Wi-Fi HaLow network sharing frequency spectra between the networks. The Wi-SUN device includes a receiver receiving packets of neighbor Wi-SUN devices, a memory storing computer executable programs including a hybrid carrier-sense multiple access with collision avoidance (CSMA/CA) control program and Wi-SUN backoff control program, a processor to execute steps of estimating a severity of Wi-Fi HaLow interference based on one or combination of the severity metrics, selecting a CSMA/CA mode between predetermined CSMA/CA modes in response to the estimated severity, detecting a channel status based on the hybrid carrier-sense multiple access, wherein if the channel status is not idle, a maximum limited number of times for re-attempting a packet transmission is checked to determine an allowability of re-attempting the packet transmission, and a transmitter to transmit packets according to a determination result of the allowability.

Hybrid carrier sense multiple access system with collision avoidance for IEEE 802.15.4 to achieve better coexistence with IEEE 802.11

A computer-implemented method is provided for discovering heterogenous neighbors in coexisting IoT networks including at least one Wi-Fi device and at least one of Zigbee coordinators, ZigBee routers and ZigBee end devices. The method includes generating a broadcast packet such that the broadcast packet emulates a ZigBee broadcast frame, transmitting the emulated broadcast packet using a transceiver of the at least one Wi-Fi device according to cross-technology communication (CTC) method, wherein the emulated broadcast packet is configured to trigger the at least one of the Zigbee coordinators and ZigBee routers having received the emulated broadcast packet to rebroadcast the received packet. The method also includes generating a unicast packet such that the unicast packet emulates a ZigBee address request frame, transmitting the emulated unicast packet using a transceiver of the at least one Wi-Fi device according to cross-technology communication (CTC) method, wherein the emulated unicast packet is configured to trigger the at least one of the Zigbee coordinators, ZigBee routers and ZigBee end devices having received the emulated unicast packet to respond with a ZigBee address response frame, and determining the at least one ZigBee end device to be a neighbor of the at least one Wi-Fi device if the scanned address response frame is transmitted by the at least one ZigBee end device.

Active cross-technology neighbor discovery

A Bluetooth-enabled device is provided, the Bluetooth-enabled device being configured to control a radio frequency (RF) chain during a first period to receive at a single antenna selected from a plurality of antennas, a constant tone extension (CTE) signal of multiple frames transmitted by the Bluetooth-enabled transmitter over multiple frequencies. Further the RF chain is controlled during a second period to switch among the plurality of antennas to receive the CTE signal at each of the plurality of antennas. An initial time-of-flight (ToF) data of the CTE signal is determined from first samples of the CTE signal received during the first period. Further, the Bluetooth-enabled transmitter is localized with respect to a location of the Bluetooth-enabled device using a signal model connecting samples of the CTE signal with an unknown angle-of-arrival of the CTE signal received at specific times, an unknown ToF conditioned on the initial ToF data.

Bluetooth-based spatial-temporal localization

A method for semi-persistent scheduling (SPS) based resource allocation in a private 5G network comprises receiving location data of a plurality of user equipment (UEs) in the private 5G network in a last subframe of respective current communication periods of the UEs and determining distance distribution of the UEs in a subsequent communication period of a candidate UE of the plurality of UEs, based on the respective location data of the UEs. The method further comprises determining signal to noise ratio expectation of the UEs for the subsequent communication period, solving an optimization problem optimizing an expectation of data to be carried by each of the one or more SPS channels in the subsequent communication period, to obtain scheduling result for each of the one or more SPS channels and generating resource allocation data for the candidate UE based on the scheduling result.

Systems, apparatuses, and methods for semi-persistent scheduling in communication networks

Growing community of inventors

Cambridge, MA, United States of America

Philip Orlik