公开(公告)号：WO2021247448A1

公开(公告)日：2021-12-09

申请号：PCT/US2021/035042

申请日：2021-05-29

Applicant: INTEL CORPORATION ,  BALAKRISHNAN, Ravikumar ,  HIMAYAT, Nageen ,  AKDENIZ, Mustafa Riza ,  DHAKAL, Sagar ,  ANAND, Arjun ,  MOSTAFA, Hesham

Inventor： BALAKRISHNAN, Ravikumar ,  HIMAYAT, Nageen ,  AKDENIZ, Mustafa Riza ,  DHAKAL, Sagar ,  ANAND, Arjun ,  MOSTAFA, Hesham

IPC: G06N20/20 ,  H04L29/08

Abstract: The apparatus of an edge computing node, a system, a method and a machine-readable medium. The apparatus includes a processor to cause an initial set of weights for a global machine learning (ML) model to be transmitted a set of client compute nodes of the edge computing network; process Hessians computed by each of the client compute nodes based on a dataset stored on the client compute node; evaluate a gradient expression for the ML model based on a second dataset and an updated set of weights received from the client compute nodes; and generate a meta-updated set of weights for the global model based on the initial set of weights, the Hessians received, and the evaluated gradient expression.

公开(公告)号：WO2022015487A8

公开(公告)日：2022-01-20

申请号：PCT/US2021/039272

申请日：2021-06-26

Applicant: INTEL CORPORATION

Inventor： AKDENIZ, Mustafa Riza ,  HIMAYAT, Nageen ,  BALAKRISHNAN, Ravikumar ,  DHAKAL, Sagar ,  EISEN, Mark R. ,  NADERIALIZADEH, Navid

IPC: H04L12/24 ,  H04L29/08 ,  H04W24/02 ,  G06N20/00

Abstract: In one embodiment, a machine learning (ML) model for determining radio resource management (RRM) decisions is updated, with ML model parameters being shared between RRM decision makers to update the model. The updates may include local operations (between an AP and UE pair) to update local primal and dual parameters of the ML model, and global operations (between other devices in the network) to exchange/update global parameters of the ML model.

公开(公告)号：WO2019132981A1

公开(公告)日：2019-07-04

申请号：PCT/US2017/068955

申请日：2017-12-29

Applicant: INTEL CORPORATION

Inventor： PEREZ-RAMIREZ, Javier ,  BALAKRISHNAN, Ravikumar ,  AZIZI, Shahrnaz ,  STACEY, Robert ,  CARIOU, Laurent

IPC: H04L27/26 ,  H04L1/18 ,  H04L1/08 ,  H04L5/00

Abstract: This disclosure describes systems, methods, and devices related to signal detection. A device may determine a first physical layer protocol data unit (PPDU). The device may store one or more segments of the first Wi-Fi PPDU. The device may send the first Wi-Fi PPDU. The device may identify an acknowledgment frame. The device may send a second Wi-Fi PPDU.

公开(公告)号：WO2021158313A1

公开(公告)日：2021-08-12

申请号：PCT/US2020/067068

申请日：2020-12-26

Applicant: INTEL CORPORATION

Inventor： AKDENIZ, Mustafa Riza ,  ANAND, Arjun ,  HIMAYAT, Nageen ,  AVESTIMEHR, Amir S. ,  BALAKRISHNAN, Ravikumar ,  BHARDWAJ, Prashant ,  CHOI, Jeongsik ,  CHOI, Yang-Seok ,  DHAKAL, Sagar ,  EDWARDS, Brandon Gary ,  PRAKASH, Saurav ,  SOLOMON, Amit ,  TALWAR, Shilpa ,  YONA, Yair Eliyahu

IPC: G06N20/00 ,  H04L12/801 ,  H04L12/927

Abstract: The apparatus of an edge computing node, a system, a method and a machine-readable medium. The apparatus includes a processor to perform rounds of federated machine learning training including: processing client reports from a plurality of clients of the edge computing network; selecting a candidate set of clients from the plurality of clients for an epoch of the federated machine learning training; causing a global model to be sent to the candidate set of clients; and performing the federated machine learning training on the candidate set of clients. The processor may perform rounds of federated machine learning training including: obtaining coded training data from each of the selected clients; and performing machine learning training on the coded training data.

公开(公告)号：WO2017111781A1

公开(公告)日：2017-06-29

申请号：PCT/US2015/000258

申请日：2015-12-23

Applicant: INTEL CORPORATION

Inventor： BALAKRISHNAN, Ravikumar ,  RASHID, Mohammad, Mamunur ,  VANNITHAMBY, Rath ,  GUPTA, Maruti

IPC: H04W76/02 ,  H04W8/18 ,  H04W28/06

CPC classification number: H04W76/11 ,  H04W8/186 ,  H04W28/06 ,  H04W76/12 ,  H04W76/15

Abstract: Methods and apparatuses for communicating in a cellular communications network, including provision of an Evolved Node B (eNB) comprising circuitry to: store a User Equipment (UE) group bearer mapping between: UE identifiers for identification by the eNB of individual UEs; and group bearer UE identifiers (GB-UEIDs) for identifying individual UEs over a group Evolved Packet System (EPS) bearer providing a single EPS bearer to be shared by a plurality of UEs forming an EPS bearer group.

Abstract translation: 用于在蜂窝通信网络中进行通信的方法和装置，包括提供演进型节点B（eNB），所述演进型节点B（eNB）包括电路，以：在以下项之间存储用户设备（UE）组承载映射： 各个UE的eNB; 和用于在提供要由形成EPS承载组的多个UE共享的单个EPS承载的组演进分组系统（EPS）承载上识别各个UE的组承载UE标识符（GB-UEID）。

公开(公告)号：WO2017023233A1

公开(公告)日：2017-02-09

申请号：PCT/US2015/000463

申请日：2015-12-26

Applicant: INTEL CORPORATION

Inventor： BALAKRISHNAN, Ravikumar ,  JHA, Satish, C. ,  SIVANESAN, Kathiravetpillai ,  VANNITHAMBY, Rath

IPC: H04W28/02 ,  H04W52/02 ,  H04W4/00 ,  H04W76/04

CPC classification number: H04W52/0216 ,  H04W4/70 ,  H04W52/0222 ,  H04W76/27 ,  H04W76/28 ,  Y02D70/1224 ,  Y02D70/1262 ,  Y02D70/142 ,  Y02D70/144 ,  Y02D70/164 ,  Y02D70/21

Abstract: An energy savings mode for a cellular communication device, such as an I nternet of Things (IoT) device. Implementing the energy savings state may involve controlling the device to transition to an I DLE state, such as a General Packet Radio Service (GPRS) Mobility Management (GMM) IDLE state, after packet transmission is complete. In one implementation, for devices that are primarily uplink only devices (e.g., IoT devices operating as remote sensors), the device may send periodicity information, relating to the expected period at which the device is l ikely to transmit upl ink data, in a network Attach Request message. Alternatively or additionally, for devices with downlink traffic or both uplink and downlink traffic, the uplink and downlink periodicity values may be provided by the device in the Attach Request message. The network may indicate the length of time the device should spend in the GMM I DLE state (the "wake up period").

Abstract translation: 用于蜂窝通信设备的节能模式，例如物联网（IoT）设备。 执行节能状态可以包括在分组传输完成之后控制设备转换到I DLE状态，例如通用分组无线业务（GPRS）移动性管理（GMM）空闲状态）。 在一个实施方案中，对于主要仅为上行链路设备（例如，作为远程​​传感器操作的IoT设备）的设备，设备可以发送周期性信息，其与设备有意地发送上一个墨水数据的期望周期有关， 网络附加请求消息。 或者或另外，对于具有下行链路业务或上行链路和下行链路业务的设备，上行链路和下行链路周期值可以由附件请求消息中的设备提供。 网络可能指示设备在GMM I DLE状态（“唤醒周期”）中花费的时间长度。

公开(公告)号：WO2022261353A1

公开(公告)日：2022-12-15

申请号：PCT/US2022/032873

申请日：2022-06-09

Applicant: INTEL CORPORATION ,  AKDENIZ, Mustafa Riza ,  ANAND, Arjun ,  BALAKRISHNAN, Ravikumar ,  DHAKAL, Sagar ,  HIMAYAT, Nageen

Inventor： AKDENIZ, Mustafa Riza ,  ANAND, Arjun ,  BALAKRISHNAN, Ravikumar ,  DHAKAL, Sagar ,  HIMAYAT, Nageen

IPC: G06N20/20 ,  G06N3/04 ,  G06N3/08 ,  H04L67/289

Abstract: An apparatus of an edge computing node, a method, and a machine-readable storage medium. The apparatus is to decode messages from a plurality of clients within the edge computing network, the messages including respective coded data for respective ones of the plurality of clients; computing estimates of metrics related to a global model for federated learning using the coded data, the metrics including a gradient on the coded data; use the metrics to update the global model to generate an updated global model, wherein the edge computing node is to update the global model by calculating the gradient on the coded data based on a linear fit of the global model to estimated labels from the federated learning; and send a message including the updated global model for transmission to at least some of the clients.

公开(公告)号：WO2019027495A1

公开(公告)日：2019-02-07

申请号：PCT/US2017/068924

申请日：2017-12-29

Applicant: INTEL CORPORATION

Inventor： AVITAL, Ziv ,  MANSOUR, Avi ,  MIN, Alexander W. ,  FANG, Juan ,  CARIOU, Laurent ,  BALAKRISHNAN, Ravikumar

IPC: H04W52/24 ,  H04W52/42 ,  H04W52/14

CPC classification number: H04W52/248 ,  H04W52/146 ,  H04W52/245 ,  H04W52/247

Abstract: In a multi-user wireless network (for example, a network conforming to wireless standard IEEE 802.11ax), an uplink power control module in an access point may determine Target RSSI and MCS parameters for each STA, transmit those values in a trigger frame, and wait for the solicited responses from the indicated STAs. Based on the RSSI measured in those responses from each STA, the AP may then place values into each Target RSSI, directing the STA to adjust its preceding transmit power up or down by a specified delta amount in the next uplink transmission. After a few passes, the transmitted power may settle down to a steady state value.

公开(公告)号：WO2017180149A1

公开(公告)日：2017-10-19

申请号：PCT/US2016/027765

申请日：2016-04-15

Applicant: INTEL CORPORATION

Inventor： ADRANGI, Farid ,  BALAKRISHNAN, Ravikumar ,  DING, Zongrui ,  RASHID, Mohammad Mamunur

IPC: H04W12/06 ,  H04L9/08 ,  H04L29/06 ,  H04W4/00 ,  H04W8/24

Abstract: Non-pre-provisioned cellular Internet of things (loT) devices can be added to an existing user's subscription with an operator and a service provider. The procedure can include obtaining a security association between a device and a user's smartphone using the operator's network. The operator and the service provider can verify the device with a certificate authority. In one embodiment, the smartphone reads (302) a URL pointer to the device certification and sends it (304) to the MME. The MME forwards (306) the URL to the HSS. The HSS verifies (312) the certificate and derives security credentials including the Master key K'. The HSS also derives another key K" used to establish security context between the loT device and the smartphone. The device uses its key deriving function KDF with K' and Rand to generate K".

Abstract translation: 非预先提供的蜂窝物联网（loT）设备可以被添加到现有用户与运营商和服务提供商的预订中。 该过程可以包括使用运营商的网络获得设备与用户智能手机之间的安全关联。 运营商和服务提供商可以使用证书颁发机构验证设备。 在一个实施例中，智能手机读取（302）指向设备证明的URL指针并将其发送（304）到MME。 MME将URL转发（306）到HSS。 HSS验证（312）证书并导出包括主密钥K'的安全凭证。 HSS还派生另一个关键K“ 用于建立loT设备和智能手机之间的安全上下文。 该设备使用其具有K'和Rand的密钥导出函数KDF来生成K“。

公开(公告)号：WO2017019118A1

公开(公告)日：2017-02-02

申请号：PCT/US2015/065895

申请日：2015-12-15

Applicant: INTEL CORPORATION

Inventor： BALAKRISHNAN, Ravikumar ,  VANNITHAMBY, Rath

IPC: H04W28/06 ,  H04W60/00 ,  H04W76/02

CPC classification number: H04W76/12 ,  H04W28/06 ,  H04W60/00

Abstract: Technology for reducing connection setup time with a network is disclosed. In an example, a mobile station including one or more processors and memory can: encapsulate, within a General Packet Radio Service (GPRS) mobility management (GMM) packet data unit (PDU) attach request message, a Session Management (SM) PDU packet data protocol (PDP) context activation request message and a PDP request indication (PRI) using a primitive from an SM layer to a GMM layer from the mobile station; transmit the GMM PDU attach request message to a Serving GPRS Support Node (SGSN); and receive from the SGSN an SM PDU PDP context activation accept or reject message and a PDP response indication (PRSI) encapsulated within a GMM PDU attach accept or reject response message.

Abstract translation: 公开了用于减少与网络的连接建立时间的技术。 在一个示例中，包括一个或多个处理器和存储器的移动站可以：在通用分组无线业务（GPRS）移动性管理（GMM）分组数据单元（PDU）附着请求消息，会话管理（SM）PDU分组 数据协议（PDP）上下文激活请求消息和使用来自移动台的从SM层到GMM层的原语的PDP请求指示（PRI）; 将GMM PDU附加请求消息发送到服务GPRS支持节点（SGSN）; 并从SGSN接收SM PDU PDP上下文激活接受或拒绝消息和封装在GMM PDU附着中的PDP响应指示（PRSI）附接接收或拒绝响应消息。