Nec Usa, Inc.

General Electric Company

Rca Inc.

Rutgers, the State University of New Jersey

Samsung Electronics Co., Ltd.

Semandex Networks Inc.

Dipankar Raychaudhuri

Ruixi Yuan

Daniel Jorge Reininger

Gopalakrishnan Ramamurthy

Rajiv S Dighe

Arup Acharya

Jun Li

Hai Xie

Dong-seek Park

Joel Walter Zdepski

Young-Kwon Cho

Charles M Wine

Glenn A Reitmeier

Kuriacose Joseph

Chang-Ho Suh

Yunxin Liu

Seok-Hyun Yoon

Marco Gruteser

Leonard N Schiff

Maximilian A Ott

Subir K Biswas

Leslie John French

Yanyong Zhang

Leonid Razoumov

Zoran Miljanic

Wuyang Zhang

Charles Martin Wine

Glen Arthur Reitmeier

Computer vision systems and methods for acceleration of high-resolution mobile deep vision with content-aware parallel offloading are provided. The system allows for acceleration of deep vision applications running on mobile devices via offloading of machine learning computations on edge cloud infrastructure, and speeds up computation via parallel processing on the edge cloud, employing one or more of the following processes: recurrent region proposal prediction, region proposal centric frame partitioning, and resource-aware offloading to multiple processors. Mobile deep vision applications can utilize the systems and methods disclosed herein to accelerate application processing. The system offloads computation to multiple servers to minimize the end-to-end latency, and includes a high-resolution mobile deep vision task acceleration system that offloads computation to multiple (edge located) servers.

Computer vision systems and methods for acceleration of high-resolution mobile deep vision with content-aware parallel offloading

Disclosed are the design and implementation of a CSI feedback channel in a multi-carrier wireless communication system. An apparatus and a method for channel state feedback using arithmetic coding are provided to ensure efficiency and reliability of a system by transmitting the CSI while compressing the CSI with a predetermined compression rate selected depending on a channel state. The apparatus for CSI feedback in a wireless communication system performing channel estimation at a transmitter or a receiver by using a communication channel includes a transmitter terminal transmitting a signal for CSI measurement by using the communication channel, and a receiver terminal receiving the signal from the transmitter terminal, checking a channel state based on the received signal, and transmitting the signal to the transmitter terminal after compressing the signal according to the channel state.

Method and apparatus for channel state feedback using arithmetic coding

The present invention changes the wireless information service paradigm from a connection-oriented unicast network model to a dynamic content-driven multicast model using a technique called 'content routing.' The present invention provides a system, method, and medium for routing content through a network from at least one content provider to at least one content consumer, wherein at least one of the content consumer or content provider communicate with the network using a wireless device. The content is routed though the network based on its content. The content is labeled with at least one content descriptor. The content descriptor characterizes or otherwise describes the content. Interest profiles are received from content consumers. Interest profiles from content consumers with a common network access point are aggregated into a single interest profile. Interest profiles can include more than the wireless consumers stated interest, such as for example, a content consumer's geographic location, the capabilities of the wireless terminal, type of service, policy preferences, and the like. Single interest profiles are shared across neighboring network access points to create a content routing table. Content, labeled with at least one descriptor, is routed through the network to content consumers based, in part, on content routing tables created from wireless content consumers'profiles.

Dynamic content based multicast routing in mobile networks

A handoff control process in a wireless ATM network replaces an old communications connection with a new communication connection. In order to guarantee that no data is lost during the replacement process, an ATM cell level mechanism is used to re-schedule the buffering and transmitting of data streams of the virtual channels (VCs) to be handed-off. In addition, this mechanism is transparent to user applications. The present invention performs three fundamental cell level scheduling functions. The first function is to mark and redirect cell transmission with operation and maintenance (OAM) cells. The second function is to disable and buffer cell transmission until the new path is connected. The third function is to enable cell transmission, starting with the buffered cells across the connected new path.

Handoff method for an ATM wireless network wherein both the switch and

Mobility in a wireless ATM network is accomplished by use of hand-off control protocols. A mobile terminal signals a first base station that a hand-off is to occur. In a first preferred embodiment the first base station signals a second base station requesting a hand-off. After the second base station signals the first base station that a datapath link is available from a hand-off switch to the second base station, the hand-off switch causes the datapath to change from the first base station to the second base station and the first base station signals the mobile terminal to commence communication with the second base station. In a second preferred embodiment, the mobile terminal signals a first base station that a hand-off is to occur. The first base station signals a second base station requesting a hand-off. At the same time a datapath link is established between the first and second base stations. When the second base station signals the first base station that the datapath link is available, the first base station signals the mobile terminal to change to the second base station and data is sent from the first base station to the second base station via the available established datapath link.

Handoff-control technique for wireless ATM

A multimedia communications and computer platform that can serve as a network interface card combined with an internal distribution network for a full range of user terminal devices. It includes an interconnection network module that serves to route all incoming and outgoing information by way of high speed buses with value added features for communication protocol acceleration. In particular, attached to the different ports of the network are an ATM module, a communications processor, a media processor and a variety of terminal devices. By migrating processing intensive functions of network protocol termination, media stream distribution and media stream adaptation into the network interface card, there are avoided the bottlenecks of the traditional CPU centric approach to ATM systems.

Multimedia network interface for asynchronous transfer mode

A data link control procedure for wireless ATM access channels based on a dynamic TDMA/TDD framework provides integrated ATM services including available bit-rate (ABR) data and constant/variable bit-rate (CBR/VBR) voice or video through the addition of wireless-specific medium access control and data link control protocol layers between the physical and ATM network layers. Generally, a data link control is used to insulate the ATM network layer from wireless channel impairments by selective retransmission of erroneous or lost cells before they are released to the ATM layer. The data link control methods disclosed use the on-demand ABR burst transmission capability of the dynamic TDMA channel to retransmit unacknowledged cells in available slots not allocated to service data. Specific error recovery procedures for both (asynchronous) ABR and (isochronous) CBR services are provided.

Data link control protocols for wireless ATM access channels

A system for variable bit-rate video coding in which encoding bandwidth as characterized by a usage parameter control (UPC) parameters is renegotiated between a video encoder and an asynchronous transfer mode network in order to maintain quality-of-service and save bandwidth. The coding system includes adjusting the video source quantization in a manner for controlling the occupancy level of a buffer while new UPC parameters are requested from an ATM network.

VBR MPEG video encoding for ATM networks with dynamic bandwidth

A medium access control (MAC) layer protocol is used in a wireless ATM system for integrated support of ATM services, including constant bit-rate (CBR), variable bit-rate (VBR) and available bit-rate (ABR) services. The MAC protocol supports both connectionless packet and connection-oriented virtual circuit modes, with appropriate service parameter and quality-of-services selection. A dynamic time division, multiple access/time division multiplex (TDMA/TDM) approach accommodates the service classes in an integrated manner. A supervisory MAC procedure integrates ATM ABR/VBR/CBR virtual circuits, providing burst-by-burst allocation of ABR cells and call-by-call allocation of VBR and CBR bandwidth parameters. A mechanism is provided for the dynamic allocation of subframe capacities, assignment of ABR slots based on desired queueing rules, assignment of VBR slots based on ATM traffic shaper parameters and assignment of CBR slots based on bandwidth requirements.

Multiservices medium access control protocol for wireless ATM system

An asynchronous transfer mode (ATM) traffic control framework is based on an integrated usage parameter control (UPC) approach, which approach provides a unified and scalable solution to the issue of quality-of-services (QOS) levels over a range of anticipated services in ATM based networks. The approach is consistent with emerging ATM Forum and CCITT standards. Additionally, a UPC-based call and burst admission control providing the desired QOS over periods of network overload by call/burst admission control and traffic shaping of source stream preferably uses a dual leaky bucket.

UPC-based traffic control framework for ATM networks

Apparatus for encoding/decoding a HDTV signal for e.g., terrestrial transmission includes a compression circuit responsive to high definition video source signals for providing ierarchically layered codewords CW representing compressed video data and associated codewords T, defining the types of data represented by the codewords CW. A priority selection circuit, responsive to the codewords CW and T, parses the codewords CW into high and low priority codeword sequences wherein the high and low priority codeword sequences correspond to compressed video data of relatively greater and lesser importance to image reproduction respectively. A transport processor, responsive to the high and low priority codeword sequences, forms high and low priority transport blocks of high and low priority codewords respectively. Each transport block includes a header, codewords CW and error detection check bits. The respective transport blocks are applied to an forward error check circuit for applying additional error check data. Thereafter the high and low priority data are applied to a modem wherein they quadrature amplitude modulate respective carriers for transmission.

An HDTV compression system

A CDMA network includes a retransmission scheduler which selectively delays the retransmission of an initially erroneously sent signal. The selected delays have an average value in a range of values which results in a single equilibrium point between input messages and output messages to avoid a bistable condition that can lower overall system throughput.

Stabilization of random access packet CDMA networks

In a communication system serving a plurality of transmitter-receivers, high capacity is achieved by asynchronous contention access of relatively short reservation request packets, which if successfully transmitted result in the scheduled, noncontention transmission of relatively long message data packets. In a first embodiment, the reservation request packets bear information relating to the duration of the message data packets, so that appropriate message data transmission time may be scheduled. Unsuccessful or colliding reservation request packets are retransmitted. In a second embodiment, the reservation requests packets do not include duration information, and the message data packets have fixed duration. In the second embodiment, collision resolution techniques allow the message data scheduling without retransmission of some colliding reservation request packets. It is also possible to operate in both modes, by scheduling random length message data transmissions as a result of noncolliding reservation request packets, and by scheduling fixed duration message data transmissions when reservation request packets collide.

Random access communication system with scheduled data transmission and

A random access communication system forms messages to be transmitted into subpackets of fixed duration. A message packet includes two or more subpackets. Each subpacket includes information relating to the source transmitter-receiver and the number of subpackets in the message. The subpackets of the message packet are transmitted contiguously over the transmission path in a contention mode. Some subpackets are successfully received and some are lost due to collisions. After a period of asynchronous contention operation, a predetermined number of subpackets are successfully received, and the system switches to a short-term synchronous, scheduled operating mode. In the scheduled operating mode, those transmitter-receivers which were the sources of message packets including some successful and some unsuccessful subpackets retransmit the data from the unsuccessful subpackets, ordering themselves in accordance with the order of receipt of successful subpackets. The system capacity is high, average delay is short and continuous global synchronization is not required.

Random access communication system with contention scheduling of

An asynchronous random access system includes stations which in one embodiment transmit information as packets at random times. After a path delay, all stations receive the packets. By comparing the period of activity at the station with the known packet length, all stations know when a collision occurs. According to the system protocol, all station inhibit new transmissions for a predetermined time which includes a retransmission time. Stations whose packets were involved in the collision know that their transmission was unsuccessful, and they can also determine whether their own packets were first or last in the collision, based upon the known path delay. Retransmission of the first and last packets occurs during the retransmission interval and is guaranteed to be successful, because all other stations are inhibited. In another embodiment, evaluation of the received data is also performed to establish which packets are second and/or penultimate in collision, and they are also retransmitted according to a prescribed schedule.

Growing community of inventors

Princeton Junction, NJ, United States of America

Dipankar Raychaudhuri