Abstract: One or more flow control modules, implemented on various types of network topologies, provide a number of functionalities for controlling the flow of IP packets (such as TCP/IP packets) over a network connection. The flow control modules may be implemented within a sender and/or receiver or may be deployed into a network as a separate device without requiring significant additional resources.
Abstract: A method of constructing a forwarding database for a data communications network comprising a plurality of network components and supporting at least first and second topologies having one or more common network components is described. The forwarding database includes a plurality of entries providing forwarding information for data destined for a related network component. The method comprises the step, performed at a constructing network component, of deriving topology dependent forwarding information for data destined for a network component. The method further comprises the step of including the topology dependent forwarding information at the related forwarding database entry.
Type:
Grant
Filed:
August 23, 2005
Date of Patent:
July 17, 2012
Assignee:
Cisco Technology, Inc.
Inventors:
Neil Alasdair James Jarvis, Crispin Nicholas Dent-Young, Alan J. Cheilek, Matthew James Collins, Christian Cassar, Jesper Skriver
Abstract: A network having a hierarchical structure includes a plurality of clusters. Each of the clusters includes a plurality of nodes and a cluster head connected to the nodes in a mesh format. Each of the nodes is connected to other clusters via the cluster head.
Type:
Grant
Filed:
October 23, 2008
Date of Patent:
July 17, 2012
Assignees:
Electronics and Telecommunications Research Institute, Fumate Co., Ltd
Inventors:
Hyeon Park, Young Hwan Ham, Sang Joon Park, Jeong Man Woo
Abstract: Systems, methods, and computer program products for processing messages and searching for locating messages in an asynchronous messaging network by using content based message tracking. The message descriptor of each message is extended by a content key generated by the message generating application. The content key provides search information derived from application data contained in the message body. Each message queuing system is extend by a message trace component that administers a message trace queue and adds entries in the message trace queue for each operation of a message, and by a message search component that provides the functionality to search for the current location of a specified message as well as to retrieve a trace of the movement of the message.
Type:
Grant
Filed:
July 26, 2005
Date of Patent:
July 17, 2012
Assignee:
International Business Machines Corporation
Abstract: To enable the peer-to-peer synchronization among multiple devices, each device in the peer network keeps track of the changes it makes to any shared data and also keeps track of its own knowledge of the changes made by other devices. When two or more peer devices communicate, they share with each other their information about any changes made to the shared data by them or by other devices. This allows the devices to synchronize with each other to the extent that each of the devices knows what changes have been made by it or by other devices in the peer network.
Type:
Grant
Filed:
March 15, 2010
Date of Patent:
July 3, 2012
Assignee:
Hewlett-Packard Development Company, L.P.
Inventors:
Srikiran Prasad, Rik Sagar, Frank Judge, Gurmeet Singh Kalra
Abstract: The present invention provides frame-interleaving systems and methods for Optical Transport Unit K (OTUK) (i.e. Optical Transport Unit 4 (OTU4)), 100 Gb/s Ethernet (100 GbE), and other 100 Gb/s (100 G) optical transport enabling multi-level optical transmission. The frame-interleaving systems and methods of the present invention support the multiplexing of sub-rate clients, such as 10×10 Gb/s (10 G) clients, 2×40 Gb/s (40 G) plus 2×10 G clients, etc., into two 50 Gb/s (50 G) transport signals, four 25 Gb/s (25 G) transport signals, etc. that are forward error correction (FEC) encoded and carried on a single wavelength to provide useful, efficient, and cost-effective 100 G optical transport solutions today. In one exemplary configuration, a 100 G client signal or 100 G aggregate client signal carried over two or more channels is frame-deinterleaved, followed by even/odd sub-channel FEC encoding and framing.
Abstract: A method for multicasting targeted advertising data to end user devices is disclosed. In an illustrative embodiment the method includes but is not limited to correlating at a regional internet protocol television (IPTV) server, multicast queue data with web data from web sites visited by each of a plurality of end user devices associated with the regional IPTV server; selecting a subgroup of end user devices from the plurality of end user devices to receive targeted advertising data correlating with a member of the multicast queue data; and multicasting targeted advertising data associated with the member of the multicast queue data to the subgroup of end user devices from the regional IPTV server. A system is also disclosed for multicasting targeted advertising data to end user devices. A data structure is disclosed for storing data used by the system and method.
Abstract: An Internet protocol (IP)-based ubiquitous sensor network (USN) system is disclosed. The IP-based USN includes a plurality of sensor nodes having unique IP addresses and an IP-USN router. The IP-USN router has a table that stores each location address of the sensor nodes and periodically broadcasts its own presence information for updating the location addresses stored in the table. The sensor nodes in the IP-based USN system have unique IP addresses. Therefore, desired information and services from any external network may be obtained.
Type:
Grant
Filed:
February 3, 2007
Date of Patent:
June 26, 2012
Assignee:
Ajou University Industry—Academic Cooperation Foundation
Inventors:
Seung-Wha Yoo, Byeong-Hee Roh, Hong-Jeon Ha, Ki-Hyung Kim
Abstract: A motion vector searching apparatus to which a reference image, an object image and weight parameters for the reference image are inputted and searches a motion vector based thereon is provides. The motion vector searching apparatus includes an inverse weighting section that generates an inverse weighted object image by performing, on the object image using the weight parameter, a weighting inverse from a weighting performed on the reference image using the weight parameter; an image comparing section that compares the generated inverse weighted object image with the reference image to obtain a comparison value indicating a degree of approximation between the two images; and a determination section that determines an optimum motion vector from the comparison value.
Abstract: A signal separation unit (101) separates an input picture made up of component pictures of RGB, each of which has an equal number of pixels as the input picture, into three component pictures, and outputs the three component pictures. Each of coding units (102 to 104) codes one of the component pictures into an intra-picture prediction coded picture or an inter-picture prediction coded picture, and outputs a bit stream corresponding to the component picture. A bit stream multiplexing unit (105) multiplexes three bit streams outputted from the three coding units into one bit stream, and outputs the bit stream. Each of the coding units (102), (103) and (104) determines a prediction method for the component picture at the time of coding.
Abstract: Methods and systems for processing a signaling message are disclosed. An exemplary method comprises: determining a first transcoding policy associated with an originator endpoint contained in a received first call offer; determining a second transcoding policy associated with an answerer endpoint contained in the first call offer; applying the first transcoding policy to a first offer codec set in the first call offer to produce a second call offer containing a second offer codec set; applying the second transcoding policy to the second codec set to produce a third call offer containing a third offer codec set; comparing a first answerer codec set in a received first answer and a second answerer codec set contained in the second call offer; and determining whether or not to perform transcoding based on a result of the comparison.
Type:
Grant
Filed:
September 15, 2006
Date of Patent:
June 19, 2012
Assignee:
Acme Packet, Inc.
Inventors:
Patrick John MeLampy, Robert F. Penfield, Clifford Spencer, David Michael Gray
Abstract: A packet processing apparatus includes a packet buffer with a queue for storing packets. An actual queue length/position discriminator acquires, at every sampling period, the latest actual queue length indicating the occupancy status of the queue, determines the positional relationship of the actual queue length to a random early detection interval, and outputs the positional relationship as position information. A discard probability computation processor calculates, at every sampling period, a packet discard probability based on the position information. A packet discard processor discards, at every sampling period and in accordance with the discard probability, packets that are not yet stored in the queue. If it is judged from the position information that the actual queue length is within the random early detection interval, the discard probability computation processor calculates an average queue length, and then calculates the discard probability from the ratio of a discard target to a reception target.
Abstract: An in-band inspection device receives a packet and/or frame, and inspects the packet and/or frame “on the fly.” The device can accept, reject, and/or modify the received packet and/or frame in accordance with the inspection. The device can operate in a fast mode wherein the received packet and/or frame can be transmitted as soon as, for example, a port address in the packet and/or frame is received, or in a slow mode wherein the received packet frame is buffered until, for example, at least the CRC of the packet and/or frame is beginning to be received.
Type:
Grant
Filed:
June 8, 2006
Date of Patent:
June 5, 2012
Inventors:
Slawomir K. Ilnickl, Martin Curran-Gray
Abstract: In one embodiment, a network device receives on a first port a first spanning tree protocol (STP) control message including a first path-tracking field corresponding to a given spanning tree instance in a network. The first path-tracking field includes a value based on one or more other network devices that have propagated the first STP control message. The network device receives on a second port a second STP control message including a second path-tracking field corresponding to the given spanning tree instance. The second path-tracking field includes a value based on one or more other network devices that have propagated the second STP control message. The network device utilizes the values from the first path-tracking field and the second path-tracking field to select a root port for the given spanning tree instance.
Type:
Grant
Filed:
December 8, 2009
Date of Patent:
May 8, 2012
Assignee:
Cisco Technology, Inc.
Inventors:
Francois Edouard Tallet, Norman W. Finn
Abstract: A computer readable medium comprising software instructions for managing resources on a host, wherein the software instructions comprise functionality to: configure a classifier located on a NIC, to forward packets addressed to a first destination address to a first HRR mapped to a first VNIC, wherein packets addressed to the first destination address are associated with a first PFC lane; configure the classifier to forward packets addressed to a second destination address to a second HRR, wherein packets addressed to the second destination address are associated with a second PFC lane; and transmit, by the first VNIC, a pause frame associated with the first PFC lane to a switch operatively connected to the physical NIC, wherein the switch, in response to receiving the pause frame, stores packets associated with the first PFC lane in a buffer without transmitting the packets.
Type:
Grant
Filed:
May 29, 2009
Date of Patent:
May 8, 2012
Assignee:
Oracle America, Inc.
Inventors:
Sunay Tripathi, Nicolas G. Droux, Kais Belgaied
Abstract: Systems and methods are provided for communicating with control objects according to a singular and unified address space. In one aspect, an industrial communications system is provided. The system includes a global address protocol that can be employed to communicate with local or remote networks. An industrial protocol is adapted to interface with the global address protocol, where a network component communicates with one or more control components via the industrial protocol and in accordance with an address supplied by the global address protocol.
Type:
Grant
Filed:
September 30, 2005
Date of Patent:
May 8, 2012
Assignee:
Rockwell Automation Technologies, Inc.
Inventors:
Kenwood H. Hall, Scot A. Tutkovics, David A. Vasko, Brian A. Batke
Abstract: Methods and apparatus for communicating IP datagrams efficiently within communication links employing variable frame sizes bundles IP datagrams into data payloads. The number of datagrams that can fit within a current frame size is determined, a bundle header indicating that the bundle header is associated with a bundled IP datagram is generated that includes information for unpacking bundled IP datagrams. The bundle header is joined to one or more IP datagrams to form a data payload that is communicated as a frame payload without indicating in a frame header that the frame includes bundled IP datagrams. Information in the bundle header enables a receiver processor to extract the IP datagrams from the frame payload without requiring modifications to the frame header. A bundle header may precede each IP datagram, or one bundle header may include a map IP datagrams in the frame payload.
Abstract: A communication system according to various exemplary embodiments can include a broadcast station. At least one encoder can be coupled to the broadcast station, wherein the at least one encoder is configured to encode in real-time signals generated from the broadcast station and to convert the signals into Transmission Control Protocol/Internet Protocol (TCP/IP) data packets. At least one server can be coupled to the at least one encoder for transmitting the TCP/IP data packets over a virtual connection between the at least one server to a content distribution network.
Abstract: A method of decoding a video bitstream is disclosed. The method generally includes the steps of (A) buffering the video bitstream in a main memory, the video bitstream comprising a first residual block based on a first motion compensated block, the first motion compensated block having been generated by a subpixel motion compensation using an N-tap interpolation on a plurality of first reference samples, (B) copying a first subset of the first reference samples from the main memory to an internal memory, (C) generating a first decode block using the subpixel motion compensation with an M-tap interpolation on the first subset in the internal memory, and (D) reconstructing a first original block by adding the first residual block to the first decode block.
Abstract: A network device includes a control unit configured to execute an extended layer three (L3) routing protocol within a control plane of the network device. The extended L3 routing protocol is extended to process a routing protocol control message having a set of layer two (L2) identifiers of customer devices attached to an autonomous system (AS) and an indicator that specifies a Route Target associated with a Virtual Private Local Area Network Service (VPLS) domain within the AS. The routing protocol control message conforms to a message type of the L3 routing protocol to initiate a flush of the set of L2 identifiers.