Abstract: A communication system includes a subscriber network unit and a communication device provided in an accommodating station. The subscriber network unit includes an acquisition unit that acquires uplink data from one or more lower-layer devices. The communication device includes: a data processing unit that acquires the uplink data from the subscriber network unit using a band of uplink communication and executes data processing on the acquired uplink data; a policy determination unit that determines a policy of band allocation of the uplink communication on the basis of the result of the data processing; and an allocation control unit that allocates the band of the uplink communication to the subscriber network unit on the basis of the policy.

Abstract: In one example, an apparatus comprises: a local on-chip memory; a computation engine configured to generate local data and to store the local data at the local on-chip memory; and a controller. The apparatus is configured to be coupled with a second device via an interconnect, the second device comprising a local memory. The controller is configured to: fetch the local data from the local on-chip memory; fetch remote data generated by another device from a local off-chip memory; generate output data based on combining the local data and the remote data; and store, via the interconnect, the output data at the local memory of the second device.

Abstract: In one embodiment, a method comprises: joining, by a network device, a network topology rooted by a root network device in a data network, and in response transmitting an advertisement indicating a position of the network device in the network topology; suppressing a second transmission based on initiating a deferred transmission operation in response to transmitting the advertisement; maintaining the deferred transmission operation to enable a prescribed minimum number of other network devices to join the network topology at respective identified lower positions than the position of the network device; and changing, by the network device, from the deferred transmission operation to an accelerated operation in response to expiration of a prescribed deferral interval or detecting the prescribed minimum number of other network devices having the respective identified lower positions, the accelerated operation enabling the network device to initiate transmission of a data packet before the other network devices.

Abstract: An improved channel access mechanism provides collision avoidance improves throughput. A communications device, e.g. an end node in a long range wireless communications system, determines that data is to be transmitted and performs a channel activity detection operation to determine if a communications channel is busy or idle. In response to determining that the channel is busy, the communications device determines a wait time based on a random number and a number of times a busy channel has been detected since determining that data is to be transmitted. The communications device uses the determined wait time in determining when to transmit.

Abstract: Systems and methods for enforcing consistency of file to storage node mappings in distributed file systems. An example method may include: receiving a request to perform a file operation with respect to a file associated with a directory of a distributed file system; identifying a directory layout version number associated with the directory; identifying a volume layout version number associated with a volume in which the directory resides; and responsive to determining that the directory layout version number matches the volume layout version number, performing the requested file operation with respect to the file.

Abstract: Embodiments include systems and methods for managing a network of nodes with delta processing. A network with a plurality of connected nodes can be stored, the nodes representing entities of the network. One or more deltas to the network that indicate updates to the connections among the plurality of nodes can be received. The deltas can be added to a queue. The connections between the nodes of the network can be updated using a delta from the queue. The network can be processed after updating the connections between the nodes to generate a vector for a given node. One or more paths for the given node can be identified based on the generated vector.

Abstract: Embodiments of the present disclosure include techniques for processing neural networks. Various forms of parallelism may be implemented using topology that combines sequences of processors. In one embodiment, the present disclosure includes a computer system comprising one or more processor groups, the processor groups each comprising a plurality of processors. A plurality of network switches are coupled to subsets of the plurality of processor groups. In one embodiment, the switches may be optical network switches. Processors in the processor groups may be configurable to form sequences, and the network switches are configurable to form at least one sequence across one or more of the plurality of processor groups to perform neural network computations. Various alternative configurations for creating Hamiltonian cycles are disclosed to support data parallelism, pipeline parallelism, layer parallelism, or combinations thereof.

Abstract: A routing server may receive a request to establish a connection from an originating node to a terminating node. The routing server may determine a route for the connection based on data in the request. Determining the route may comprise selecting at least two of the originating node, at least one virtual node, and the at least one terminating node. The originating node and/or at least one virtual node may establish at least one rule for routing data therethrough. The routing server may establish the connection between the originating node and the at least one terminating node. The routing server may route the data between the originating node and the at least one terminating node according to the at least one rule.

Abstract: An all-to-all communication operation which is carried out in a fabric of networked entities by defining in each of the entities a plurality of memory regions of contiguous memory addresses holding messages therein, and exchanging the messages repeatedly with all the other entities. Relatively small messages are copied using a CPU and larger messages are transmitted using scatter/gather facilities.

Abstract: A method for configuring and executing card content management (CCM) operations in a declarative manner includes composing a CCM operation declaration, wherein each CCM operation includes one or more CCM scripts and storing the CCM operation declaration in memory. When provisioning is needed, applicable scripts for the CCM operation declaration are fetched from the memory. An execution context needed for each script in the CCM operation declaration is prepared. The scripts are executed in an order specified in the CCM operation declaration.

Abstract: A computing node includes an active Non-Volatile Random Access Memory (NVRAM) component which includes memory and a sub-processor component. The memory is to store data chunks received from a processor core, the data chunks comprising metadata indicating a type of post-processing to be performed on data within the data chunks. The sub-processor component is to perform post-processing of said data chunks based on said metadata.

Abstract: A method is provided for sharing digital information between first and second mobile devices. The method may include, at an embedded device, receiving the digital information from the first mobile device; at the embedded device, notifying the second mobile device of the digital information via a push notification; and at the embedded device, maintaining communication with the first and second mobile devices. In some embodiments, the method further includes, at the embedded device, receiving an embedded device wide area network (WAN) Internet protocol (IP) address. Maintaining communication with the first and second mobile devices may include notifying the first and second mobile devices of the new embedded device WAN IP address via respective additional push notifications, or using a DDNS server.

Abstract: Embodiments of the invention provide a method, system and computer program product for message merging in a messaging queue. In an embodiment of the invention, a method for message merging in a messaging queue can be provided. The method can include receiving a request to add a new message to a message queue in a message queue manager executing in memory by a processor of a host computing platform. The method can also include a merge indicator to stipulate whether or not a merge should take place. The method also can include identifying an association key associating the new message with an existing message in the message queue and locating an associated message in the message queue corresponding to the identified association key. Finally, the method can include merging the new message with the located associated message in the message queue.

Abstract: In one embodiment, a device determines a set of sources and used destinations for traffic in a computer network, where nodes of the network are configured to send all traffic to the used destinations through a root node of the computer network according to a directed acyclic graph (DAG). The device may then also determine a set of capable nodes as common ancestors to source-destination pairs that provide a more optimal path between the source-destination pairs than traversing the root node, and instructs the set of capable nodes to store downward routes to forward traffic for one or more of the used destinations according to the stored downward route rather than through the root node.

Abstract: In a method for managing transfer of data from a source machine cluster to a destination machine cluster, information relevant to the transfer of data from the source machine cluster to the destination machine cluster is accessed. In addition, a data transfer operation that substantially optimizes the transfer of the data based upon the accessed information is determined. Furthermore, the determined data transfer operation is implemented to transfer the data from the source machine cluster to the destination machine cluster.

Abstract: A method of monitoring symptoms of a person includes repeating, over a period of time, the steps of: selecting, by the person, one or more symbolic representations corresponding to one or more symptoms from a predefined set of symbolic representations presented to the person; and electronically recording data regarding the one or more symbolic representations selected by the person such that the data is electronically accessible later for generating a history of the symptoms of the person over the period of time. The data is transferred over virtual networks using virtual dispersive routing and stored in dispersive storage area networks (SANs). The data is classified as trusted, corrupted, invalid or uncertain, and/or as trusted, ambiguous and inadequate.

Abstract: Disclosed are systems, hybrid compute environments, methods and computer-readable media for dynamically provisioning nodes for a workload. In the hybrid compute environment, each node communicates with a first resource manager associated with the first operating system and a second resource manager associated with a second operating system. The method includes receiving an instruction to provision at least one node in the hybrid compute environment from the first operating system to the second operating system, after provisioning the second operating system, pooling at least one signal from the resource manager associated with the at least one node, processing at least one signal from the second resource manager associated with the at least one node and consuming resources associated with the at least one node having the second operating system provisioned thereon.

Abstract: A database system may process multiple column-oriented tasks in parallel for a database being stored according to a row-partitioning protocol. The database system may determine when the query should process the column-oriented task serially or in parallel. For parallel processing, the database system may generate processing tasks for each unique column-oriented task contained in the query used to retrieve column data and to process the column data according to the column-oriented tasks requested. A method of operating the database system may determine that multiple column-oriented tasks included in a query are to be processed in parallel. The method may further include generating a processing task for each unique column-oriented task included in the query. The method may further include performing the column-oriented tasks in parallel based on the processing threads. The method may implement various considerations in determining to process to the column-oriented tasks in parallel.

Abstract: Each server in a server group of a data center can run a data collection agent. The agent can collect data from a respective server in the server group. The data can include a performance characteristic (i.e., performance data) associated with the respective server. If the performance characteristic falls outside an allowable range, the agent can broadcast this information to other agents at other servers in the server group. If the other agents at the other servers in the server group detect a similar performance characteristic outside the allowable range, they can broadcast as well. If there is a sufficiently high quantity of broadcasts, then that can indicate a high likelihood that there is a potential problem with respect to servers in the server group. The problem can be reported to the central controller, which can then handle the problem appropriately (e.g., issue an alarm, contact support technician, etc.).

Abstract: A network device may be configured to define a Layer 2 domain, wherein the Layer 2 domain defines a data link layer broadcast domain and assign an access interface to the Layer 2 domain. The network device may be further configured to assign a double Virtual Local Area Network (VLAN) identifier to the access interface in the Layer 2 domain, wherein the double VLAN identifier includes an inner VLAN identifier and an outer VLAN identifier, receive a data unit via the access interface, wherein the data unit includes the double VLAN identifier, and associate the data unit with the defined Layer 2 domain, based on the data unit being received via the access interface and based on the data unit including the assigned double VLAN identifier.

Abstract: A system for performing storage operations using hierarchically configured storage operation cells. The system includes a first storage manager component and a first storage operation cell. The first storage operation cell has a second storage manager component directed to performing storage operations in the first storage operation cell. Moreover, the first storage manager component is programmed to instruct the second storage manager regarding performance of storage operations in the first storage operation cell.

Abstract: A calculation processing apparatus includes a monitor, a CPU, a memory and a hard disk. The hard disk stores an initial program input from outside, a network creation program, a network modifying program, network information, node operation definition, and learning variables. The CPU executes the network creation program, and creates network information related to a network representing an algorithm structure of the initial program. Further, the CPU executes the network modifying program and modifies the network information based on the result of calculation by the network, using a learning algorithm.

Abstract: For a network that includes several managed edge switching elements and several managed non-edge switching elements that are for implementing a logical switching element, some embodiments provide a method of distributing packet processing across the several managed non-edge switching elements. The method receives a packet for processing through the logical switching element. Based on a determination that the packet needs to be processed by a managed non-edge switching element, the method determines a particular managed non-edge switching element of the several managed non-edge switching elements to forward the packet. The method forwards the packet to the particular managed non-edge switching element for the particular managed non-edge switching element to process the packet.

Abstract: Some embodiments provide a network architecture that includes several lower level managed switching elements for forwarding network data to several of network hosts. The network architecture includes a set of higher level managed switching elements. The several lower level managed switching elements and the set of higher level managed switching elements implement several logical datapath sets. Communication channels are established among the several lower level managed switching elements and the set of higher level managed switching elements based on a mesh topology.

Abstract: A novel method for logically routing a packet between a source machine that is in a first logical domain and a destination machine that is in a second logical domain is described. The method configures a managed switching element as a second-level managed switching element. The method configures a router in a host that includes the second-level managed switching element. The method communicatively couples the second-level managed switching element with the router. The method causes the router to route a packet when the router receives a packet from the first logical domain that is addressed to the second logical domain.

Abstract: In order to create a data gathering/data processing device for video/audio signals which includes a plurality of signal processors and which has an optimized hardware architecture, it is provided that the signal processors or a subset of the signal processors are coupled to a network having a star-shaped topology.

Abstract: In a hierarchical switching architecture that includes at least one lower level managed switching element that connects to several higher level managed switching elements, some embodiments provide a method of identifying a higher level managed switching element to which the lower level managed switching element forwards a packet for further processing. The method computes a value based on a set of attributes of the packet. The method identifies a record from a hierarchy traversal table based on the computed value. The record specifies (1) a first higher level managed switching element as a primary higher level managed switching element and (2) a second higher level managed switching element as a secondary higher level managed switching element. The primary and secondary higher level managed switching elements are for forwarding the packet for further processing. The method forwards the packet to one of the higher level managed switching elements.

Abstract: Identifying a collective operation within an application executing on a parallel computer; identifying a call site of the collective operation; determining whether the collective operation is root-based; if the collective operation is not root-based: establishing a tuning session and executing the collective operation in the tuning session; if the collective operation is root-based, determining whether all compute nodes executing the application identified the collective operation at the same call site; if all compute nodes identified the collective operation at the same call site, establishing a tuning session and executing the collective operation in the tuning session; and if all compute nodes executing the application did not identify the collective operation at the same call site, executing the collective operation without establishing a tuning session.

Abstract: Identifying a collective operation within an application executing on a parallel computer; identifying a call site of the collective operation; determining whether the collective operation is root-based; if the collective operation is not root-based: establishing a tuning session and executing the collective operation in the tuning session; if the collective operation is root-based, determining whether all compute nodes executing the application identified the collective operation at the same call site; if all compute nodes identified the collective operation at the same call site, establishing a tuning session and executing the collective operation in the tuning session; and if all compute nodes executing the application did not identify the collective operation at the same call site, executing the collective operation without establishing a tuning session.

Abstract: Embodiments include a network switch for use in a local area network that interconnects the switch and one or more network powered devices, the switch also being operatively connected to at least one power supply that powers the switch, the switch comprising a plurality of ports to which network powered devices are connected, at least one management module for controlling the operation of the switch, the module having memory for storing information relating to the power and operating requirements of each connected network device, and a processor for controlling the distribution of power from the power supply among the ports, including selectively time slicing power provided to at least selected ones of the ports.

Abstract: Described is a method for improving both the bandwidth efficiency and bit rate of data streams flowing through binary tree networks. In an embodiment, the invention provides a method for efficiently utilizing end-consumer computers possessing a variety of upstream bandwidth capabilities to deliver online streaming video in a distributed manner.

Abstract: The efficiency of scaling of a network of computing devices can be improved by grouping highly-connected portions into deployment units wherein devices only have to be aware of other devices in the respective deployment unit. The various deployment units can be connected by a backbone mechanism. In some embodiments, the backbone utilizes a torroidal connection scheme to connect the backbone switches and pass routing protocol information. Host traffic can still utilize connections between tiers, such that standard routing and networking protocols can be utilized. In other embodiments, logical interfaces can be initialized in each backbone switch, such that each switch itself can function as a point backbone. The devices of the various deployment units then can communicate with each other via the point backbones using standard networking and routing protocols.

Abstract: Known network management systems for managing traffic signals in a network store node/link parameter signals and traffic parameter signals and calculate solutions defining intermediate nodes/links situated between sources and destinations for transporting said traffic signals from said sources to said destinations via said intermediate nodes/links and calculate route information per solution. By storing node/link parameter signals and traffic parameter signals for several situations (like several predictions in the future) and calculating solutions for each situation and then selecting a solution per situation, the network management system will manage the network more advantageously, especially when solutions are compared with each other, and when a solution is selected per situation in dependence of comparison results (comprising similarities/differences between solutions of different situations).

Abstract: To reduce the delivery delay of video data to a display device in which a control right for adjusting parameters of the video data is vested, a video delivery apparatus modifies a delivery tree for delivering the video data to multiple display devices, and delivers the video data to the multiple display device via a network in accordance with the delivery tree.

Abstract: The invention provides an Ethernet-based data transmission method, which is applied to a control system with a plurality of nodes. The method comprises: a first node caches the data needed to be sent (S101); if the current communication macro-cycle comprises the time segment for the first node to send data, the first node sends the cached data in the time segment for the first node to send data within the current communication macro-cycle; if the current communication macro-cycle does not comprise the time segment for the first node to send data, the first node does not send the cached data in the current communication macro-cycle; wherein other nodes do not send data in the time segment for the first node to send data within the current communication macro-cycle, and the communication macro-cycles have same durations and each communication macro-cycle consists of a plurality of time segments with a fixed length (S102). The invention further provides Ethernet nodes and a control system.

Abstract: In one embodiment, a source routing device (e.g., root device) pre-computes diverse source-routed paths to one or more nodes in a computer network. Upon receiving a particular packet, the device forwards the particular packet on a source-routed first path of the pre-computed diverse paths. In the event the device implicitly detects failure of the first path, then it forwards a copy of the particular packet on a source-routed second path of the pre-computed diverse paths in response. In one embodiment, implicit failure detection comprises seeing a second (repeated) packet with the same identification within a certain time since the first packet, and the second packet is forwarded on the second path. In another embodiment, implicit failure detection comprises not seeing a link-layer acknowledgment returned or receiving an error notification from a node along the broken path, and a stored copy of the particular packet is forwarded on the second path.

Abstract: A multi-address space mobile network architecture is disclosed, in which edge networks (ENs) are connected to a core network (CN) through core edge routers (CERs). A distributed hash table (DHT) set in the CN is adapted to not only store or update a mapping relation between a CER ID and CER location information in the CN, but also receive registration information sent from a host through the CER, and store or update mapping relations between a host ID host identifier (HI) and the CER ID and between the HI and host location information. A method for registering host information includes the steps as follows. The host acquires routable host location information in an EN to which the host currently added and information of the CER in the EN. The host sends registration information to the CN through the CER in the EN. Further, a method for sending data is described. The network architecture and methods of the present invention may improve update and query efficiency of the registration information from the host.

Abstract: A server creates a communication path table indicating a communication path for packet data, based on connection information reported by a parent device and a relay device, and distributes the created communication path table to the parent device, the relay device and a child device. The child device transmits the packet data to a relay device indicated in the communication path table, based on the distributed communication path table. The relay device transfers the packet data transmitted from the child device, to a relay device or a parent device indicated in the communication path table, based on the distributed communication path table.

Abstract: According to one aspect of the present disclosure, a method and technique for hierarchical network failure handling in a clustered node environment is disclosed. The method includes: detecting a network failure by a node in a cluster, the cluster having plural nodes arranged in a hierarchy, wherein the network failure is associated with a subordinate node in the hierarchy to the detecting node; communicating the network failure from the detecting node to a superior node in the hierarchy; determining whether the network failure affects nodes higher than the detecting node in the hierarchy; and responsive to determining that the network failure does not affect nodes higher than the detecting node in the hierarchy, the detecting node initiating a protocol to expel the subordinate node from the cluster.

Abstract: In one embodiment, a method is provided. The method of this embodiment provides if an OOB (out-of-band) state machine has not received a confirmation from an operating system that the operating system is functional, the operating system using a primary IP (internet protocol) address, then if the OOB state machine is in a passive state, transitioning to an active state and obtaining an OOB IP address if the OOB state machine currently has no OOB IP address; and receiving one or more first packets using the OOB IP address.

Abstract: A network apparatus comprising a trunk end point associated with an Ethernet-tree (E-Tree) service in a network domain and configured to forward a frame that comprises a tag according to the tag in the frame, wherein the tag in the frame is a root tag that indicates a root source of the frame or a leaf tag that indicates a leaf source of the frame, and wherein the trunk end point is coupled to a second end point associated with the E-tree service outside the network domain.

Abstract: An automated service discovery and monitoring utility employs automatically generated policies (rules) for monitoring network health of a service based on network discovery using flow data derived from transmissions sent over segments between computing components defining the service. An interactive discovery application employs flow data based on transmissions associated with a service. Analysis of the flow data identifies associated components from inspection of the transmissions. Iterative component selection allows network traversal of components based on flow data of the segments used for providing the service.

Abstract: In one embodiment, the present invention includes a component to be coupled to a peripheral device to enable the peripheral device to appear to be locally connected to a computer of a local area network, although the peripheral device is not physically connected to the computer. The component may include a first set of registers to store a mirrored copy of control register information present in a second set of registers of a host controller interface of the computer. Other embodiments are described and claimed.

Abstract: The invention relates to an image processing system for an x-ray installation or such like having at least one image source, a plurality of computing units for processing different algorithms, at least one non-volatile storage and a data sink, with the image source, the plurality of computing units, the non-volatile storage and the data sink each being physically connected to at least one control unit operating as a circuit module in a physical network, as a result of which a star-shaped logical topology can be configured by the control unit.

Abstract: Methods, apparatuses, and computer program products are provided for supporting remote hosting without using network address translation. A method may include supporting, at a hub, a plurality of redundant tunnel end points for each of a plurality of spokes using a technology based at least in part on multipoint generic routing encapsulation protocol. The method may further include using policy based routing to determine a destination spoke for outgoing data traffic. The method may additionally include using service level agreement tracking to select a tunnel end point from the plurality of redundant tunnel end points for the determined destination spoke. The method may also include causing the outgoing data traffic to be routed to the determined destination spoke via the selected tunnel end point. Corresponding systems, apparatuses and computer program products are also provided.

Abstract: A method and user device for limiting a time for which location data sharing is enabled for a user device of a data sharing group, the data sharing group comprising at least two user devices inclusive of the user device, each user device of the data sharing group configured to store data shared by user devices of the data sharing group on the respective user device and to maintain a list of the user devices in the data sharing group. The method comprises receiving an instruction through a user interface on the user device indicating that location data sharing is to be enabled for a limited time period after which location sharing is to be disabled, obtaining location data for the user device during the limited time period; and sending the location data to at least one other user device of the sharing group during the limited time period.

Abstract: Internet-protocol based, virtual private networks are configured by first identifying the basic components of such networks and then efficiently assigning MPLS-based RD and RTs to each component or efficiently generating VR-based access lists for each component.

Abstract: A method to organize nodes into an overlay network and to disseminate multicast messages within this overlay both through an efficient tree embedded in the overlay and through gossips exchanged between overlay neighbors. Regardless of the size of the system, this invention incurs a constant low overhead on each node.

Abstract: A method comprises distributing a plurality of descriptors of file encoded with comprising a plurality of recipient nodes, wherein at least one descriptor is distributed from the first node to each recipient node of the at least a portion of the first group. The at least a portion of the first group communicate their respective descriptors received from the first node to other nodes of the first group. A system comprises an origin node operable to distribute all of a plurality of descriptors of a MDC file to a first group of recipient nodes, wherein the origin node does not attempt to communicate all of the plurality of descriptors to all of the recipient nodes of the first group. The recipient nodes of the first group are each operable to communicate a descriptor that it receives from the origin node to other nodes of the first group.

Abstract: A hierarchical and distributed system architecture for a container monitoring and security system is provided. The architecture may be a hierarchical chain of separate, related processing elements. The partitioning of functions and distribution of processing among these or other similar hierarchical elements in the network is provided. The elements may further be described in successive layers, each have a greater level of network intelligence than the former.