Abstract: A privilege elevation flaw detection analysis is performed on a host system on a network. In addition, accounts on the host system are identified that have access to, or corresponding accounts on, other systems on the network. Privilege elevation analyses are performed on one or more of the network systems corresponding to the identified accounts. A privilege elevation graph is generated of the host system from the privilege elevation analysis. The graph includes account nodes and edges illustrating the detected privilege elevations between the accounts on the host system. In addition, nodes for the network systems are added to the graphs along with edges connecting to the nodes corresponding to the accounts identified as having access to the particular network systems. The user may then select a particular network system node and view its detected privilege elevations in relation to the host system.

Abstract: A network comprising a plurality of trees each comprising at least one ingress leaf node, at least one interior node, and at least one egress leaf node, wherein at least some of the ingress leaf nodes and the egress leaf nodes are common to the trees, and wherein the ingress leaf node is configured to transport data to the egress leaf node using any of the trees is disclosed. Also disclosed is a network component comprising a processor configured to implement a method comprising selecting one of a plurality of trees associated with information contained within a frame, directing the frame to the selected tree, and maintaining a filtering database (FDB) entry in an interior node in the unselected tree or trees.

Abstract: A computer apparatus comprising a processor and a forwarding engine arranged to forward LDP multicast traffic along a multicast tree having a primary and a backup path in a converged network topology, the processor being configured to cause the forwarding engine to forward traffic via the backup path upon a topology change and send a changed topology label and path vector to at least one neighbor node in the changed topology.

Abstract: A system and method for enabling high-speed, low-latency global collective communications among interconnected processing nodes. The global collective network optimally enables collective reduction operations to be performed during parallel algorithm operations executing in a computer structure having a plurality of the interconnected processing nodes. Router devices ate included that interconnect the nodes of the network via links to facilitate performance of low-latency global processing operations at nodes of the virtual network and class structures. The global collective network may be configured to provide global barrier and interrupt functionality in asynchronous or synchronized manner. When implemented in a massively-parallel supercomputing structure, the global collective network is physically and logically partitionable according to needs of a processing algorithm.

Abstract: A method and associated apparatus may implement spontaneous topology discovery in a computer network. The network may include a plurality of entities. When a change in the topology is detected, a spanning tree is created and topology information is propagated from individual entities through the spanning tree towards an entity that functions as a “root.” Arbitrarily large topologies may be discovered this way without requiring arbitrarily large amounts of memory in participating switches.

Abstract: The invention is directed to techniques for maintaining a map of node relationships for a network of nodes (e.g., network of computers). In one example, the map of node relationships represents relationships overlaying and typically different from the network of physical connections among the nodes. Each child node periodically checks in with its parent nodes, and the parent nodes can thus determine when a child node has terminated a relationship with the parent or created a new relationship with a new parent. Changes in relationships propagate upward through the network of nodes so that each node maintains a map of the relationships among the descendants of that node. A root node receives the propagated change relationship information and maintains a map of the entire network and valid pathways through the network.

Abstract: A Point-of-Sale networked computing system includes a workstation with a first Personal Computer (PC)-based controller, a workstation with a second industrial PC-based controller having different performance characteristics than the first controller, and an Ethernet network coupling the workstations. The Ethernet network may include home-run or star Ethernet topology, and the controllers may be TCP/IP addressable. The computing system may be used in a restaurant, kitchen, shop, or department store. The computing system may also include a bump bar keyboard or keypad and a monitor coupled to the controllers Another Point-of-Sale networked computing system includes at least two workstations, each of which includes an industrial non-PC-based controller, and an Ethernet network coupling the workstations.

Abstract: A method, apparatus and computer program product for building multipoint Pseudo Wires (PWs) using Multicast Label Distribution Protocol (MLDP) is presented. A multipoint PW leaf identifier (mTAI) is configured on one or more Layer 2 (L2) leaf interface(s) of a multipoint PW. A multipoint PW root identifier identifying the root of the multipoint PW is also configured and an opaque value is computed, the opaque value used in MLDP label mapping messages that will flow upstream towards a the root of the multipoint PW. The opaque value is computed from the multipoint leaf identifier and the multipoint root identifier. The PE leaf routers direct MLDP label mapping to the root PE router via a core network.

Abstract: A network including a hierarchical structure of nodes is described. The structure of nodes includes n layers including n?1 layers of switch nodes and 1 layer of computational nodes. Each layer in the structure can include mn-L nodes grouped into units, where m represents a number of nodes in a unit and L represents a layer in the structure, where L=0 represents a lowest layer and L=n?1 represents a highest layer. Each node in a layer other than the computational layer can include a switch node for a unit in a next lower layer. For each unit, each node can be connected to each other node by a point to point link and to a local switch node for the unit by a point to point link. Each node can be connected to each other node and to the local switch node by a local broadcast network for the unit.

Abstract: A Kitchen Display System (KDS) includes a workstation including a first Personal Computer (PC)-based controller, a fanless workstation including a second PC-based or a non-PC-based controller having lower cost than the first PC-based controller, an Ethernet network operatively coupling the controllers, and a bump bar, keyboard, touchscreen, or keypad operatively coupled to at least one of the controllers. The first PC-based controller stores and displays video, multimedia, or a build card. At least one of the controllers includes a rugged computing module, which includes a circuit board including an outer perimeter and traces associated therewith, an integrated circuit mounted on the circuit board, at least one interface connector, a housing, and a thermal transfer device.

Abstract: Methods, systems, and computer program products are provided for managing contact proxies and security proxies in networks that are organized as forests. For instance, contact proxies may be generated to represent user accounts and groups in forests other than the home forests of the user accounts and groups. Security proxy objects may be generated to represent group members (e.g., security principals and groups) in groups in forests other than the home forests of the group members. Furthermore, when both a contact object and a security proxy object exist for a member added to a group, one of the contact object or the security proxy object may be selected to represent the member in the group.

Abstract: Various systems and methods for rerouting multicast traffic in response to detecting imminent network disruption are disclosed. One method involves detecting an imminent topology change and, in response, identifying a new multicast distribution tree for a multicast group. A join message for the multicast group is then sent towards a root of the new multicast distribution tree. Multicast traffic addressed to the multicast group continues to be forwarded via the current multicast distribution tree, subsequent to sending the join message. The multicast traffic is not forwarded via the new multicast distribution tree until one or more multicast data packets have been received via the new multicast distribution tree.

Abstract: A method and apparatus for providing redundancy for an access network are disclosed. For example, the method receives an access network structure, and removes one or more unnecessary 2D edges or 3D edges from a simplex cover for the access network structure. The method then performs a search on the simplex cover; and, generates a new access network structure that includes the redundancy.

Abstract: Certain exemplary embodiments provide a method that can comprise establishing a tunnel between a monitor and at least one router; sending a message to join a multicast transmission; and/or transmitting a packet via the tunnel to a router. The packet can comprise a source address of the network monitor and a destination address comprising a multicast address.

Abstract: A method for implementing multicast fast reroute includes: determining a route for a backup LSP according to the tree topology of a point-to-multipoint primary LSP and establishing the backup LSP according to the determined route for the backup LSP. In the process of establishing the backup LSP, messages for the primary LSP and its corresponding backup LSPs are further merged to decrease the number of network signaling states. A node includes a transceiver unit and a backup LSP path establishing unit. The present invention can save protection bandwidth, optimize network resource utilization, and enhance utilization efficiency of resources.

Abstract: Dynamically upsizing and/or downsizing a network farm in response to network demand. An application message router routes messages to the network farm. When the network farm approaches or is anticipated to be approaching capacity, a group of one or more servers may be added to the network farm. When the added server(s) are capable of participating in the network farm, the application message router is triggered to route also to the added servers. When the network farm has excess capacity, a group of one or more servers may be dropped from the network farm. This may be accomplished by triggering the application message router to no longer route messages to the removed servers. The removed servers may be either immediately or gracefully removed from service.

Abstract: A system and method for preventing a count-to-infinity problem in a network using epochs of sequence numbers in protocol messages to eliminate stale protocol information in the network and to allow the forwarding topology to recover in merely one round trip time across the network.

Abstract: A method may include receiving a reconfiguration to a first Virtual Local Area Network (VLAN)/spanning tree table, where the first VLAN/spanning tree table has a first identifier and is associated with a region of a network; updating the first VLAN/spanning tree table to generate a second VLAN/spanning tree table based on the reconfiguration; determining a second identifier of the second VLAN/spanning tree table; and generating a list of identifiers associated with the region of the network, the list including the first identifier and the second identifier.

Abstract: The attaching of labels to an OPEN frame and applying label switched routing to SAS expanders is disclosed to eliminate the need for large routing tables in SAS networks. A label stack is inserted into the OPEN frame by the initiator, prior to the OPEN frame being transmitted. Each label contains the egress port for a SAS expander in the transmit path. Each SAS expander to be participating in the connection reads the labels to determine the egress port to connect to and through which data is to be sent. The SAS expander marks its label invalid or discards it and forwards the OPEN frame to the egress port where the next SAS expander will look for the first valid label. The process repeats until the OPEN frame reaches the edge device, at which time all labels are discarded and the OPEN frame is forwarded to the end device.

Abstract: Embodiments of the invention provide a method for building a classification tree based on a plurality of rules. Each rule includes multiple fields having different priority levels. The classification tree is built to provide for single pass classification of a received packet according to the multiple rules. The classification tree is built level-by-level, where each level represents one of the priority levels. Building a single level of the classification tree includes: creating, for each of the logical decision paths received from a next higher priority level, a respective sub-tree. The sub-tree includes only nodes and related logical decision paths that are necessary to satisfy rules associated with a respective next higher priority level node. Building of the single level further includes: creating, for each non-lowest priority level, logical decision paths from nodes within the current priority level to a next lower priority level.

Abstract: A method is described for forwarding a tunneled packet in a data communications network. A header of the tunneled packet is removed at a forwarding node neighboring the tunnel end point, and the payload is forwarded.

Abstract: A method for voice based communication between a number of mobile terminals of a telecommunication network, the method comprising the steps of creating a group of said number of terminals and storing group information in each of the number of terminals, recording a voice message in one of the terminals of the group, dispatching the voice message to the terminals of the group from the one of the terminals, and receiving the voice message at least all other terminals of the group. The step of creating a group of terminals further comprises the steps of coupling together a number of mobile terminals, initiating a group creation command from at least one of the terminals, sending predetermined information from terminal to terminal to each terminal of the number of terminals for forming a group information, and transmitting the group information to all of the terminals thus forming a group of the terminals.

Abstract: The present invention relates to a digital multimedia network of apparatuses each comprising a control device, wherein a device parameter of an apparatus is controlled by sending a command message (CMD) to said control device of said apparatus containing a tree-structured hierarchical parameter address (HPA) which consists of parameter grouping identifiers each corresponding to a hierarchy level of a predetermined tree-structured parameter hierarchy used for addressing device parameters throughout said digital multimedia network.

Abstract: A mechanism is provided to configure a plurality of transport trees in a transport network, each of which correspond to a native tree (e.g., a bidirectional multicast tree). In embodiments of the present invention, each of the plurality of transport trees has a unique root node so that in the event of a failure of any root node, the transport trees with surviving root nodes can be used to transport traffic from the native tree. The present invention provides for each transport network edge router being independently responsible for selection of a transport tree that the edge router will use to transmit a datastream, while also being capable of receiving packets from any transport tree. Through the use of such configured transport trees along with independent selection of a transport tree, the present invention provides a reduction in the disruption of datastream transmission due to a root node failure.

Abstract: In order that a loop does not occur when switching upstream terminal apparatus, to provide an information distribution system capable of preemptively detecting a distribution route loop, and avoiding the loop, a terminal apparatus and a program used in the same system, as well as a loop connection avoidance method. A terminal apparatus B which has received loop connection detection information from upstream determines whether or not a downstream terminal apparatus of a control path which is not a distribution route along which contents information is flowing, and a source of the loop connection detection information match and, in the event that they match, determines that a loop will occur. By this means, it is possible to determine in advance that a loop will occur, before switching an upstream connection destination.

Abstract: An Internet user transfers directly to a domain within an e-community without returning to a home domain or reauthenticating by providing to a web browser by a home domain server a home identity cookie with an extensible data area and an enrollment token; performing enrollment through an e-community for a web-browser user by redirecting the home identity cookie via the web browser to each affiliated domain in the e-community until each has been visited once; responsive to each visit to each affiliated domain, sending an affiliated domain identity cookie to the web browser including an enrollment successful indicator; accumulating received enrollment success indicators in the extensible data area of the home identity cookie; and subsequently, vouching for an identity of the user at an affiliated domain through exchange of a vouch-for request and vouch-for response between the home domain server and an affiliated domain server.

Abstract: Resource allocations in stream data processing systems are performed to allocate the resources in a cost-effective manner by formulating the resource allocation problem as a linear program. For a problem with a single output, a backward algorithm or method is used that produces an optimal solution in linear time. For a problem with multiple outputs and the network has a tree structure, a backward shrink algorithm or method is used to provide an optimal solution in linear time. These algorithms are fully distributed, they require only local exchange of parameters between neighboring nodes, and are adaptive to the dynamic changes in network conditions and flow rate fluctuations.

Abstract: A grand computer network is formed from layer 2 (L2) networking technology in which groups of Provider L2 bridges are organized into formations, and different formations are interconnected via network-network interface (NNI) links. Customer sites are coupled to the formations. Customers identify their traffic, e.g., frames, by labeling or tagging it with a Customer Virtual Local Area Network (VLAN) Identifier (C-VLAN ID) or Customer Service Instance (CSI). Within the formations, the C-VLAN ID is mapped to a Service VLAN ID (S-VLAN ID) or Provider Service Instance (PSI), and the S-VLAN ID is appended to the customer traffic. The PSIs are hierarchical, such that each PSI belongs to at most one other “outer” or higher-level PSI, but may itself own any number of “inner” or lower-level PSIs. As a given frame traverses through the different formations of the Grand Network via the NNI links, the frame acquires an encapsulation, sheds an encapsulation or exchanges its current encapsulation for a different one.

Abstract: In a hierarchical tree-based protection scheme, a node in a mesh network is designated as a root node of a spanning hierarchical protection tree and subsequently invites each adjacent node to become its child within the tree. If the inviting node provides a more capacious protection path to the root node than is currently enjoyed by the invitee, the invitee designates the inviting node as its primary parent and assumes a new tree position. Otherwise, the invitee designates the inviting node as a backup parent. A node assuming a new tree position invites all adjacent nodes except its parent to become its child. The invitations propagate throughout the network until a spanning hierarchical protection tree is formed. Upon a subsequent failure of a straddling link, the tree may be used to re-route data. Further, given a tree link failure, protection switching is quickly achieved at a disconnected node through use of a backup parent as the new primary parent.

Abstract: A star network (1) having a star coupler (14) and at least a first and second network branch (112, 122), wherein each network branch (112, 122) comprises at least one network node (110, 120) and a bus driver (11, 12) connected between the star coupler (14) and the at least one network node (110, 120). A first bus driver (11) of the first network branch (112) is adapted to detect a symbol pattern comprising at least two predetermined equal control symbols, which are transmitted from the first network node (110) to the second network node (120), wherein a control symbol is part of a control pattern. The control pattern is used to force at least the second network node (120) to switch into a predetermined state. The first bus driver (11) is disabling the first network branch (112) from the star network (1) based on whether the transmitted symbol pattern is detected. By using the protection mechanism a reliable detection of babbling idiots is provided.

Abstract: Generating a digital signature of an entire embedded code project is provided while maintaining certain exclusion areas so that a productivity application can incorporate application-specific information into the embedded code project without hampering the digital signature. A tree structure of data may be serialized into a data stream. The tree structure may include multiple branches and one or more elements identified as an exclusion area. A digital signature of the data stream may be created and included in a document associated with the tree structure.

Abstract: Methods and apparatus for addressing and/or routing packets in a network are described. A networked device in the network can include a hierarchical structure of nodes and a processor. The hierarchical structure of nodes includes n layers including n?1 layers of switch nodes and 1 layer of computational nodes. L represents a layer in the hierarchical structure and is an integer with L=0 representing a lowest layer and L=n?1 representing a highest layer. The networked device can include a processor configured for processing n groups of bits received in a packet, where each computational node is fully addressed by the n groups of bits and each switch node of a layer L is fully addressed by n?L groups of most significant bits.

Abstract: A method of generating spanning trees in a network in which a plurality of network nodes are interconnected by links. The spanning trees are utilized for handling link and node failures. For link failures, each link has at least one tree that does not include that link. For node failures, each node has at least one tree to which the node is connected by a single link. A first spanning tree connects all of the nodes, and from each node one link is left unconnected. A second spanning tree includes all of the nodes and all of the unconnected links. Thus, none of the links is included in both trees. If a node failure prevents other nodes from communicating, a third spanning tree is needed. The method minimizes the number of required trees in large networks of any topology and can be implemented off-line.

Abstract: A Join message load control system in a network using a Protocol Independent Multicast-Source Specific Multicast (PIM-SSM) protocol, the Join message load control system including a source access hop router creating a SOURCE-INACTIVE message and transmitting the created SOURCE-INACTIVE message when a multicast traffic is not received from the multicast server during a set Keep-Alive Time (KAT), and the SOURCE-INACTIVE message including information on an update set time of a PIM Join message; and a group access hop router resetting the update set time of the PIM Join message when the SOURCE-INACTIVE message is received from the source access hop router. Accordingly, it is possible to reduce the periodical transmission of unnecessary Join(S, G) messages and thereby reduce traffic load in a network by adjusting a JP timer of the Join(S, G) message depending on whether a channel(S, G) is active or inactive.

Abstract: A system comprising a first node and a second node located in a single multiprocessor system, the first node including a first router and a first replicated service executing on a first operating system, the second node including a second router and a second replicated service executing on a second operating system, and a mesh interconnect connecting the first node to the second node using the first router and the second router.

Abstract: A wireless infrastructure network system (100) includes one or more first wireless transceivers (107) for communicating data through the wireless mesh network such that the first wireless transceivers (107) act solely to transmit and receive data. The mesh network system further includes one or more second wireless transceivers (101) for communicating data through the wireless mesh network system (100) such that the second wireless transceivers (101) are each integrated with an overhead task light (201). Operation of the overhead task light (201) is controlled through the wireless network enabling each individual overhead task light (201) to be easily controlled without the need for extensive control wiring to be added to each of the second wireless transceivers.

Abstract: A connection state control device is provided that can promptly perform a reproducing operation by smoothly switching channels at each node and reduce the load on the management server in a case where operations are performed so as to sequentially change distribution information to be received at each node in a network system having a tree structure. A node receives content being distributed over a network, and performs an operation of reproducing the received content. When the node is to receive different content being distributed over a different network, the node is connected to the different network. After that, the node receives and reproduces the different content. The node continues to receive the original content but suspends the operation of reproducing the original content after connected to the different network.

Abstract: In a hierarchical tree-based protection scheme, a node in a mesh network is designated as a root node of a spanning hierarchical protection tree and subsequently invites each adjacent node to become its child within the tree. If the inviting node provides a more capacious protection path to the root node than is currently enjoyed by the invitee, the invitee designates the inviting node as its primary parent and assumes a new tree position. Otherwise, the invitee designates the inviting node as a backup parent. A node assuming a new tree position invites all adjacent nodes except its parent to become its child. The invitations propagate throughout the network until a spanning hierarchical protection tree is formed. Upon a subsequent failure of a straddling link, the tree may be used to re-route data. Further, given a tree link failure, protection switching is quickly achieved at a disconnected node through use of a backup parent as the new primary parent.

Abstract: An apparatus includes unit storing scripts includes a tree structure, each of the scripts including at least one of a child-node script and content, unit selecting a parent-node script among the scripts, and acquiring to analyze a child-node script if the child-node script exists, unit acquiring content if content exists, unit retrieving substitute content if the acquisition of content ends in failure and if content indicates retrieving substitute content, unit retrieving the substitute-child-node script if the acquisition of the child-node script ends in failure and if the child-node script indicates retrieving the substitute-child-node script, and retrieving the substitute-parent-node script when content or the child-node script shows retrieving the substitute-parent-node script, unit acquiring substitute content when substitute content is found, unit acquiring the substitute-child-node script and the substitute-parent-node script when the substitute-child-node script and the substitute-parent-node script a

Abstract: An abnormal point locating method includes: a link primary abnormality judgment procedure that acquires information on a link through which each observation flow passes, locates a link through which an observation flow observed at an observation point passes, and judges whether the located link is normal or abnormal based on the observation flow; and a link secondary abnormality judgment procedure that judges a probability of abnormality based on a result of the link primary abnormality judgment procedure by using a connection relationship as seen from each observation point.

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 method is provided of planning routes and allocating route identifiers in a managed frame-forwarding network. The network comprises a plurality of nodes interconnected by links, with each node being arranged to forward data frames according to a combination of an identifier and a network address carried by a received data frame and forwarding instructions stored at the node. A first step of the method identifies a sub-set of nodes which are core nodes of the network. The remaining nodes are termed outlying nodes. A spanning tree is then built off each of the identified core nodes, with the spanning tree stopping one link short of any other core node. Each spanning tree defines a loop-free path between a core node at the root of the spanning tree and a set of outlying nodes. Connections are planned between roots of the spanning trees and a different identifier is allocated to each planned connection between a pair of spanning trees.

Abstract: A Kitchen Display System (KDS) includes a workstation including a first Personal Computer (PC)-based controller, a fanless workstation including a second PC-based or a non-PC-based controller having lower cost than the first PC-based controller, an Ethernet network operatively coupling the controllers, and a bump bar, keyboard, touchscreen, or keypad operatively coupled to at least one of the controllers. The first PC-based controller stores and displays video, multimedia, or a build card. At least one of the controllers includes a rugged computing module, which includes a circuit board including an outer perimeter and traces associated therewith, an integrated circuit mounted on the circuit board, at least one interface connector, a housing, and a thermal transfer device.

Abstract: Locating hardware faults in a parallel computer, including defining within a tree network of the parallel computer two or more sets of non-overlapping test levels of compute nodes of the network that together include all the data communications links of the network, each non-overlapping test level comprising two or more adjacent tiers of the tree; defining test cells within each non-overlapping test level, each test cell comprising a subtree of the tree including a subtree root compute node and all descendant compute nodes of the subtree root compute node within a non-overlapping test level; performing, separately on each set of non-overlapping test levels, an uplink test on all test cells in a set of non-overlapping test levels; and performing, separately from the uplink tests and separately on each set of non-overlapping test levels, a downlink test on all test cells in a set of non-overlapping test levels.

Abstract: The present invention provides a storage network structure based on Peterson graph and data read-write method thereof. The aforementioned storage network structure consists of a primary distributed storage network and a Peterson graph-based storage network, in which, the said primary distributed storage network may be a cluster distributed storage network or a DHT (Distributed Hash Table)-based peer-to-peer storage network consisting of n nodes; and the said Peterson graph-based storage network is formed by 10 nodes selected from aforementioned storage node set according to policies.

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: Methods, apparatus, and products are disclosed for configuring compute nodes of a parallel computer in an operational group into a plurality of independent non-overlapping collective networks, the compute nodes in the operational group connected together for data communications through a global combining network, that include: partitioning the compute nodes in the operational group into a plurality of non-overlapping subgroups; designating one compute node from each of the non-overlapping subgroups as a master node; and assigning, to the compute nodes in each of the non-overlapping subgroups, class routing instructions that organize the compute nodes in that non-overlapping subgroup as a collective network such that the master node is a physical root.

Abstract: Provided herein are systems and methods for influencing shoppers while they consider making purchasing decisions in a retail establishment. Specifically, the systems and methods for influencing shoppers may occur during a shopper's first moment of truth by transmitting through a communication network one or more influential messages to a handheld wireless electronic device or intelligent shopping cart, which is in close proximity to a shopper during a shopping trip throughout a retail establishment.

Abstract: Provided is a method of constructing a tree network which includes a first node. The method comprises: (a) the first node determining if there is a child node and requesting a parent node to allow the first node to withdraw if there is no child node; (b) selecting a child node based on remaining amounts of battery power of each child node if there are child nodes of the first node; (c) exchanging positions of the first node and the selected child node and requesting the parent node to allow the first node to withdraw if the child node selected in operation (b) does not have child nodes; and (d) repeating operations (b) and (c) if the child node selected in operation (b) has child nodes. The present invention solves the problem of decreasing lifetimes of upper nodes that consume more battery power than lower nodes due to more traffic transmission and reception.

Abstract: A system and method for enabling high-speed, low-latency global tree network communications among processing nodes interconnected according to a tree network structure. The global tree network enables collective reduction operations to be performed during parallel algorithm operations executing in a computer structure having a plurality of the interconnected processing nodes. Router devices are included that interconnect the nodes of the tree via links to facilitate performance of low-latency global processing operations at nodes of the virtual tree and sub-tree structures. The global operations performed include one or more of: broadcast operations downstream from a root node to leaf nodes of a virtual tree, reduction operations upstream from leaf nodes to the root node in the virtual tree, and point-to-point message passing from any node to the root node.