Abstract: Motivation.- Network survivability models using node types.- Survivable network design under connectivity constraints - a survey.- Decomposition.- Basic inequalities.- Lifting theorems.- Partition inequalities.- Node partition inequalities.- Lifted r-cover inequalities.- Comb inequalities.- How to find valid inequalities.- Implementation of the cutting plane algorithm.- Computational results.

Abstract: A distributed computer system has a number of computers coupled thereto at distinct nodes. The computer at each node of the distributed system has a trusted computing base that includes an authentication agent for authenticating requests received from principals at other nodes in the system. Requests are transmitted to servers as messages that include a first identifier provided by the requester and a second identifier provided by the authentication agent of the requester node. Each server process is provided with a local cache of authentication data that identifies requesters whose previous request messages have been authenticated. When a request is received, the server checks the request's first and second identifiers against the entries in its local cache. If there is a match, then the request is known to be authentic. Otherwise, the server node's authentication agent is called to obtain authentication credentials from the requester's node to authenticate the request message. The principal identifier of the requester and the received credentials are stored in a local cache by the server node's authentication agent. The server process also stores a record in its local cache indicating that request messages from the specified requester are known to be authentic, thereby expediting the process of authenticating received requests.

Abstract: Properties of neural network performance are investigated by studying the modelling of non-linear dynamical systems Network complexity, node selection, prediction and the effects of noise are studied and some new metrics of performance are introduced The results are illustrated with both simulated and industrial examples

Abstract: A communication system includes a remote mobile node which acquires time-correlated data of its actual position from a global positional system (GPS), and securely transmits the information as a position history to a central facility. The mobile node includes encrypted programming material such as copyrighted video. At the central facility, a comparison is made between the received position history and predetermined signature data representing acceptable time-position histories. If a positive match is detected, a decryption key associated with the matched history is forwarded to the mobile node for decoding of the encrypted programming material.

Abstract: A computer cluster architecture including a plurality of CPUs at each of a plurality of nodes. Each CPU has the property of coherency and includes a primary cache. A local bus at each node couples: all the local caches, a local main memory having physical space assignable as-shared space and non-shared space and a local external coherency unit (ECU). An inter-node communication bus couples all the ECUs. Each ECU includes a monitoring section for monitoring the local and inter-node busses and a coherency section for a) responding to a non-shared cache-line request appearing on the local bus by directing the request to the non-shared space of the local memory and b) responding to a shared cache-line request appearing on the local bus by examining its coherence state to further determine if inter-node action is required to service the request and, if such action is required, transmitting a unique identifier and a coherency command to all the other ECUs. Each unit of information present in the shared space of the local memory is assigned, by the local ECU, a coherency state which may be: exclusive (the local copy of the requested information is unique in the cluster); 2) modified (the local copy has been updated by a CPU in the same node); 3) invalid (a local copy either does not exist or is known to be out-of-date); or 4) shared (the local copy is one of a plurality of current copies present in a plurality of nodes).

Abstract: We present a technique for computing upper bounds on the distribution of individual per-session performance measures such as delay and buffer occupancy for networks in which sessions may be routed over several “hops.” Our approach is based on first stochastically bounding the distribution of the number of packets (or cells) which can be generated by each traffic source over various lengths of time and then “pushing” these bounds (which are then shown to hold over new time interval lengths at various network queues) through the network on a per-session basis. Session performance bounds can then be computed once the stochastic bounds on the arrival process have been characterized for each session at all network nodes. A numerical example is presented and the resulting distributional bounds compared with simulation as well as with a point-valued worst-case performance bound.

Abstract: A network or telemetry system which allows virtual services at the application or presentation layer to communicate with other virtual services without regard to the physical interconnections. Each message, called a parcel, includes the information to be transmitted along with a virtual address header. The parcel is provided to a gateway, which inserts the parcel without modification into a packet with address information for the physical through session layers in the packet header. The packet is then transmitted to another network node, which receives and delivers the unmodified parcel to the addressed destination virtual service. A number of parcels from the same or different virtual services can be packed into a single packet for transmission from the gateway in cases where these parcels are all directed to virtual services at the same destination node. Once a session is established, such as between a gateway and a workstation, virtual services at the gateway node and the workstation can communicate with each other without requiring a lot of header overhead for each transmission. Instead, the session need simply be identified. Each gateway typically has one session at a time, but a workstation can support up to 64 sessions simultaneously.

Abstract: A system for making procedure calls can be used with a network of computers. An application program on a local node calls a desired library procedure. The library procedure can be available on the local node or a remote node, and the location need not be known by the application. If the library procedure is available on a remote node, a remote router procedure communicates a procedure identifier to the remote node. The procedure is executed, and any results are returned to the locol node, to be returned to the application program.

Abstract: In a packet (ATM) network, a source node at the entry of the network is responsive to a connection request from a user terminal for invoking a CAC (call admission control) algorithm to accept or reject the request depending on the amount of resource requested by the user, and allocates a portion of a free bandwidth resource exclusively to the user for the duration of the call according to established contract parameter values. Each node of the network responds to a reallocation request from the user for transmitting a copy of the request to a downstream node to elicit an acceptance message therefrom, and reserving a portion of a pool bandwidth resource and invoking the CAC algorithm to additionally reserve a portion of the free bandwidth resource. The node proceeds to allocate the reserved pool bandwidth to the user in response to the acceptance message indicating that the same amount of the reserved pool bandwidth is available in the downstream node. The allocation of the pool bandwidth is temporary. When a portion of the free bandwidth resource is reserved using the CAC algorithm, this portion is exclusively allocated to the user until the end of the call and the temporarily allocated pool bandwidth is returned to the pool resource for other users.

Abstract: An improved system for administration of license terms for a software product on the network, having an arrangement, for tracking software product usage, with one of the computers acting as a license server. This arrangement permits the license server (i) to identify the current set of nodes that are using the software product, (ii) to handle license data concerning conditions under which usage of the software product is permitted at any given node, and (iii) to determine whether at any given time the conditions would be satisfied if a given node is added to this set of nodes. The software product may thus include instructions to interface with the license server to cause enforcement of the license terms. The improvement, in one embodiment, to the system includes a policy server database maintained on each node, containing data specifying conditions under which usage of the software product is permitted on the corresponding node. Each node also has a policy server "daemon" in association with the corresponding policy server database, for (i) communicating with the license server, (ii) interfacing with both the software product and the corresponding policy server database, (iii) enforcing the license terms applicable to the software product at a given local node on the basis of both license policy maintained at the local node as well as applicable data from the license server.

Abstract: A digital switching network for providing simultaneous connections among user processors of processor-based communications equipment. The user processors act as originators and destinations of data communications packets. Packet data connections are through node controllers, which communicate with gateways, which are connected to a switching network via packet links. Control messages are communicated between node controllers, gateways, and an interchange control subsystem via various control message links. All control message processing and packet data transmissions are synchronized with a packet frame synchronization signal, and processing tasks performed by each of the network subsystems are pipelined so that they occur simultaneously. Service requests are queued in a central queue in the interchange control system. The synchronization and queueing simplify the control messages that are required to set up and release the connections.

Abstract: A partial taxonomy for cognitive maps is provided. The notions of NPN (negative-positive-neural) logic, NPN relations, coupled-type neurons, and coupled-type neural networks are introduced and used as a framework for cognitive map modeling. D-POOL a cognitive-map-based architecture for the coordination of distributed cooperative agents, is presented. D-POOL consists of a collection of distributed nodes. Each node is a cognitive-map-based metalevel system coupled with a local expert/database system (or agent). To solve a problem, a local node first pools cognitive maps from relevant agents in an NPN relation that retains both negative and positive assertions. New cognitive maps are then derived and focuses of attentions are generated. With the focuses, a solution is proposed by the local node and passed to the remote systems. The remote systems respond to the proposal, and D-POOL strives for a cooperative or compromised solution through coherent communication and perspective sharing. The utility of D-POOL is demonstrated using two examples in distributed group decision support. >

Abstract: A method and apparatus for use in a cell network node (10) for initial routing and rerouting of connections to a slave node based on ordering candidate slave connections (502), searching existing connections (507) for candidate best routes and validating the selected best route comparing actual route delays and available bandwith with maximum allowable delays and loading of the candidate slave connection. The invention also provides for programmable queuing delays by controlling queue lengths (212, 214, 216, 218) and for improved justification of reassembled information using actual maximum node delays.

Abstract: A method for restoring a communication failure in a network comprises the steps of providing a predetermined order for selection to the links included in the network, searching an alternative communication path upon a failure in communication occurring in the network, producing, at a first arbitrary node included in the alternative communication path, a first path setup message containing at least information about the number of the paths to be selected in a first link extending from the first node for setting up the alternative communication path, and an identification code of the path that is to be selected at first in the first link for establishing the alternative communication path, transmitting the first path setup message to a second node also included in the alternative communication path and connected to the first node by a second link, for effecting a setup of connection therein. The second node selects paths with the number specified by the path setup message according to said predetermined order of selection and effects a connection to the paths in the first link according to the order of selection.

Abstract: A multihop packet radio network is considered with a single traffic class and given end-to-end transmission requirements. A transmission schedule specifies at each time instant the set of links which are allowed to transmit. The purpose of a schedule is to prevent interference among transmissions from neighboring links. Given amounts of information are residing initially at a subset of the network nodes and must be delivered to a prespecified set of destination nodes. The transmission schedule that evacuates the network in minimum time is specified. The decomposition of the problem into a pure routing and a pure scheduling problem is crucial for the characterization of the optimal transmission schedule. >

Abstract: A self healing, passive protected ring network and a method of operation are disclosed. The ring network includes a plurality of active nodes interconnected by optical fibers into a working ring. The ring network also includes a passive optical fiber protection ring which interfaces with the working ring through a small number of optical switches located at each node. In the event of a failure in the optical fibers or nodes of the working ring, the optical switches are set to interconnect the protection ring to the nodes on either side of the failure and to bypass all other nodes. It is a particular advantage of the present invention that the protection ring may be installed without modifying or adding to the active electronic equipment at each node and remains unchanged when the working ring's equipment is upgraded to a higher line transmission rate.

Abstract: The practical application and implementation of online distributed system-level diagnosis theory is documented. Proven distributed diagnosis algorithms are shown to be impractical in real systems due to high resource requirements. A distributed system-level diagnosis algorithm called Adaptive DSD is shown to minimize network resources and has resulted in a practical implementation. Adaptive DSD assumes a distributed network, in which network nodes can test other nodes and determine them to be faulty or fault-free. Tests are issued from each node adaptively and depend on the fault situation of the network. Test result reports are generated from test results and forwarded between nodes in the network. Adaptive DSD is proven correct in that each fault-free node reaches an accurate independent diagnosis of the fault conditions of the remaining nodes. No restriction is placed on the number of faulty nodes; any fault situation with any number of faulty nodes is diagnosed correctly. An implementation of the Adaptive DSD algorithm is described. >

Abstract: A Transport Layer Protocol Boundary (TLPB) architecture is described which will permit an application program to run over a non-native transport protocol without first generating a protocol compensation package tailored to the transport protocols assumed by the program's application programming interface and by the available transport provider. All transport functions required by the program are converted to standardized or TLPB representations. When a connection between the first application program and a second remote application is requested, the individual required TLPB transport functions are compared to corresponding functions supported by the transport provider. Compensations are invoked only where there is a mismatch. The node on which the remote application program runs is informed of the compensations so that necessary de-compensation operations can be performed before the data is delivered to the remote application program.

Abstract: In this paper we describe briefly a dynamic multi-path routing scheme that has been considered for connection oriented homogeneous high speed networks. The fundamental objective of the scheme is to bridge the gap between routing and congestion control as the network becomes congested. Because propagation delay far out shadows queueing and transmission delay in high speed networks, the proposed routing scheme works as a shortest path (minimum hop) first algorithm under light traffic conditions. However as the shortest path becomes congested, the source node uses multiple paths when and if available in order to distribute the load and reduce packet loss. The scheme is a cross between Alternate Path routing and Trunk Reservation.We compare the performance of the proposed scheme with the Shortest Path Only algorithm, the Alternate Path routing algorithm, the Random Routing algorithm, and the Trunk Reservation scheme. The throughput and packet loss performance are compared via simulations. These have been carried out concentrating on a 5 node network with varying traffic patterns, the intention being to gain insight into the strengths and weaknesses of the various schemes.

Abstract: A bus arbitration scheme is implemented in a system where an arbitrary assembly of nodes on a system bus have been resolved into an acyclic directed graph The hierarchical arrangement of nodes has one node designated a root while all other nodes have established parent/child relationships with the nodes to which they are linked Each node may have a plurality of connected child ports with a predetermined acknowledgment priority scheme established Fair bus access arbitration provides for bus granting in a sequence corresponding to the predetermined port priorities allowing all nodes a turn on the bus The root node may always assert its priority access status to gain bus access which is useful for accommodating a root node which requires isochronous data transfer Alternatively, a token passing arbitration scheme may be implemented where the token for bus access is passed around the nodes according to the above-described predetermined port priority scheme Preemptive bus initialization may be triggered by any node upon detection of a necessitating error or addition or removal of a connection to an existing node

Abstract: A multistage interconnect network (MIN) capable of supporting massive parallel processing, including point-to-point and multicast communications between processor modules (PMs) which are connected to the input and output ports of the network. The network is built using interconnected switch nodes arranged in 2 logb N stages, wherein b is the number of switch node input/output ports, N is the number of network input/output ports and logb N indicates a ceiling function providing the smallest integer not less than logb N. The additional stages provide additional paths between network input ports and network output ports, thereby enhancing fault tolerance and lessening contention.

Abstract: A distributed network monitor system distributes the network monitoring function among each of the nodes of a multiple network system, such that monitor software resident in each node is responsible for providing status information about that node and its communications links. At predetermined monitoring intervals, a circulating status table (CST) (FIGURE 4) is circulated to all of the on-line nodes, with each node updating the CST with its link and status information. The monitor software includes a servicer task (22), a node monitor task (24), and a packet manager task (26), with intertask data transfers being implemented through a monitor region (28) in memory. In addition to link and node status information, the CST includes information about links that are in an intermittent condition (i.e., links with significantly degraded statistical performance). Intermittent link conditions are determined by a voting procedure in which each node votes on the condition of its links with other nodes, with the results of the votes being distributed in the CST and used by each node to determine those nodes with links in an intermittent condition. For those links without sufficient message traffic to make a clear determination of condition, volunteer nodes send additional link test messages until sufficient statistical information is available.

Abstract: A method and apparatus for enhanced electronic mail distribution which permits expeditious distribution of electronic mail documents to multiple host systems and/or external networks via a single host agent. A distributed data processing system having multiple host systems and/or external networks permits electronic mail documents to be distributed from local networks to nodes within various host systems or networks by establishing a host agent reference table at selected host agents within the local network. Each host agent reference table includes an identification of selected destination nodes associated with an identified host agent for those nodes. A referral to the host agent reference table may then be utilized to determine the appropriate host agent for an electronic document destined for a selected node. Thereafter, existing interprocess local network communications may be utilized to communicate the electronic document to the appropriate host agent. In the depicted embodiment of the present invention a default host agent may also be listed within the host agent reference table for all destination nodes not explicitly listed within the table. Additionally, an ordered list of likely host agents may be established and utilized sequentially in a selected order for unlisted destination nodes, in response to a failure of a document to achieve a confirmation of delivery within a specified period of time.

Abstract: A plurality of nodes in a network (100) are connected in a tree arrangement. A master control node (110) is at the root of the tree. Each connection (121-128) between a parent node and a child node (114-118) in the tree consists of only one pair of wires, over which data bits are transmitted serially. To avoid contention, communications are always initiated by the parent node. The child node detects the end of message when no more data is received for a set timeout period. The child node then has a specified time interval for response to the communication, during which interval it has control of the line. After the end of the interval, control reverts to the parent. If a message has been received from the child within that time, normal status is resumed; otherwise, the parent retries the message or takes other error recovery actions. In the preferred embodiment, the network is used to monitor power conditions at a plurality of nodes in a computer system.

Abstract: An adaptive distributed system and method for fault tolerance is provided in a network (12) of at least three nodes (14) that are able to execute a distributed diagnosis algorithm in communication with the network. Each node has a device (16, 18) for testing whether a node is in a desired state or an undesired state and for determining which node to test. The method begins by using a tester node to test another node, determining whether the other node is in a desired or undesired state. If the tested node is in an undesired state, the testing is repeated on at least one further node until a node in a desired state is tested. If the tested node is in a desired state, an activation signal is provided to the tested node, causing it to become the tester node.