Showing papers on "Node (networking) published in 1987"

Coherent Cooperation Among Communicating Problem Solvers

Abstract: When two or more computing agents work on interacting tasks, their activities should be coordinated so that they cooperate coherently. Coherence is particularly problematic in domains where each agent has only a limited view of the overall task, where communication between agents is limited, and where there is no ``controller'' to coordinate the agents. Our approach to coherent cooperation in such domains is developed in the context of a distributed problem-solving network where agents cooperate to solve a single problem. The approach stresses the importance of sophisticated local control by which each problem-solving node integrates knowledge of the problem domain with (meta-level) knowledge about network coordination. This allows nodes to make rapid, intelligent local decisions based on changing problem characteristics with only a limited amount of intercommunication to coordinate these decisions. We describe three mechanisms that improve network coherence: 1) an organizational structure that provides a long-term framework for network coordination to guide each node's local control decisions; 2) a planner at each node that develops sequences of problem-solving activities based on the current situation; and 3) meta-level communication about the current state of local problem solving that enables nodes to dynamically refine the organization. We present a variety of problem-solving situations to show the benefits and limitations of these mechanisms, and we provide simulation results showing the mechanisms to be particularly cost effective in more complex problem-solving situations. We also discuss how these mechanisms might be of more general use in other distributed computing applications.

How to emulate shared memory

Abstract: We present a simple algorithm for emulating an N processor CRCW PRAM on an N node butterfly. Each step of the PRAM is emulated in time O(log N) with high probability, using FIFO queues of size O(1) at each node. The only use of randomization is in selecting a hash function to distribute the shared address space of the PRAM onto the nodes of the butterfly. The routing itself is both deterministic and oblivious, and messages are combined without the use of associative memories or explicit sorting. As a corollary we improve the result of Pippenger [8] by routing permutations with bounded queues in logarithmic time, without the possibility of deadlock. Besides being optimal, our algorithm has the advantage of extreme simplicity and is readily suited for use in practice.

Method of maintaining a topology database

Abstract: Each network node in a communications network maintains its own copy of the network topology database defining network resources. Each resource record contains a "timer" field which is initially set to a maximum value but which may be decremented on a daily basis. If the timer field is decremented to zero without being reset, the node unilaterally removes the resource record from its copy of the database. The timer field will normally reach zero only for obsolete resource records since each network node responsible for a resource broadcasts a timer-resetting message for the resource (1) each time the resource status changes, (2) when the node first joins or rejoins the network, and (3) on a periodic (weekly) basis regardless of whether conditions (1) or (2) have occurred.

Locating resources in computer networks

Abstract: A LOCATE search dynamically locates resources (e.g., logical units (LUs) and transaction program and files associated with LUs) in a computer network so that a session can be established between the origin and the destination of the search. In a network which includes end nodes associated with server nodes, a resource contained in any node is located by a requesting node. The requesting node initiates a search of the resources residing within itself. If the resource does not reside in the requesting node, the server node searches the resources known to the server node that reside anywhere in the nework. If the resource is not known by the server node, it searchers all resources that reside in its associated end nodes. If the resource does not reside in the associated end nodes, either a request is sent to a central directory if one exists or a search of all resources in the network is made.

Distributed Nodes Organization Algorithm for Channel Access in a Multihop Dynamic Radio Network

Abstract: This paper proposes a solution to providing a collision free channel allocation in a multihop mobile radio network. An efficient solution to this problem provides spatial reuse of the bandwidth whenever possible. A robust solution maintains the collision free property of the allocation under any combination of topological changes. The node organization algorithm presented in this paper provides a completely distributed, maximally localized execution of collision free channel allocation. It allows for parallel channel allocation in stationary and mobile networks with provable spatial reuse properties. A simpler version of the algorithm provides also a highly localized distributed coloring algorithm of dynamic graphs.

Network adapter for connecting local area network to backbone network

Abstract: A table memory stores an address transformation table (AT table) which contains no address pairs therein before initialization of the system. A hash function memory stores a plurality of hash functions which are selectively used in correspondence with the sizes of a backbone network (BN) and a local area network (LAN), when an address pair consisting of a node address and a LAN address, of a LAN to which a node indicated by the node address belongs, is registered in the AT table. When a control unit receives frame data from the LAN or the BN, it automatically generates the above address pair consisting of the node address and the LAN address in the AT table by selecting one of the plurality of hash functions. During registration, if the AT table reaches the limit of its available storage space, and a new address is to be registered, the control unit then generates an AT table larger in size than the current AT table, using another hash function, and registers the already registered address pairs and the new address pair in the newly enlarged AT table.

Modeling and performance analysis of multihop packet radio networks

Abstract: The design of packet radio networks involves a large number of issues which interact in a very complex fashion. Many of these pertain to the RF channel and its use, others pertain to the operational protocols. Clearly, no single model can be formulated which incorporates all the necessary parameters and leads to the optimum solution. The one essential element which complicates matters is that, contrary to point-to-point networks in which each channel is utilized by a single pair of nodes, the radio channel in packet radio networks is a multiaccess broadcast resource: i) in a given locality determined by radio connectivity, the channel is shared by many contending users, hence the need for channel access protocols; ii) radio is a broadcast medium and thus the action taken by a node has an effect on the actions taken by neighboring nodes and their outcome. Despite the complexity of the problem, there has been significant progress worth reporting on. The work accomplished so far has been either the analysis of specific examples of networks or an attempt to create models that would be useful in the design of general networks. The purpose of this paper is to survey the various modeling techniques that have been used for the performance analysis of packet radio networks, and to discuss the assumptions underlying these models, their scope of applicability, and some of the results obtained.

Hypernet: A Communication-Efficient Architecture for Constructing Massively Parallel Computers

Abstract: A new class of modular networks is proposed for hierarchically constructing massively parallel computer systems for distributed supercomputing and AI applications. These networks are called hypernets. They are constructed incrementally with identical cubelets, treelets, or buslets that are well suited for VLSI implementation. Hypernets integrate positive features of both hypercubes and tree-based topologies, and maintain a constant node degree when the network size increases. This paper presents the principles of constructing hypernets and analyzes their architectural potentials in terms of message routing complexity, cost-effective support for global as well as localized communication, I/O capabilities, and fault tolerance. Several algorithms are mapped onto hypernets to illustrate their ability to support parallel processing in a hierarchically structured or data-dependent environment. The emulation of hypercube connections using less hardware is shown. The potential of hypernets for efficient support of connectionist models of computation is also explored.

Tree-Based Broadcasting in Multihop Radio Networks

Abstract: This paper considers the issue of broadcasting protocols in multihop radio networks. The objective of a broadcasting protocol is to deliver the broadcasted message to all network nodes. To efficiently achieve this objective the broad- casting protocol in this paper utilizes two basic properties of the multihop radio network. One is the broadcast nature of the radio which allows every single trasmission to reach all nodes that are in line of sight and within range of the transmitting node. The other, spatial reuse of the radio channel, which due to the multihop nature of the network allows multiple simultaneous transmissions to be received correctly. The proposed protocol incorporates these properties to obtain a collision free forwarding of the broadcasted message on a tree. Centralized and distributed algorithms for the tree construction are presented. The obtained trees are unique in incorporating radio oriented time ordering as part of their definition. In this way multiple copies of one or more broadcasted messages can be transmitted simultaneously without collision, requiring only a small number of message transmissions. Consequently, the protocol not only guarantees that the broadcasted message reaches all network nodes in bounded time, but also ensures that the broadcasting activity will use only limited channel bandwidth and node memory. The proposed broadcast protocol thus possesses the advantages of TDM solutions while allowing the channel bandwidth to be shared, concurrently with the broadcast, with other transmission activities as dictated, for instance, by data link protocols. Some NP-completeness proofs are also given.

Input/output network for computer system

Abstract: An I/O network channel (140) achieves relatively high rates of data communication between a plurality of widely physically dispersed devices interconnected by a LAN-type media (170). A processor (172) connected to each node (174) of the media controls the data transmitting and receiving functions at each node to establish network level, transport level and session level data communication and control functions.

Virtual Time CSMA Protocols for Hard Real-Time Communication

Abstract: We study virtual time CSMA protocols for hard real time communication systems, i, e., systems where messages have explicit deadlines. In this class of CSMA protocols, each node maintains two clocks; a real time clock and a virtual time clock. Whenever a node finds the channel to be idle, it resets its virtual clock. The virtual clock then runs at a higher rate than the real clock. A node transmits a waiting message when the time on the virtual clock is equal to some parameter of the message. Using different message parameters in conjunction with the virtual clock, different transmission policies can be implemented. In particular, use of message arrival time, message length, message laxity, and message deadline implements FCFS, Minimum-Length-First, Minimum-Laxity-First, and Minimum-Deadline-First transmission policies, respectively.

Synchronizer for a fault tolerant multiple node processing system

Abstract: A synchronizer for each node in a multiple node processing system having a message interface for receiving sync and pre-sync time-dependent message, counter means for generating a local time, a time stamp memory having an entry for each node in the multiple node processing system, a time stamper responsive to receiving a time-dependent message from a node for storing the local time in the entry of said time stamp memory for that node to generate a time stamp. The synchronizer has a time stamp voter for generating a medial time stamp value from all the time stamps in the time stamp memory, a sync correction generator for generating a sync delta having a value corresponding to the difference between the voted time stamp and the node's own time stamp, means for adding said sync delta to a nominal transmission timing interval to generate an actual transmission timing interval, and a message generator for generating a pre-sync time dependent message passed to the transmitter at the end of the nominal transmission timing interval and for generating a sync time-dependent message passed to the transmitter at the end of said actual transmission timing interval.

Architecture of a message-driven processor

Abstract: We propose a machine architecture for a high-performance processing node for a message-passing, MIMD concurrent computer. The principal mechanisms for attaining this goal are the direct execution and buffering of messages and a memory-based architecture that permits very fast context switches. Our architecture also includes a novel memory organization that permits both indexed and associative accesses and that incorporates an instruction buffer and message queue. Simulation results suggest that this architecture reduces message reception overhead by more than an order of magnitude.

Method and apparatus for eliminating unsuccessful tries in a search tree

Abstract: A circuit switching system in an M-ary, n-cube connected network completes a best-first path from an originating node to a destination node by latching valid legs of the path as the path is being sought out. Each network node is provided with a routing hyperswitch sub-network, ("HSN") connected between that node and bidirectional high capacity communication channels of the n-cube network. The sub-networks are all controlled by routing algorithms which respond to message identification headings ("headers") on messages to be routed along one or more routing legs. The header includes information embedded therein which is interpreted by each sub-network to route and historically update the header. A logic circuit, available at every node, implements the algorithm and automatically forwards or back-tracks the header in the network legs of various paths until a completed path is latched.

A new distributed algorithm to find breadth first search trees

Abstract: A new distributed algorithm is presented for constructing breadth first search (BFS) trees. A BFS tree is a tree of shortest paths from a given root node to all other nodes of a network under the assumption of unit edge weights; such trees provide useful building blocks for a number of routing and control functions in communication networks. The order of communication complexity for the new algorithm is O(V^{1.6} + E) where V is the number of nodes and E the number of edges. For dense networks with E \geq V^{1.6} this order of complexity is optimum.

Digital signature system and method based on a conventional encryption function

Abstract: A method of generating digital signatures for signing an infinitely expandable series of messages M i . An infinitely expandable tree of signature nodes is used, where each node can be used to sign a message. Each node is also used to sign up to k subnodes, where k is an integer greater than one. Each signature used, both for signing messages and for signing subnodes, is a one time signature, which in the preferred embodiment is based on a one-way function F. The function F is made public. To sign a message M i the signer selects a previously unused node (i.e., node i) from the signature tree. The message signing key at this node is then used to sign this message. The sequence of nodes from the root of the tree (i.e. node l) to node i is then used to verify that the message signature is correct and has not been tampered with. Furthermore, this process proves that the message has not been tampered with. Advantages of the invention include the infinite expandability of the signature tree, dependable verification of messages based on the use of secure one time signatures (e.g., which may be based on one way functions), the small amount of computation required to set up a signature tree, the small amount of storage required to maintain a tree, and the ability to implement the invention using high speed conventional encryption equipment and methods.

Local area network priority control system

Abstract: A LAN priority control system is described which uses an interrupt priority control structure which suspends transmission of a packet from a lower priority node and allows a higher priority interrupting node, which desires to transmit a higher priority message, to obtain rapid access to the transmission medium and to transmit the message The message is interrupted using a within packet interrupt which is taken to include any command which results in the transmission of a packet onto a common-medium being halted before the complete packet has been transmitted In addition, the priority of nodes can be adjusted by a centralized network manager or by intelligent terminals distributed among the nodes, allowing priorities to be adjusted at nodes to correspond to the relative priorities of messages currently at these nodes rather than having a fixed allocation of priorities In a token ring implementation, suspension of the message causes a busy token to be released which recirculates round the ring A token is then freed to the interrupting node or nodes which claims the token and uses the token to transmit the higher priority message and, after transmission cycle of the message, the token is returned to its free state and its original node to complete transmission of the original message, in the absence of another active higher priority node The within packet interrupt can be implemented in software or hardware and is suitable for use with token passing local area networks and especially a token passing ring

Competitive Location on a Network

Abstract: In this paper we study the problem of locating facilities on a network in the presence of competition. Customers at each node in the network choose from the available facilities so as to minimize the distance traveled. The problem is to find a set of facilities that is stable in the sense that each facility is economically viable and no competitor can successfully open any facilities. We define several versions of stability and establish certain relationships between them. We then present integer programming formulations that identify stable sets, and describe an enumeration algorithm for constructing a profit-maximizing stable set.

Adaptive session-level pacing

Abstract: A technique called "Adaptive Pacing" permits a receiving node, in a computing networking, to change the size of an information window during an active session. To effect the change, the receiving node sends an "Isolated Pacing Message" (IPM) to the sending node. Thereafter, the sending node adjusts the window size to conform with a window size value in the IPM. The IPM includes a type field, a reset window indicator and a next window size field. There are three types of isolated pacing messages. Two of the messages are used by the receiving node to control the flow of data; while the third message is used, by the sending node, to confirm the beginning of a new window and the ending of an old one.

Loosely coupled distributed computer system with node synchronization for precision in real time applications

Abstract: A loosely coupled distributed computer system provided with node synchronization for precision in real time applications includes a number of loosely coupled node computers. Each computer includes a local real time clock, I/O subsystems, and a communication unit. The computers are connected to each other by a transmission medium for serial communication. Each computer contains a synchronization unit for synchronization of the local real time clocks in each of the computer nodes with the real time clocks of other nodes. Each synchronization unit is connected with a digital output line of its associated local real time clock for supply of unsynchronized time signals from the digital output line to a digital input of the synchronization unit. The latter also has a digital output supplying global synchronized time signals to digital inputs of other I/O subsystems and is operatively connected with the communication unit for obtaining therefrom information as to the point of time of sending a message. The synchronization unit generates global synchronized time signals on its digital output line as a function of the time of an information from a sender and the arrival time information of other computer systems.

Time stamping for packet-system nodes.

Abstract: Delay that a packet experiences in a network node is measured by employing a single local time signal in the node, and by separating a time stamp update function into a network node entry stamp function and a network node exit time stamp function. The node entry time stamp function is to generate a packet originate time value referenced to the local time signal value upon the packet entering the node. This packet originate time value is inserted into a packet header time stamp field to be transported through the node. The node exit time stamp function is to generate an updated time stamp value from the packet originate time value from the header time stamp field and the local time signal value upon the packet exiting the node. The undated time stamp value is inserted into the packet header time stamp field in place of the packet originate time value.

Lookahead bus arbitration system with override of conditional access grants by bus cycle extensions for multicycle data transfers

Abstract: A node for obtaining access to a bus. In this arbitration method, the node receives a conditional grant. The node determines whether access to the bus will actually transfer to it. The node contains distributed logic that examines an extend bus cycle signal to determine whether it can become a transmitter to transfer messages on the bus. When the node becomes a transmitter, it generates an extend bus cycle signal to maintain access to the bus when executing a multi-cycle transfer, even though other nodes, perhaps with higher priorities, require access to the bus.

Design of a Self-Timed VLSI Multicomputer Communication Controller,

Abstract: : We describe the design of the network design frame, (NDF), a self-timed routing chip for a message-passing concurrent computer The NDF uses a partitioned data path, low-voltage output drivers, and a distributed token-passing arbiter to provide a bandwidth of 450Mbits/sec into the network Wormhole routing and bidirectional virtual channels are used to provide low latency communications, less than micro seconds latency to deliver a 216 bit message across the diameter of a 1K node machine To support concurrent software systems, the NDF provides two logical networks, one for user messages and one for system messages, that share the same set of physical wires To facilitate the development of network nodes, the NDF is a design frame The NDF circuitry is integrated into the pad frame of a chip leaving the center of the chip uncommitted

Architectures and Performance of Multichannel Multihop Packet Radio Networks

Abstract: Packet radio networks that employ several parallel multiple-access channels are considered. An architecture for such a network dictates the selection of channels for packet transmission. We propose and analyze two multichannel architectures. In the first, each node employs a single radio and is assigned a channel on which it listens when it does not transmit. To transmit a packet, the node tunes its radio to the channel of the intended receiver for that transmission only. The second architecture requires each radio to use a single channel for both transmission and reception, but provides some of the nodes with more than one radio each, allowing them to serve as bridges between channels. Within these architectures, one can further select the amount of routing information held by each node and the channel-access protocol, both of which greatly affect the network performance. To ascertain the effects of the various parameters, we calculate the throughput in both architectures. The channel-access protocols we consider are slotted ALOHA and CSMA with and without capture. We also evaluate the effect of increasing the amount of routing information held by the nodes.

Interprocess communication queue location transparency

Abstract: A system for an efficient message handling technique implemented in AIX, an operating system derived from UNIX System V, is described for use in a distributed services network include a plurality of multi-processing, multi-tasking nodes among which interprocess communication occurs via queues, the actual node locations of which are transparent at the application interface.

Modeling of Concurrent Task Execution in a Distributed System for Real-Time Control

Abstract: In a distributed system that implements real-time control, computational tasks are distributed over different nodes for execution to improve response time and system reliability. To model system behavior, tasks in each node are first decomposed into activities. The activities and precedence constraints among them are then modeled by a generalized stochastic Petri net (GSPN). Finally, a sequence of homogeneous continuous-time Markov chains (CTMC's) is built from the GSPN to model the concurrent task execution in the system.

Computer bus apparatus with distributed arbitration

Abstract: A bus apparatus for interconnecting a plurality of nodes is disclosed. The nodes may comprise processors, input/output subsystems, or the like. Each node maintains a unique priority number; the priority numbers are determined independently by each node. Separate updating of the priority numbers occurs for acknowledgement packets as compared to data transmissions. This provides for quick, efficient acknowledgement of transmissions and does not unfairly penalize a popular receiving node. Two different interface circuits are described, one particularly suitable for use with an input/output subsystem, and the other for a processor.

Node initiating xid exchanges over an activated link including an exchange of sets of binding signals between nodes for establishing sessions

Abstract: Enhanced type 2.1 nodes for SNA networks provide network routing of information between non-adjacent network nodes and program-to-program linkage across the network. These nodes are called "advanced peer to peer networking" (abbreviated APPN) nodes. Network state information, describing the topology and resources of the network is disseminated betweeen APPN nodes by a first data link control exchange of identification (XID) information indicating node type (and thereby distinguishing APPN nodes from other nodes) and other exchanges conducted over a newly defined control point to control point (abbreviated CP--CP) session having a message architecture unique to APPN type nodes. The XID exchange is conducted whenever a link between nodes is activated, and when the exchanging nodes are both APPN, and if both are available to participate in as CP--CP session, they automatically engage in a second exchange of session binding signals which establish a pair of sessions between the nodes. By internal conventions, the node components which use the session pair for exchanging their network state information adapt the sessions to be used in simplex modes with one session flowing from one node to the other and the other flowing in reverse from the other node to the one node. Within these sessions, for the sake of efficiency, communications are not directly acknowledged. Instead, a process transparent to the sessions monitors for link failure and upon such occurrence notifies all affected components to terminate their use of the session and deactivate the session.