High-speed Communication Networks Research Articles

On the capacity of multiwavelength optical-star packet switches

The optical star network and its power budget and spectral-width limits are reviewed. A third capacity limit due to the bandwidth available within the switch for communicating the switch-control functions to the crosspoint elements of the switch is investigated. A pipelined switch processor and the communications bandwidth required within the switch to support the maximum amount of information transfer within that pipelined architecture are considered. Calculation results are presented for two examples, the switching of asynchronous transfer mode (ATM) cells and of data packets of longer duration, and the differing conclusions to which these two cases lead are compared. For short packets. such as in ATM, the power budgets are not a significant factor, and the bandwidth of the control channel is dominant. Increasing that bandwidth by the use of a parallel bus is indicated. For long packets, the power budget is much more important and the control channel less so. In this case, coherent detection adds significantly to the total switch capacity. The impact that these considerations may have on high-speed communications networks is discussed.< <ETX xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">&gt;</ETX>

A Hitchhiker’s Guide to Life on the Internet

Summary This paper provides a survey-level introduction to the Internet, the high-speed communications network that provides access to remote supercomputers, libraries, and other specialized information services. In addition to describing the physical and logical topology of the network, it describes the supercomputing and network-oriented visualization capabilities offered by one node on the Internet, the Natl. Center for Supercomputing Applications (NCSA). Engineering studies are used as vehicles throughout the paper.

Performance comparison of error control schemes in high-speed computer communication networks

The authors examine the performance of two different approaches for handling the loss and/or corruption of messages as they are transmitted between two end users in a high-speed network. In the link-by-link approach, two adjacent nodes in an end-to-end path locally detect and recover from message loss or corruption along their joining link. In the end-to-end approach, recovery is done solely on the basis of a single end-to-end protocol. The authors develop analytic performance models, validated with simulation, for comparing the performance of these two approaches. The authors find that for the range of network parameters of practical interest, an end-to-end approach towards error control is superior to a link-by-link approach, even under assumptions that would overly favor the link-by-link approach, while at the same time requiring fewer network resources (e.g. buffers, computation time) than the link-by-link approach. The performance differences arise primarily from the increased buffer requirements of the link-by-link approach. >

Evaluation of the Random Token Protocol for High-Speed and Radio Networks

In high-speed communication networks, the ratio between the end-to-end propagation delay to packet transmission time is large, causing increased scheduling overhead in demand assignment protocols and increased collision probabilities in random access schemes. These lead to rapid degradation of the channel utilization in both channel access control approaches. In this paper, we present a "random token" oriented protocol where channel access is scheduled by random, implicit token passing leading to lower channel access control penalty. By optimally balancing the collision and scheduling penalties, the protocol allows the network to reach better performance than that obtained from random access schemes in networks with and without collision detection, without imposing additional system operational assumptions. Specifically, the random token protocol does not require knowledge of the number of stations, their identities, or synchronization in periods of silence. Therefore, the protocol is also suitable for high-speed networks with frequent reconfiguration and for mobile radio networks.

Neocortex: A high speed analogic conditionable complex-pattern recognizer

Neocortical intrinsic signalling is posited to be predominantly dendro-dendritic. Dendro-dendritic synaptic transmission is at electron-spin-resonance (ESR) tempo between pulsating -NH 2 entities at the tips of the dendrites' neurofibrils. Signals along a fibril consist of ESR-tempo fluxes in excitation state of its helical C-chain macromolecule. At each dendrite-soma ‘stem region’ fibril excitation-flux transduces (reversibly) into soma-membrane ionic polarization flux. Outgoings of signals from the soma-membrane into dendrites are influenced, probabilistically, by the ‘transduction strength’ at each stem region. Transduction strength depends upon the density of H-bondings, locally, within the fibril. The density of H-bondings is proportional to the number of times the fibril has been dupli-excited. Dupli-excitation occurs whenever activation fronts from two axon-input boundaries (eg. from conditioned and unconditioned percepts) meet in precise phase at synapsing dendrite tips, causing momentary duplication in -NH 2 pulsation amplitude, which causes transient recursive duplicate excitation amplitude in the fibrils, which energises sustained addition of a quantum of H-bonds within the fibril at the stem region of the two dendrites involved: ie. H-bond transducer quanta form experience-variable memory-unit weightings within a very high speed (VHS) communication network. Such organization bypasses the speed-constrictions of axonal propagation and chemo-synaptic transmission and provides the neocorticate animal with VHS analogical processing of information during learning and pursuit of prey or avoidance of predators in familiar surroundings: ie. with VHS pattern-recognition favoring survival.

Uses of APL in a manufacturing environment

This paper describes the use of APL as a node of a high speed communications network. APL is used by the engineering community for statistical data analysis and for the creation of reports. These results are used to keep management informed of manufacturing status and to improve process control.

Microprocessors: Their impact on real-time computing in psychology

This paper attempts to evaluate the impact of microprocessors and other products using large-scale integrated circuit technologies with a set of examples of past, current, and proposed computer applications in psychological research. I first discuss a number of developments in real-time computing that have taken place during the last 10 years. The major part of this paper describes future developments in the Computer Laboratory for Instruction in Psychological Research. We are developing an intelligent high-speed communication network that will enable computers and terminals to communicate with each other. In addition, we are developing a microprocessor-based real-time computer system. The network, the real-time system, and the general assumption underlying the development of the next generation of CLIPR facilities are described.

APCOL: A programming system for computer-controlled psychology laboratories

The development of on-line computer systems can be facilitated by higher-level programming languages that are more powerful and easier to learn than machine-level Assembly languages. In general, it is not possible to speak ofa programming language without also describing the operating system and functional problem domain of the small computer system for which it is intended. This is because of the great variety in the architecture and configuration of small computers. However, it is possible to state more general requirements and design criteria for on-line computer languages in psychology. APCOL is an example of an integrated programming system. One of the important distinguishing features of the system is that it is implemented both on the small-scale laboratory computer and on a large general-purpose computer linked to the smaller machine via a high-speed communications network. This arrangement permits firsthand access to interactive program construction with immediate test of program function by using the on-line implementation as well as access to the text-editing and mass storage facilities of the large-scale machine.