Showing papers in "IEEE Communications Magazine in 1984"

Automatic-repeat-request error-control schemes

Abstract: ERROR DETECTION incorporated with automatic-repeatrequest (ARQ) is widely used for error control in data communications systems. This method of error control is simple and provides high system reliability. If a properly chosen code is used for error detection, virtually error-free data transmission can be attained. This paper surveys various types of ARQ and hybrid ARQ schemes, and error detection using linear block codes.

Spread-spectrum signal acquisition: Methods and technology

Abstract: A system and method of driect-sequence spread-spec trum TDMA (or TDD) digital communication, wherein acquisition and tracking occurs for a plurality of frames having a predetermined number of time slots. One of the predetermined number of time slots in each frame is assigned for acquisition purposes and carries acquisition and sync digital information. The remaining time slots in each frame have assigned header bytes for tracking purposes. The assigned acquisition time slot is spread with an acquisition direct-sequence spreading code. The remaining time slots are spread with a con munication direct-sequence spreading code. During acquisition, the signal strength in each successive time slot for each spread frame is measured for a given num ber of time frames. The peak is located within the given number of time frames through a major acquistion sweep and a refinement sweep. The major acquisition sweep and refinement acquisition sweep locate the frame and time slot boundaries of the transmitted signal. Once acquired, tracking occurs during the header bytes of each successive remaining time slot. Acquisition, tracking and demodulation of the digital data is per formed with the same circuitry.

Local network performance

Abstract: Because the number of truly different local network configurations is manageably small, a comparative analysis of local network performance is possible. The question of performance is of concern in the design. or selection of a local network for a specific application. Given a certain c,ollection of devices, with certain traffic characteristics, a fundamental requirement is that the local network has adequate capacity for the expected load. Table I, based on studies by the IEEE 802 Local Network Standards Committee, indicates the type of load that may be offered to a local network by various devices. We would like the local network to be able to sustain a throughput that keeps up with the load, and does so without undue delays. This paper aims to show which factors are significant in determining local network. performance' and to summarize recent comparative studies. The first section below shows that two basic characteristics of a local network, propagation delay and data rate, set an upper bound on performance independent of the mediumaccess control protocol. Next, some simple models are developed for comparing three protocols: CSMA/CD, token bus, and token ring. These are protocols for which standards have been developed [2], and it is likely that most local network products will use a variant of one of them. Finally, some comparative studies are summarized. The results cover CSMA/CD, token bus, and token ring, as weil as two other ring protocols slotted ring and register insertion.

Optical fibers in local area networks

Abstract: < , . I 0 I63-6804/84/0800-0022 E io I .OO 8 1984 IEEE I N THE following paragraphs, we wil l examine the application of optical transmission technology to the local interconnection structures that we call local area networks (LAN’s). These networks are a singular technological development with an application scope which may eventually overlap that of older conventional network structures, such as the local subscriber loop in telephone .systems. By.the term LAN we understand a data communications system that allows a number of independent, nonhomogeneous devices to communicate with each other [l]. LAN’s are usually distinguished from other types of data networks in that communication is confined to a modestly sized geographic area such as a single office building or a complex of buildings and laboratories such as a university campus. LAN application environments include the commercial, the industrial, and the institutional. Perhaps the major thrust of LAN’s will be in office applications. LAN’s must, therefore, support services such as file transfer, graphics applications, word processing, electronic mail, distributed data bases, interconnection to other LAN’s, digital telephony, and, eventually, some kind of video service. Moreover, LAN’s .must support a wide variety of data devices: computers of all vintages (micro, mini, and maxi), video terminals, mass torage devices, printers, plotters, facsimile printers, and gateways to other networks.

A local area network architecture overview

Abstract: HIS PAPER is a brief overview of existing Local Area Network (LAN) architectures and is intended as a guide for this special issue. It begins with an attempt to define an LAN, followed by some background information on the LAN environment and user needs. To provide some technical perspective, we next compare LAN's and conventional PBX's, and then we focus on current LAN architectures and the associated protocols and technology. i An LAN is a resource-sharing data-communications network , which is limited in geographic scope to the range of 0.1-10 km, provides high bandwidth communication (above 1 Mb/s) over inexpensive transmission media, and is usually privately owned. LAN's differ from conventional long-haul networks in transmission bandwidth, network protocols, and topologies. Usually, in long-haul networks, the transmission bandwidth is expensive and network design is focused on the link utilization efficiency. Designers of LAN's, unlike designers of long-haul networks, do not consider that optimizing bandwidth use is a critical issue, since bandwidth in an LAN environment is not a precious resource [1,2]. The economic and technical characteristics of LAN's offer simple solutions for network problems. Complicated routing and control algorithms for long-haul networks are not needed for LAN's with their wide bandwidth and short-delay attributes. With their small geographic scope, any one of several different LAN topologies and resource sharing mechanisms can be selected to provide economic communications for their users. The typical environments where LAN's are applicable today include [3]: Offices where integrated voice and data communications support office automation functions such as electronic mail, text processing, document distribution, and voice storage; Universities that need to provide a communications network for accessing central or distributed processing facilities and special-purpose application software such as word processing; Factories that are increasingly moving toward automation , including automated manufacturing techniques such as CAD/CAM, robotics, and numerically controlled manufacturing processes; Laboratories that need networks to support distributed computing and distributed process control for experiments; Hospitals where communications are crucial for patient file retrieval, status monitoring, and diagnostics; Distribution and sales where communications links are needed for order entry and inventory control systems. The ,LAN's primary purpose is to provide connectivity and switching functions for premises equipment [4] (such as terminals, workstations, personal computers, and shared resources like disk storage and high-speed printers), thus allowing users to efficiently accomplish required tasks. Such tasks include accessing databases or application software residing in a mainframe, distributing files, mailing …

A new local area network architecture using a centralized bus

Abstract: LOCAL AREA NETWORKS are currently enjoying tremendous popularity as a means for providing wideband interconnection and communications among data terminals, host computers and other types of digital equipment located throughout a single building or a campus of buildings. Such networks are typically based on bus, ring, or star architectures, each of which manifests its own set of advantages and disadvantages. In this paper, an architectural approach is described that draws upon and integrates the advantages found separately in these three different architectures, while avoiding the major disadvantages found in any one. This new architecture employs a centrally located short bus that provides an extremely efficient packet-switching service to the devices attached to the network. Bandwidth on the short bus is dynamically allocated in response to instantaneous demands by means of a highly efficient but flexible prioritybased bus contention scheme. The approach permits multiple priority classes with fair allocation of bandwidth within each class, along with a capability for integrated circuit and packet switching. The architecture can also make use of existing twisted-pair building wiring, and at the same time take advantage of emerging optical-fiber technology. In addition, the architecture provides a means to expand the network beyond a local area, resulting in a wide-area network capability.

Information theory: The copernican system of communications

Abstract: IT IS ARGUED that Shannon's theory of information is the scientific basis of commumcataons in the same sense that Copernicus' heliocentric theory is the scientific basis of astronomy. Some parallels between Copernicus and Shannon are drawn. It is further argued that the fundamental thesis of information theory is that all communications is essentially digital and that the main theorem of information theory is the separation theorem for source and channel coding. Some observations about the source coding theorem and the channel coding theorem are offered.

Bandwidth allocation and routing in ISDN's

Abstract: The goal of communications network design is to satisfy user requirements with the minimum amount of investment. This paper presents a method for the optimal design of ISDN's. First, three multiplexing systems that allow the integration of circuit- and packetswitching traffic are described. Then, a key design problem - bandwidth allocation and routing for integrated networks - is formulated, and efficient methods for its solution are presented.

Considerations for ISDN planning and implementation

Abstract: THE INFORMATION AGE has created such a diversity between the Communications needs of the business user and the residential Community that one common-denominator public telephone network can no longer serve everyone. The time when evolutionary network enhancements can be justified and implemented will come too late to be acceptable to a large body of business users, because of this diversity of needs. The emerging proliferation of technological bypass of local telephone facilities in the United States is symptomatic of the need for one or more shared wideband/ broadband overlay networks to serve the business community. This environment requires accelerated introduction of digital technologies in the telephone network for provision of high-speed end-to-end digital services including data, facsimile, and video. This necessitates the use of Overlay facilities and networks to meet present business customer demands until the loop plant in the common-denominator network can evolve to optimally meet the varied demands of the whole community. This paper describes past progress and current plans for the introduction of digital technologies for switching, interexchange trunking, and the local loop for an end-to-end digital network. The paper also discusses the implementation of common-channel signaling, based on CCITT's SSN07 in the circuit-switched network, as well as the use of an overlay packet-switched network for provision of data commumcations and videotex services.

Emerging telecommunications needs of the card industry

Abstract: THE CARD INDUSTRY offers a wide range of payment system products and services, including credit and charge cards, automatic-teller machine networks, interchange networks, debit card systems, authorization and processing services, and point-of-sale equipment. The trend in the industry is to extend automation in the forms of both on-line and off-line transactional systems to a vastly increased number of points of sale and to personal terminals in business and home locations. To this end, the industry is increasingly interested in interchange among proprietary and industry networks, traffic concentration and other facilities sharing arrangements, conversion of voice traffic to data traffic, use of cable television and other local-loop bypass options, integration of public packet-switched networks with private networks, and enhanced transactional security and authentication.

Deep space communication - Past, present, and future

Abstract: This paper reviews the progress made in deep space communication from its beginnings until now, describes the development and applications of NASA's Deep Space Network, and indicates directions for the future. Limiting factors in deep space communication are examined using the upcoming Voyager encounter with Uranus, centered on the downlink telemetry from spacecraft to earth, as an example. A link calculation for Voyager at Uranus over Australia is exhibited. Seven basic deep space communication functions are discussed, and technical aspects of spacecraft communication equipment, ground antennas, and ground electronics and processing are considered.

Electro magnatic detection of trapped miners

Abstract: The Bureau of Mines has conducted field studies in coal mines throughout the United States to determine the effectiveness of electromagnetic techniques in locating miners trapped underground following a mine accident. Data from these tests have been used to generate models of expected signal and noise distributions as found above these mines. These distributions have aided in placing the expected performance of a through-the-earth electromagnetic communications technique into a probabilistic framework. Results show that at a 10% false-alarm rate, the expected probability of detecting a miner's signal from a depth of 1000 ft is 54%; at 500 ft it is 95%. These depths exceed the actual depths of 90 and 50%, respectively, of United States coal mines. Sensitivity studies have shown that at a depth of 1000 ft, the probability of detection will improve approximately 2% for each dB of increase in signal-to-noise (SNR) ratio.