Network Bibliography Search Results

Query: bandwidth 


Chris Adams and Stephen Ades, "Voice experiments in the UNIVERSE project," in Conference
Record of the International Conference on Communications (ICC), (Chicago, Illinois), pp.
927-935 (29.4), IEEE, June 1985. 

 Abstract: In the UNIVERSE project, experiments were conducted on a series of
 Cambridge Ring LANs linked together by satellite bridges and local bridges. The
 Cambridge Ring is a slotted ring which has the considerable advantage for packet voice,
 when compared with for example CSMA systems, that efficient transmission of small
 packets is feasible and the ring inherently shares bandwidth amongst contenders in a
 manner favorable to voice in an ISDN environment. An adaptive packet voice protocol
 (which adjusts to the load and jitter conditions which prevail during the course of any
 particular call) has been designed, implemented and tested. It proved to be resilient and to
 give satisfactory sound quality whilst running through bridges, the bridges being subjected
 to widely varying loads. Design requirements for bridges/gateways in an ISDN, which have
 to bear both voice and data traffic, are discussed in the light of experience with this
 protocol.

 Keywords: packet voice; ring network; local area network

 Annotation: discusses silence substitution (last one second repeated, leading to repeat of
 clicks)

Stephen Ades, Roy Want, and Roger Calnan, "Protocols for Real Time Voice Communication on a
Packet Local Network," in Conference Record of the International Conference on
Communications (ICC), (Toronto, Canada), pp. 525-530 (17.1), IEEE, June 1986. 

 Abstract: There is currently much interest in alternative architectures and networks forthe
 provision of integrated services. A project at the University Computer Laboratory has been
 investigating integrated services provision in a local area context using a slotted ring; the
 scope of this work ranges widely from protocol design for simultaneous transport of voice,
 data and images to user-level integrated services facilities, e.g., multimedia editors and
 image manipulation. This paper addresses the lowest level - simultaneous transmission of
 voice and data on one network. In particular, since our network is packet based whereas
 traditional voice-carrying networks have been circuit switched, it considers the design of
 protocols for packet voice to produce delays no greater than those in current
 circuit-switched PABXs - a capability without which packet systems simply are not
 practical for general use. Our work on packet-based protocols has been fuelled by the fact
 that commonly accepted telecommunication approaches to voice/data transport are based
 on assumptions, from some time ago, that bandwidth is expensive and that its usage must
 be maximized and optimized. These assumptions are no longer true within current local
 area network technology.

 Keywords: packet voice; playout synchronization; local area networks; PLL

 Annotation: Presents detailed description of packet voice on a 100 Mbits/s slotted ring. 2
 ms packets without silence suppression and silence fill-in. Discusses circular buffering, 8
 kHz synchronization with slow adjustment (packet-PLL). Argues against silence
 suppression under normal circumstances to prevent change in background noise during
 silence periods; mark silence for storage or WAN applications. Packets carry sequence
 numbers, but no time stamps. Delays set amount j after receival of first packet. Cites study
 that bursts of missing packet occurring up to 1% are not noticeable as long as bursts are
 less than 4 ms.

L. Bhuyan and D. Agrawal, "Design and Performance of Generalized Interconnection Networks,"
IEEE Transactions on Computers, vol. C-32, no. 12, pp. 1081-1090, 1983. 

 Keywords: Bandwidth switching; cost function; interconnection

 Annotation: This paper introduces a general class of self-routing interconnection
 networks for tightly coupled multiprocessor systems. The proposed network, named a
 "generalized shuffle networl" is based on a new interconnection pattern called a
 generalized shuffle and is capable of connecting any number of processors M to any
 number of memory modules N. The technique results in a variety of interconnection
 networks depending on how M and N are factore

Radamis Botros, Onsy Abdel-Alim, and Peter Damaske, "Stereophonic speech
teleconferencing," in IEEE International Conference on Acoustics, Speech, and Signal Processing,
vol. 2, (Tokyo, Japan), pp. 1321 - 1324 (26.6), IEEE, Apr. 1986. 

 Abstract: Speech intelligibility in monophonic loudspeaking telephones has been subject to
 some limitations from both the transmission medium and room acoustics. The medium has a
 limited bandwidth in addition to inherent noise and distortions. Room acoustics usually add
 the undesired impurities of the ``rain-barrel effect'', reverberance and background noise to
 direct speech signals. Binaural listening offsets many of these effects, enhancing speech
 intelligibility. In this work, a stereophonic conference telephone has been designed and
 tested. The test results showed the feasibility of talker localization and improvements in
 talker identification. If two or more persons are talking simultaneously, the stero telephone
 makes it possible to concentrate with less strain on one of the talkers. Speech intelligibility
 was found to be higher with the stereophonic telephone.

 Keywords: speech transmission; audioconferencing; teleconferencing

M. Butto and G. Colombo, "Performance evaluation of DA/SS-TDMA satellite systems with
wide bandwidth traffics," in Tenth International Teletraffic Congress, (Montreal, Canada), pp.
1-7 (4.3a.6), 1983. 

 Keywords: Data traffic; computer network

L. Bic and R. R. Hartmann, "AGM: A Dataflow Database Machine," Technical Report
ICS-TR-85-24, University of California, Irvine, Department of Information and Computer
Science, Aug. 1985. 

 Abstract: In recent years, a number of database machines consisting of large numbers of
 parallel processing elements have been proposed. Unfortunately, one of the main limitations
 to parallelism in database processing is the I/O bandwidth of the underlying storage
 devices. One way to solve this problem is to use multiple parallel disk units. The main
 problem with this approach, however, is the lack of a computational model capable of
 utilizing the potential of any significant number of such devices. This paper presents a
 database model which is based on the principles of datadriven computation. According to
 this model, the database is represented as a network in which each node is conceptually an
 independent processing element, capable of communicating with other nodes by
 exchanging messages along the network arcs. To answer a query, one or more such
 messages, called tokens, are created and injected into the network. These then propagate
 asynchron ously through the network in the search of results satisfying the given query. To
 investigate the performance of the proposed system, we have implemented the model on a
 simulated computer architecture. The results of the simulation experiments indicate that the
 model is capable of exploiting the potential I/O bandwidth of a large number of disk units as
 well as the computational power of the associated processing elements.

Wai Yip Chan, "A speech detector for mobile radio," Master's thesis, Carleton University, Ottawa,
Ontario, Canada, Sept. 1982. 

 Abstract: Mobile radio bandwidth can be efficiently utilized by not transmitting a talker's
 voice signal during pauses, which on the average accounts for 60% of a party's time in a
 conversation. A speech detector (SD) is needed to identify the speech-plus-noise and
 noise-only intervals within the talker's audio signal. In this way, both voice and data users
 can share the radio frequencies on demand. The SD should be simple, as each voice
 terminal must possess one. Deployed in a background of highly variable acoustic noise, the
 SD should be adaptive and robust. Also, the detector should produce acceptable speech
 quality and yet realize a bandwidth-compression gain close to the reciprocal of the
 intrinsic speech activity. Lastly, the detector's talkspurt/pause statistics should be such that
 the speech transmission is tolerant to variable delay induced by a
 dynamic-resource-sharing network. This thesis describes the design and development of
 a SD to meet the above requirements. Two signal features are used to make the binary
 decision: magnitude-energy and zero crossings. The SD also uses a hangover to admit
 weak utterances as speech and to produce continuous speech flow. The novel part of the
 SD is to use a signal-variability measure to adapt the parameters of the discrimination
 devices to the background noise characteristics. The SD has been implemented and
 developed on the 2920 single-chip digital signal processor. From observations, subjective
 evaluation, and objective measurements, the SD is shown to meet the said requirements.

 Keywords: silence detection; speech transmission; packet voice; packet radio

Imrich Chlamtac, "An Ethernet compatible protocol for real-time voice/data integration," 
Computer Networks Journal, vol. 10, pp. 81-96, Sept. 1985. 

 Abstract: This paper addresses the issues involved in providing an integrated real time
 voice and data service on the Ethernet local network. The paper proposes a protocol that is
 functional, robust and compatible with current Ethernet specifications. The proposed voice
 protocol meets the real time voice performance requirements for local and potentially
 remote distribution of voice. As demonstrated by detailed simulation, the addition of voice
 traffic to an existing Ethernet via the proposed protocol improves the total channel
 utilization and leaves more data-usable bandwidth when compared to the same load of
 voice being added via CSMA-CD. Controllers implementing this protocol can co-exist on
 an Ethernet with stations using conventional controllers. Additionally, the introduction of the
 voice controllers requires no changes to the data link procedures used by the already
 existing data stations in the network. The protocol can be considered for other real time
 applications, such as process control, as well.

 Keywords: packet voice; Ethernet; integrated services; multiple access

Danny Cohen, "Issues in Transnet Packetized Voice Communications," in Proceedings of the
Fifth Data Communications Symposium, (Snowbird, Utah), pp. 6-10 - 6-13, ACM, IEEE, Sept.
1977. 

 Abstract: In recent years, very important progress was made in both real-time digital
 computer communication and real-time digital voice communication, due to the growing
 importance of geographically distributed computation, the growing need for resource
 sharing, and the need for secure voice communication. The progress in digital voice
 communication - mainly aimed at low data rate and high fidelity speech communication -
 has led to various vocoding techniques such as LPC and CVSD. The progress in real-time
 computer communication - which is mainly aimed at low delays, high bandwidth and high
 cost-effectiveness - led to technologies such as packet switching, packet radio and
 satellite communications, as well as intercomputer communication protocols. It is the
 natural result of the progress in both areas that the use of packet switching technology be
 considered for voice communication. In this paper, the issues involved in the application of
 packet-switching networks to real-time voice communication are identified and explained.

 Keywords: packet voice; playout synchronization; Internet; LPC; CVSD

 Annotation: Outlines delay components and graphs synchronization recovery, describing
 tradeoffs between loss and discontinuity.

Danny Cohen, "Flow control for real-time communication," ACM Computer Communication
Review, vol. 10, pp. 41-47, January/April 1980. 

 Abstract: Flow control is a problem of resource allocation in communication systems;
 typically the critical resources are the bandwidth of the medium, memory space and
 processing capabilities. Traditionally, protecting the resources of the communicating
 processes is called flow control, and the protecting of the communication system is called
 congestion control. Hence, the flow control is the protection of one process from the others,
 and the congestion control is protection of the communication system from the processes.
 This note argues that real-time communication requires different flow control and
 congestion control schemes than the familiar ones used for non-real-time communication.

 Keywords: flow control; congestion control; real-time traffic

 Annotation: distinguishes 'milk rule' (discard oldest) from 'wine rule' (discard youngest)

Ch. Das and L. Bhuyan, "Bandwidth Availability of Multiple-Bus Multiprocessors," IEEE
Transactions on Computers, vol. C-34, pp. 918-926, Oct. 1985. 

 Keywords: Multiprocessor system; interconnection network; bus

 Annotation: Multiprocessor systems should be designed considering both performance
 and reliability issues. They sholud support graceful degradation by isolating the failed
 components and by reconfiguring to a new state with decreased performance. We present
 in this paper the effect of failures on the performance of multiple-bus multiprocessors.
 Bandwidth expressions for this architecture are derived for uniform and nonuniform
 memory references.

R. E. Eaves, "ALOHA/TDM systems with multiple downlink capacity," IEEE Transactions on
Communications, vol. COM-27, pp. 537-541, 1979. 

 Keywords: TDM; time division multiplex; ALOHA

 Annotation: The concept of a processing satellite with multiple slotted aloha uplinks to
 efficiently fill a TDM downlink is extended to include multiple downlinks. Throughput is
 derived for a general configuration with n ALOHA uplinks and m TDM downlinks. An
 optimization procedure is described to select the n and m which maximize throughput for a
 given bandwidth constraint.

O. Enomoto and H. Miyamoto, "An analysis of mixtures of multiple bandwidth traffic on time
division switching networks," in Seventh International Teletraffic Congress, (Stockholm), p. 635,
1973. 

 Keywords: Computer network; performance evaluation; multichannel

 Annotation: Full availability; limited availability

Michael Fine and Fouad A. Tobabri, "Packet voice on a local area network with round robin
service," IEEE Transactions on Communications, vol. COM-34, pp. 906-915, Sept. 1986. 

 Abstract: In this paper, we propose a combined voice/data protocol suitable for multiple
 access broadcast networks that provide round robin service to the stations. Such networks
 are well suited to the integration of voice and data since they guarantee bounded delay and
 provide high utilization even for high bandwidth channels. Using one such network proposal
 - namely Expressnet - as a representative scheme, we examine the characteristics of the
 service that voice traffic experiences under the voice/data protocol. We show that the
 access protocol is able to utilize the channel efficiently to support a large polpulation of
 voice sources while maintaining low packet delay and guaranteeing some prespecified
 minimum bandwidth for data traffic. In addition, we show the advantages of silence
 suppression, i.e., discarding speech that constitutes silent periods, and we examine the cost
 of overloading the network in terms of the amount of speech discarded.

 Keywords: packet voice; voice-data integration; local area networks; LAN; round robin

Donald P. Gaver and John P. Lehoczky, "Channels that cooperatively service a data stream and
voice messages," IEEE Transactions on Communications, vol. COM-30, pp. 1153-1162, May
1982. 

 Abstract: This paper presents an analysis of the performance of a special type of
 voice-data queueing system. The system consists of an integrated circuit and
 packet-switched multiplexer resembling that found in the SENET network. A fluid-flow
 approximation for the data component is introduced to ease the study of the voice loss rate
 and data packet waiting times. In this paper it is shown that very long data queues are
 possible, with the mean data queue length being proportional to the ratio of the holding time
 for voice to the holding time for data. The equilibrium distribution is determined, and aspects
 of the transient behavior are characterized. The question of how much bandwidth to
 reserve for exclusive use of data and how much to share between voice and data is
 addressed.

 Keywords: integrated services; SENET; packet voice; multiplexer; performance
 evaluation; fluid flow model

Kathryn Hanson, Wushow Chou, and Arne Nilsson, "Integration of voice, data, and image traffic
on a wideband local network," in Computer Networking Symposium, (Gaithersburg, Maryland),
pp. 3-11, IEEE, Dec. 1981. 

 Abstract: A byrid access method (HAM) is proposed for data, voice, and image traffic on
 a wideband, local communication network. The mixture of bursty and stream traffic from a
 large population of users requires: uninterrupted bandwidth for stream traffic; fast
 response time for bursty traffic; access by simple interface devices; method suitability for
 wide channel bandwidth, in particular, for optical fiber; and no bandwdidth monopolization
 by either type of traffic. HAM meets these constraints by dividing the channel into time
 frames, each with a fixed number of minislots for reservations, a specified maximum
 number of slots reserved by stream traffic, and a specified minimum number of slots for
 bursty traffic. The boundary between the bursty and stream portions of the frame shifts in
 response to stream traffic demand. HAM is analyzed by modified reservation-Aloha and
 slotted-Aloha, and by diffusion approximation or analysis of models based on M/M/n (with
 varying service rate) or M/G/1 queues, depending on traffic load. Results show HAM
 appropriate for a range of mixes of traffic types and network operating characteristics.

 Keywords: integrated services; hybrid switching; movable boundary; local area networks;
 multiple access

A. Huang and Scott C. Knauer, "Wideband digital switching network," 1983. United States Patent
Nr. 4542497 Appl. 479669. 

 Keywords: ATM; digital network; switching network; self routing

 Annotation: Disclosed is a wide bandwidth self-routing switch. One-to-one,
 one-to-many, and many-to-many modes of communications are achieved with time
 multiplexed signal packets of multi-service users with a system having a bank of
 demultiplexers which demultiplex incoming signals and supply them to a broadcast
 network. The broadcast network includes a concentrator, a "sort on source"
 sortingnetwork and a copy network.

Alan Huang and Scott Knauer, "Starlite: a wideband digital switch," in Proceedings of the IEEE
Conference on Global Communications (GLOBECOM), (Atlanta, Georgia), pp. 121-125 (5.3),
IEEE, Nov. 1984. 

 Abstract: Starlite is a wideband digital switch intended for future visual and data
 communication needs. It provides arbitrary amounts of receive and transmit per-user 
 bandwidth (in fixed increments) for either bursty or continuous signals. In addition to the
 traditional one-to-one connection, the switch provides for simultaneous reception of
 information from separate sources, simultaneous transmission of different information to
 separate destinations and broadcast transmission to separate destinations. The switch is
 constructed out of self-routing, non-blocking interconnection networks. Bursty data
 (variable length packets) and continuous data are reformatted into small synchronized,
 fixed length ``switch packets'' which are used to route the networks. At the switch input,
 unused switch packets are discarded by a concentrator network. A sort-to-copy and a
 copy network are used to provide the broadcast mechanism. A sort-to-destination and an
 expander network route the switch packets to their destinations. The use of multiple
 channels per user allows the variety of interconnections. For n user channels, the sorting
 networks have O(n(\log_2 n)^2 elements each; the other networks have O(n\log_2 n).
 VLSI chips implementing the networks are also described; each handles 64 5 Mbit
 channels.

 Keywords: fast packet switches; hardware implementation; self-routing networks

D. Huynh, H. Kobayashi, and F. F. Kuo, "Optimal design of mixed-media packet-switching
networks: routing and capacity assignment," IEEE Transactions on Communications, vol.
COM-25, pp. 158-169, 1977. 

 Abstract: This paper considers a mixed-media packet-switched computer
 communication network which consists of a low-delay terrestrial store-and-forward
 subnet combined with a low-cost high-bandwidth satellite subnet. We show how to route
 traffic via ground and/or satellite links by means of static, deterministic procedures and
 assign capacities to channels subject to a given linear cost such that the network average
 delay is minimized.

 Keywords: packet switching; ALOHA; routing algorithm; time in system; cost; multiple
 access

John H. Hartman and John K. Ousterhout, "Zebra: A Striped Network File System," Tech. Rep.
UCB/CSD 92/683, University of California, Berkeley, Apr. 1992. 

 Abstract: Zebra is a striped network file system. The intent of such a system is to increase
 I/O bandwidth by allowing concurrent access to data which is distributed over multiple file
 servers, without the need for expensive specialized hardware. Unfortunately, such a
 scheme is generally efficient only for large files, since the overhead involved for small files
 is relatively high. Zebra addresses this problem by striping on the basis of client streams,
 rather than individual files. It borrows this philosophy (i.e., saving up small writes to enable
 larger transfers) from Sprite's Log-Structured File System. Other goals addressed by
 Zebra include uniformity of server loads, a simple parity mechanism (for efficient failure
 recovery), scalability of the system, and greater efficiency and cost effectiveness of the
 (storage) servers.

R. G. Herrtwich and L. Wolf, "A system software structure for distributed multimedia systems," in 
5th ACM SIGOPS European Workshop, (Le Mont Saint-Michel, France), Sept. 1992. 

 Abstract: Digital audio and video are different from traditional media in their
 time-criticalness and high bandwidth requirements. These requirements and the fact that
 typical multimedia applications perform only a few operations on the continuous-media
 data suggest the use of new techniques for data handling in distributed, integrated, digital
 multimedia systems. This paper proposes a system software structure which encapsulates
 the processing of continuous-media data into stream handlers of a real-time environment.
 This environment is controlled by traditional non-real-time functions for resource, buffer
 and continuous-media data stream management. On top of this system, distributed
 multimedia applications can be built.

 Keywords: multimedia; operating systems; continuous media

J. S. Kaufman, "Blocking in a shared resource environment," IEEE Transactions on
Communications, vol. COM-29, no. 10, pp. 1474-1481, 1981. 

 Keywords: Blocking; queueing network; loss system; exponential queueing network;
 BCMP; Erlang B formula

 Annotation: In recent years, considerable effort has focused on evaluating the blocking
 experienced by "customers" in contending for a commonly shared "resource". The
 customers and resource in question have typically been messages and storage space in
 message storage applications or data streams and bandwidth in data multiplexing
 applications. The model employed in these studies, a multidimensional generalization of the
 classical Erlang loss model, has bee

J. J. Kulzer and W. A. Montgomery, "Statistical switching architectures for future services," in 
International Switching Symposium, (Florence), p. paper 43a1, 1984. 

 Abstract: This paper explores the trends to find efficient ways to trans- port variable 
 bandwidth or "bursty" information, a continuum of techniques that can be used for such
 transport and some of the implications that such approaches have on system architecture.
 It concludes that a statistical switching technique based on a generalized form of packet
 switching , FPS, is particularly attractive.

 Keywords: ATM

Belka Kraimeche and Mischa Schwartz, "Circuit access control strategies in integrated digital
networks," in Proceedings of the Conference on Computer Communications (IEEE Infocom),
(San Francisco, California), pp. 230-235, IEEE, Apr. 1984. 

 Abstract: Traffic access control strategies are investigated for circuit-switched demand
 access to a common digital transmission facility by a population of heterogeneous users,
 with different average holding times and differing bandwidth (bit rate) requirements. In
 order for this type of integrated communication system to handle its traffic demands with
 high efficiency and flexibility, close control of access and switching at the input node is
 required. We introduce a class of restricted access (RA) control strategies capable of
 providing improved system performance. These access strategies consist of clustering the
 set of user types and limiting the number of resource (bandwidth) units occupied by user
 types in each cluster. A recursive scheme for efficient system performance computation is
 presented. The optimum design of an RA strategy is formulated as a combinatorial
 optimization problem, and special case is treated via a simple algorithm. The RA strategy
 augmented by a priority scheme is shown to provide improved performance.

 Keywords: access control

Edward K. Lee, Peter M. Chen, John H Hartman, Ann L. Chervenak~Drapeau, Ethan L. Miller,
Randy H. Katz, Garth A. Gibson, and David A. Patterson, "RAID-II: A scalable storage
architecture for high-bandwidth network file service," Tech. Rep. UCB/CSD 92/672, University
of California, Berkeley, Feb. 1992. 

 Abstract: RAID-II is an experimental storage architecture to provide both high-
 bandwidth transfer of large data files, and low-latency access to smaller files. Is
 development reflects the incresing number of applications for network-based workstation
 environments requiring high I/O bandwidth (e.g, multimedia), while acknowledging the
 continuing need for such environments to provide low-latency service for smaller files. The
 advent of high-bandwidth network technology makes possible RAID-II's approach,
 which is to use the network as the backplane for its storage system. This provides the
 system with a far greater scalability than current mass storage systems used for super
 computers. Other features include the separation of file servers from storage servers to
 allow high bandwidth connections which bypass the file server, as well as two modes of
 transfer, since low-latency transfers have different needs. The Log-Structured File
 System (LFS) is used because the large sequential data segments stored on disk facilitate
 high speed transfer.

Phillip Lougher and Doug Shepherd, "The Design of a Storage Server for Continuous Media," 
Computer Journal, vol. 36, pp. 32-42, Feb. 1993. 

 Abstract: Present file systems are inadequate for continuous media storage for three
 major reasons: inadequate capacity, inadequate bandwidth and no provision for the
 real-time scheduling needed to guarantee steady continuous retrieval of data streams. The
 issues are addressed by the Continuous Media Storage Server (CMSS). Use of disk
 striping (multiple disks) enhances capacity and provides an increase in bandwidth linear
 with the number of extra disks (by using concurrent retrieval). Specialized disk layout,
 including a separate disk for meta data and a new block indexing scheme, along with a disk
 scheduler which handles hard and soft deadlines, make continuous retrieval of multiple
 streams possible.

 Keywords: file system; storage server; multimedia; continuous media; disk storage

T. Lang, M. Valero, and I. Alegre, "Bandwidth of crossbar and multiple-bus connections for
multiprocessors," IEEE Transactions on Computers, vol. C-31, pp. 1227-1234, Dec. 1982.

Marco Ajmone Marsan and Guido Albertengo, "Integrated voice and data network," Computer
Communications, vol. 5, pp. 119-128, June 1982. 

 Abstract: A bus-structured local area communications network architecture is presented.
 The network is based on a unidirectional bus system, over which packets are broadcast.
 Stations are connected to the communications channel by means of three passive taps. The
 multiple-access protocol is an extension of the register-insertion scheme used for loop
 networks; it is extremely efficient and guarantees that the packet access delay is less than
 a known maximum. Message-based priority functions can be introduced with a minimum
 overhead. The channel-access protocol allows an efficient integration of real-time traffic,
 such as packetized voice, with bursty data traffic. Simulation results quantitatively
 demonstrate the protocol performance.

 Keywords: computer networks; local area networks; multiple access protocols; simulation;
 packet voice; integrated services

 Annotation: reserved voice bandwidth

Samy A. Mahmoud, Wai Yip Chan, J. Spruce Riordon, and Salah E. Aidarous, "An Integrated
Voice/Data System for VHF/UHF Mobile Radio," IEEE Journal on Selected Areas in
Communications, vol. SAC-1, pp. 1098-1111, Dec. 1983. 

 Abstract: This paper presents the basic architecture and performance of a mobile radio
 multiaccess voice/data system. Natural pauses in conversational speech allow bandwidth
 saving through interleaving of data packets and talkspurts from different voice sources. A
 speech detector designed specifically for the mobile environment is presented. Blocking
 and delay performance of the multiaccess uplink is analyzed for voice traffic, assuming no
 traffic effects from the low priority data packets. Performance results from simulation are
 then presented for two downlink strategies in a two-hop virtual circuit in which a base
 station acts as a relay. The results verify also that the uplink analysis is valid for low voice
 traffic. For the data traffic, simulation results are presented in terms of data packet
 transmission delay and probability of collision with talkspurts. The results indicate that data
 flow may be limited by the collision factor. This work concludes that relative to
 conventional radio telephoning in which two channels are dedicated to each
 transmitter/receiver pair, a bandwidth reduction of 30-35 percent can be achieved.

 Keywords: packet voice; speech coding

 Annotation: CVSD coder at 16 kb/s. Describes silence detector consisting of zero crossing
 detector and averaging of energy. Also provides statistics on first moments and distribution
 of talkspurts and pauses. The exponential distribution fits the talkspurts well, but not the
 silence periods.

John Nagle, "Congestion control in IP/TCP internetworks," ACM Computer Communication
Review, vol. 14, pp. 11-17, Oct. 1984. 

 Abstract: Congestion control is a recognized problem in complex networks. We have
 discovered that the Department of Defense's Internet Protocol (IP), a pure datagram
 protocol, and Transmission Control Protocol (TCP), a transport layer protocol, when used
 together, are subject to unusual congestion problems caused by the interactions between
 the transport and datagram layers. In particular, IP gateways are vulnerable to a
 phenomenon we call congestion collapse, especially when such gateways connect
 networks with widely different bandwidth. We have developed solutions that prevent
 congestion collapse.

 Keywords: congestion control; TCP; IP; source quench

R. Nocker, "A time-division multiplex communication network featuring decentralized switching
and reduced bandwidth," Siemens Forschungs- und Entwicklungsberichte, vol. 6, no. 4, pp.
198-203, 1977. 

 Abstract: If a communication network level is without a switching center, but each station
 has a switching unit of its own, that network level is said to feature decentralized switching.
 Communication networks of this type in which all stations have multiple access to the
 transmission medium have been under discussion for some time.

 Keywords: communication network; TDM; time division multiplex; decentralized control;
 reduced bandwidth

John Ousterhout and Fred Douglis, "Beating the I/O Bottleneck: A Case for Log-Structured File
Systems," Tech. Rep. UCB/CSD 88/467, University of California, Berkeley, Oct. 1988. 

 Abstract: As CPU speeds continue to outgrow I/O bandwidth, it is likely that I/O will
 increasingly be a bottleneck in the performance of systems. As secondary storage (e.g.,
 disks) becomes less expensive, it will likely be economically beneficial to trade off disk
 space for bandwidth if this is feasible. This paper examines a few possible techniques for
 improving I/O bandwidth, in particular file caching, caching with battery backup, cache
 logging, and, most significantly, log-structured file systems. Since file caching has the
 potential to eliminate 90% or more of disk read accesses, writes will dominate disk access
 time (which itself is dominated by seek times). Seek times (for writes) can be practically
 eliminated by using a log-strucured append-only file system. A scheme for occasional disk
 reads from such a system with performancwe at least as good as traditional (i.e., UNIX) file
 systems is described. When log-structured file systems are used in conjunction with disk
 arrays (of, say, 100 disks), I/O bandwidth increases of up to 1000 fold may be possible.

David A. Patterson, Garth Gibson, and Randy H. Katz, "A Case for Redundant Arrays of
Inexpensive Disks (RAID)," in Proceedings of the 1988 ACM Conference on Management of
Data (SIGMOD), (Chicago, IL), pp. 109-116, June 1988. 

 Abstract: As processor and memory speeds increase at an exponential rate and single
 disk access times remain relatively constant, it is apparent that I/O bandwidth is likely to
 become a bottleneck in the performance of systems. One way to address this problem is by
 using disk arrays, i.e., sets of relatively inexpensive disks which can improve I/O bandwidth
 via parallel access. The problem with this approach is that simply using disk arrays can
 drastically reduce reliability. The approach of RAID is to use redundant disks of check data
 to bring reliability up to acceptable levels (i.e., failure rates better than expected useful life
 of the disks). Five levels of the RAID design are presented to address the issues of
 overhead cost (in terms of number of disks), useable storage capacity, and efficiency per
 disk for various read and write scenarios (i.e, large vs. small). These issues were
 considered in terms of \em data rates (supercomputer applications) and \em I/O rates
 (transaction systems). Level 5 RAID provides the best all around performance by
 distributing check data across the data disks to increase parallelism.

Lawrence G. Roberts, "The evolution of packet switching," Proceedings of the IEEE, vol. 66, pp.
1307-1313, Nov. 1978. 

 Abstract: Over the past decade data communications has been revolutionized by a
 radically new technology called packet switching. In 1968 virtually all interactive data
 communication networks were circuit switched, the same as the telephone network.
 Circuit switching networks preallocate transmission bandwidth for an entire call or session.
 However, since interactive data traffic occurs in short burst 90 percent or more of this 
 bandwidth is wasted. Thus, as digital electronics become inexpensive enough, it became
 dramatically more cost-effective to completely redesign communication networks,
 introducing the concept of packet switching where the transmission bandwidth is
 dynamically allocated, permitting many users to share the same transmission line
 previously required for one user. Packet switching has been so successfull, not only in
 improving the economics of data communications but in enhancing reliability and functional
 flexibility as well, that in 1978 virtually all new data networks being built throughout the
 world are based on packet switching. An open question at this time is how long will it take
 for voice communications to be revolutionized as well by packet switching technology. In
 order to better understand both the past and future evolution of this fast moving technology,
 this paper examines in detail the history and trends of packet switching.

 Keywords: packet switching; survey; history of computing; ARPAnet; X.25; Tymnet;
 virtual circuits; datagram

Kenneth Salem and Hector Garcia-Molina, "Disk Striping," in Proceedings of the 2^{nd}
International Conference on Data Engineering, pp. 336-342, ACM, Feb. 1986. 

 Abstract: This paper investigates the effectiveness of a technique called \em disk striping
 for increasing I/O bandwidth. The basic idea of disk striping is to divide data blocks into
 subblocks which can be stored on separate disks and accessed concurrently. Ideally, using
 n disks would reduce access times by a factor of 1/n. However, striping can incur additional
 overheads, such as processing time and disk access latencies (as n increases, the latencies
 will approach their maximum values since the overall access is bound by the slowest
 access). A computational model is presented to investigate the trends in effectiveness of
 striping when varying n. The effectiveness of various enhancements to reduce the
 overheads are also examined. An application is modelled to study the trends in more
 realistic situations. In addition, future CPU speeds and file sizes are considered, to show
 what improvements are possible when performance is not CPU bound.

E. Szurkowski, "A High Bandwidth Local Computer Network," in Proceedings Computer
Communications Networks; Computers and Communications: Interfaces and Interactions
(Compcon), pp. 98-103, Fall 1978. also in \cite Fouad Tobagi and Noel Gonzalez-Cawley, "On
CSMA-CD local networks and voice communication," in Proceedings of the Conference on
Computer Communications (IEEE Infocom), (Las Vegas, Nevada), pp. 122-127, IEEE,
March/April 1982. 

 Abstract: We consider in this paper local networks of the CSMA-CD broadcast bus type,
 exemplified by Ethernet, and investigate their performance when supporting voice
 communication. For such real-time application, we define network performance as the
 maximum number of voice source accomodated for a given maximum delay requirement
 and a tolerable packet loss rate. We study the effect on this performance of various system
 parameters such as channel bandwidth, vocoder rate, delay requirement and packet loss
 rate.

 Keywords: Ethernet; CSMA/CD; packet voice; local area networks; integrated services

D. J. Wilcox and I. S. Gibson, "Numerical Laplace transformation and inversion in the analysis of
physical systems," International Journal for Numerical Methods in Engineering, vol. 20, pp.
1507-1519, 1984. 

 Abstract: The paper develops a new discrete transform pair which economizes on the
 number of samples of F(s) and f(t) compared with the standard pair. Whereas the standard
 pair used uniform sampling along the Bromwich contour, the new method allows a
 progressive increase in the sampling interval, thereby reducing the number of samples
 required to meet a specified bandwidth requirement. In the case of the computation of
 transient responses of physical systems, a reduction by a factor of 10 is not unrealistic. The
 new transform pair is based on the generalization of fundamental sampling principles. In
 particular, this leads to an integration contour different from the Bromwich contour.

 Keywords: Laplace transform; numerical methods; inversion

C. J. Weinstein, A. J. McLaughlin, and T. Bially, "Efficient multiplexing of voice and data in
integrated digital networks," in Conference Record of the International Conference on
Communications (ICC), (Seattle, Washington), pp. 21.1.1-21.1.7, IEEE, June 1979. 

 Abstract: The development of digital communications systems which can accomodate
 voice and data traffic in an integrated manner is a subject of much current interest and
 activity. In this paper, we describe some basic issues and tradeoffs in the statistical
 multiplexing of voice and data in common digital links, with particular attention to the
 statistical properties of voice traffic and to the effectiveness of various proposed techniques
 for integrating voice and data in a bandwidth efficient manner. Techniques for the efficient
 multiplexing of voice and data in a variety of system configurations, including
 demand-assigned broadcast satellite networks, multi-node terrestrial topologies , and
 distributed-control local networks, are discussed. Communication strategies considered
 include packet techniques for both voice and data, and hybrid methods which employ
 TDMA-based circuit-switching for voice and packet-switching for data. Special attention
 is directed at the exploitation of the silence periods in voice conversation and at the effect of
 variable voice transmission rates.

 Keywords: packet voice; silence detection; voice source model; integrated services

Catherine G. Wolf, "Video conferencing: delay and transmission characteristics," in 
Teleconferencing and electronic communication (L. A. Parker and C. H. Olgren, eds.), pp.
184-188, unknown publisher, 1982. 

 Abstract: This paper discusses the effects of audio-video transmission delay on
 communication in a two-way, interactive video conferencing system. The study was
 conducted with the Bell System's market trial video conferencing system,
 PICTUREPHONE Meeting Service. This system allows groups of up to six people in each
 of two locations to confer ``face-to-face'' with graphics support. There are at least two
 potential sources of transmission delay in video conferencing. First, video conferencing is
 typically employed over long distances where satellite transmission may be most
 economical. Due to the distance of a satellite from the earth, satellite transmission adds a
 round-trip increment of about 510 msec to transmission time. Secondly, attempts to reduce
 the bandwidth required for transmission of the video signal characteristics employ picture
 processing techniques which add delay to the transmission time. The exact amount of delay
 depends on the picture processing techniques employed.

 Keywords: delay; multimedia; videoconferencing; human factors

 Annotation: Tested delays of 0, 330, and 840 ms. The 840 ms delay and half-duplex
 exacerbates simultaneous speech problems, but does not affect task performance. 330 ms
 shows only very small difference in subjective evaluation of being interrupted or difficulty to
 communicate compared to 0 ms. [Sc]