% This is the "Multimedia Operating Systems" bibliography file which
% contains references to papers and technical reports on operating
% systems support for multimedia or continuous media. 
%
% You may ftp the latest version from 
%   ftp.cse.ucsc.edu:/pub/bib/mmos.bib
% 
% The bibliography is maintained by:
%   Takeshi Aoki
%   Computer Network Systems Laboratory
%   Fujitsu Laboratories Ltd., JAPAN
%   @string{email="aokit@flab.fujitsu.co.jp"}
%
% Any help and suggestion to keep the file updated is greatly appreciated. 
% Please email to me at the above email address.
%
% Acknowledgements:
%   I owe a lot of inputs to
%   Roland Mechler, University of British Columbia.
%   The abstracts of some of papers are his own descriptions.
%
% History of this file:
%   ver.1.0	Oct 1993
%   ver.2.1     Apr 1994

@STRING{ACMCOMSYS=tocs}

@STRING{ACMINFOSYS="ACM Transactions on Information Systems"}

%%
%% University of California at San Diego
%%
%% ftp site:
%%    cs.ucsd.edu:/pub/multimedia
%%

@TECHREPORT{Rang9302:Research,
AUTHOR="Rangan, P. V.",
TITLE="Research Contributions of the {UCSD Multimedia Laboratory}",
TYPE="Report",
INSTITUTION="Multimedia Laboratory, Department of Computer Science and
Engineering, University of California, San Diego",
MONTH=feb,
YEAR=1993,
ENTRYBY=Ao
}

@INCOLLECTION{Rang9104:C,
AUTHOR="Rangan, P. V. and Vin, H. M.",
TITLE="Multimedia Conferencing as a Universal Paradigm for
Colaboration",
BOOKTITLE="Multimedia - Principles, Systems, and Applications",
EDITOR="Kjelldahl, Lars",
CHAPTER=14,
PUBLISHER="Springer-Verlag",
NOTE="(Proceedings of Eurographics Workshop on Multimedia Systems,
Applications, and Interaction, Stockholm, Sweden)",
MONTH=apr,
YEAR=1991,
ABSTRACT="A paradigm for collaborative work via multimedia is presented,
with the key abstraction being the {\em conference}. Conferences are
treated as first class objects, and may contain both simple participants
(i.e., users) or other conferences (i.e., {\em sub-conferences}) as
participants. Conferences also provide means for access control. A
taxonomy for multimedia conferencing, as well as a set of operations on
conferences are presented here. Examples of how the paradigm may be used
to model scenarios of multimedia collaboration are included, but no
details of implementation or practical results are given.",
ENTRYBY=Ao
}

%%%
%%% UCB Continuous Media File System
%%% 

@ARTICLE{Ande9211:File,
AUTHOR="Anderson, D. P. and Osawa, Y. and Govindan, R.",
TITLE="A File system for Continuous Media",
JOURNAL=tocs,
VOLUME=10,
NUMBER=4,
PAGES="311-337",
MONTH=nov,
YEAR=1992,
ABSTRACT="This paper describes the Continuous Media File System (CMFS),
a file system designed to handle the storage and retrieval of continuous
media data on disk, with emphasis on the provision of real-time
guarantees. The model used for file access is that of {\em sessions},
where multiple concurrent sessions are possible, each with a guasranteed
data rate. The issues examined are disk layout for deterministic
latency, acceptance criteria (for granting new sessions while
maintaining guarantees for existing ones), as well as disk scheduling
policies and related performance issues.",
ENTRYBY=Ao
}

@TECHREPORT{Ande9108:Real,
AUTHOR="Anderson, David P. and Osawa, Yoshitomo and Govindan, Ramesh",
TITLE="Real-Time Disk Storage and Retrieval of Digital Audio/Video Data",
TYPE="Technical Report",
INSTITUTION="Computer Science Divison, University of California,
Berkeley",
NUMBER="UCB/CSD 91/646",
MONTH=aug,
YEAR=1991,
ABSTRACT={The Continuous Media File System, CMFS, supports real-time
storage and retrieval of continuous media data (digital audio and
video) on disk.  CMFS clients read or write files in "sessions", each
with a guaranteed minimum data rate.  Several sessions can exist
concurren tly, sharing a single disk drive.  Clients can concurrently
access non-real-time files on the same disk.  CMFS addresses several
interrelated design issues: 1) real-time semantics of sessions; 2)
disk layout; 3) acceptance test for new sessions, and 4) disk
scheduling policy.  We use simulation to compare different design
choices and to estimate the performance of CMFS under various load
conditions and hardware assumptions.},
URL="ftp://tr-ftp.cs.berkeley.edu/pub/tech-reports/csd/csd-91-646",
ENTRYBY=Ao
}

@TECHREPORT{Ande9010:Meta,
AUTHOR="Anderson, D. P.",
TITLE="Meta-scheduling for distributed continuous media",
TYPE="Technical Report",
INSTITUTION="University of California, Berkeley",
NUMBER="UCB/CSD 90/599",
MONTH=oct,
YEAR=1990,
ABSTRACT="In using system resources (e.g., CPU, disk, network), a CM
application typically requires guarantees on factors such as maximum
delay and minimum throughput. A single request may span several
resources (perhaps across several network hosts), and the application
may require guarantees on the total delay and throughput. The scheduling
required to provide this is called end-to-end scheduling, or {\em
meta-scheduling}. This paper presents a meta-scheduling system based on
the {\em CM-resource model}. {\em Workahead scheduling} and negotiation
of service parameters are used to acheive these goals. Other issues
addressed by the system are an abstract interface for the scheduler
(e.g., for requesting resources), and co-existence of real-time and non
real-time work.",
URL="ftp://tr-ftp.cs.berkeley.edu/pub/tech-reports/csd/csd-90-599",
ENTRYBY=Ao
}

@TECHREPORT{Govi9207:Operating,
AUTHOR="Govindan, Ramesh",
TITLE="Operating Systems Mechanisms for Continuous Media",
INSTITUTION="University of California, Berkeley",
NUMBER="UCB/CSD 92/697",
MONTH=jul,
YEAR=1992,
ABSTRACT={Support for digital audio and video (continuous media) as
part of the human/computer interface is an important direction for
computer systems research.  There are various ways to incorporate
continuos media (CM) in computer systems; in the integrated approach,
CM data is h andled by user-level programs on general-purpose
operating systems such as Unix or Mach.  Integrated CM applications
handle data at high rates, with strict timing requirements and often
in small "chunks".  Conventional operating systems support for program
execution and local communication may be non-optimal for such
applications.  In particular, conventional mechanisms for process
scheduling and inter address space communication can add significant
overhead for some CM programs.  User/kernel interactions, by which
user-level programs invoke ystem functions, are partly responsible for
this overhead.  We describe new mechanisms for process scheduling and
CM stream communication between virtual address spaces.  These
mechanisms, split-level scheduling and memory-mapped streams, reduce
or eliminate user/kernel interactions by using user/kernel shared
memory for exchanging scheduling and communcation information.  We
demonstrate that, compared with conventional mechanisms for process
scheduling and stream communication, these new mechanisms can reduce
overhead by up to a factor of four.},
URL="ftp://tr-ftp.cs.berkeley.edu/pub/tech-reports/csd/csd-92-697/",
ENTRYBY=Ao
}

@TECHREPORT{Ande8910:Support,
AUTHOR="Anderson, David P. and Tzou, Shin-Yuan and Wahbe, Robert and
Govindan, Ramesh and Andrews, Martin",
TITLE="Support for Continuous Media in the {DASH} System",
TYPE="Technical Report",
INSTITUTION="University of California, Berkeley",
NUMBER="CSD 89/537",
MONTH=oct,
YEAR=1989,
ABSTRACT="Future distributed systems will support continuous media
such as digital audio and video, allowing user programs to convert,
process, store, and communicate continuous-media data.  The DASH
project is developing such a system.  Our work consists of two related
parts.  First we have defined the DASH resource model as a basis for
reserving and scheduling resources (disk, CPU, network, etc.) involved
in end-to-end handling of continuous-media data.  The model uses
primitives that express workload characteristics and performance
requirements, and degines an algorithm for negotiated reservation of
distributed resources.  This algorithm is embodied IP4, a
backwards-compatible extension ofthe Internet Protocol (IP).
        
Second, we have developed a distributed system kernel for use as an
experimental testbed.  The DASH kernel implements the DASH resource
model for scheduling of CPU and network access.  Its virtual memory
system provides efficient data transfer between address spaces. 
Finally, its implementation is structured using object-oriented
programming and message-passing.",
URL="ftp://tr-ftp.cs.berkeley.edu/pub/tech-reports/csd/csd-89-537",
ENTRYBY=Ao
}

%%%
%%% University of Cambridge
%%% Pandora and Pegasus projects
%%% Desk Area Network (DAN)
%%% Multi-Service Storage Architecture
%%%

@TECHREPORT{Mull9209:Pegasus,
AUTHOR="Mullender, S. J. and Leslie, I. M. and McAuley, D.",
TITLE="Pegasus Project Description",
TYPE="Technical Report",
INSTITUTION="University of Cambridge Computer Laboratory",
NUMBER=281,
MONTH=sep,
YEAR=1992,
URL="ftp://ftp.cl.cam.ac.uk/papers/reports/",
ENTRYBY=Ao
}

@ARTICLE{Lesl9301:Pegasus,
AUTHOR="Leslie, Ian M. and McAuley, Derek R. and Mullender, Sape J.",
TITLE="Pegasus --- Operating System Support for Distributed Multimedia
Systems",
JOURNAL=acmosrev,
VOLUME=27,
NUMBER=1,
PAGES="69-78",
MONTH=jan,
YEAR=1993,
ABSTRACT="This paper states the goals and preliminary design of a
distributed operating system which gives support for multimedia
applications. The principal idea is that all types of data (e.g., text,
video, audio) should be given equal status, and their manipulation
should be under operating system control, allowing the system to
maintain all of the properties usually desired of distributed systems,
such as resource sharing, data sharing, security and fault tolerance.
Real-time guarantees (as required for multimedia applications) will be
based on guarantees of the underlying ATM network. The system
configuration is based on individually networked terminals, network
devices (e.g., ATM cameras) and servers. Storage management may be based
on the Log-Structured File System from Sprite. Real-time guarantees may
require periodic multimedia tasks to be divided into mandatory tasks and
optional tasks.",
KEYWORDS="operating systems; multimedia; desk area network; ATM;
packet video; DAN",
ENTRYBY=Ao
}

@PHDTHESIS{Jard9208:Network,
AUTHOR="Jardetzky, P. W.",
TITLE="Network File Server Design for Continuous Media",
SCHOOL="Computer Laboratory, University of Cambridge",
KEY="Jardetzky 92",
NOTE="Published as Technical Report No. 268",
MONTH=aug,
YEAR=1992,
ABSTRACT="This dissertation concentrates on issues related to the
provision of a network based storage facility for digital audio and
video data.  The goal is to demonstrate that a distributed file
service in support of these media may be built without special purpose
hardware.  The main objective is to identify those parameters that
affect file system performance and provide the criteria for making
desirable design decisions.",
URL="ftp://ftp.cl.cam.ac.uk/papers/reports/TR268-pwj-cmfileserver.ps.Z",
ENTRYBY=Sc
}

@PHDTHESIS{Sree9210:Synchronisation,
AUTHOR="Sreenan, C. J.",
TITLE="Synchronisation Services for Digital Continuous Media",
SCHOOL="Computer Laboratory, University of Cambridge",
NOTE="Published as Technical Report No. 292",
MONTH=oct,
YEAR=1992,
ABSTRACT="The development of broadband ATM networking makes it
attractive to use computer communication networks for the transport of
digital audio and motion video.  Coupled with advances in workstation
technology, this creates the opportunity to integrate these continuous
information media within a distributed computing system.  Continuous
media have an inherent temporal dimension, resulting in a set of
synchronisation requirements which have real-time constraints.  This
dissertation identifies the role and position of synchronisation, in
terms of the support which is necessary in an integrated distributed
system.  This work is supported by a set of experiments which were
performed in an ATM inter-network using multi-media workstations, each
equipped with an Olivetti Pandora Box.",
URL="ftp://ftp.cl.cam.ac.uk/papers/reports/TR292-cjs-synchronisation.ps.Z",
ENTRYBY=Sc
}

@PHDTHESIS{Nico9105:Distributed,
AUTHOR="Nicolaou, C.",
TITLE="A Distributed Architecture for Multimedia Communication Systems",
SCHOOL="Computer Laboratory, University of Cambridge",
NOTE="Published as Technical Report No. 220",
MONTH=may,
YEAR=1991,
URL="ftp://none",
ENTRYBY=Ao
}

@PHDTHESIS{McAu90:Protocol,
AUTHOR="McAuley, D. R.",
TITLE="Protocol Design for High Speed Networks",
SCHOOL="Computer Laboratory, University of Cambridge",
NOTE="Published as Technical Report No. 186",
ABSTRACT="Improvements in fibre optic communication and in VLSI for
network switching components have led to the consideration of building
digital switched networks capable of providing point to point
communication in the gigabit per second range.  Provision of
bandwidths of this magnitude allows the consideration of a whole new
range of telecommunications services, integrating video, voice, image
and text.  These multi-service networks have a range of requirements
not met by traditional network architectures designed for digital
telephony or computer applications.  This dissertation describes the
design, and an implementation, of the Multi-Service Network
architecture and protocol family, which is aimed at supporting these
services.  Asynchronous transfer mode networks provide the basic
support required for these integrated services, and the Multi-Service
Network architecture is designed primarily for these types of
networks.  The aim of the Multi-Service protocol family is to provide
a complete architecture which allows use of the full facilities of
asynchronous transfer mode networks by multi-media applications.  To
maintain comparable performance with the underlying media, certain
elements of the MSN protocol stack are designed with implementation in
hardware in mind.  The interconnection of heterogeneous networks, and
networks belonging to different security and administrative domains,
is considered vital, so the MSN architecture takes an internetworking
approach.",
YEAR=1990,
URL="ftp://ftp.cl.cam.ac.uk/papers/reports/TR186-dm-thesis.ps.gz",
ENTRYBY=Sc
}

@PHDTHESIS{Dixo9201:System,
AUTHOR="Dixon, M. J.",
TITLE="System Support for Multi-Service Traffic",
SCHOOL="Computer Laboratory, University of Cambridge",
MONTH=jan,
YEAR=1992,
ENTRYBY=Ao
}

%% Lancaster University, U.K.
%%
%% More papers available from the following ftp site:
%%     ftp.comp.lancs.ac.uk:/pub/mpg

@ARTICLE{Loug9302:Continuous,
AUTHOR="Lougher, Phillip and Shepherd, Doug",
TITLE="The Design of a Storage Server for Continuous Media",
JOURNAL=compj,
VOLUME=36,
NUMBER=1,
PAGES="32-42",
MONTH=feb,
YEAR=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",
ENTRYBY=Ao
}

@TECHREPORT{Blai92:Meeting,
AUTHOR="Blair, Gordon S. and Coulson, Geoff",
TITLE="Meeting the Real-time Synchronization Requirements of
Multimedia in Open Distributed Processing",
INSTITUTION="Lancaster University",
NOTE="(internal report)",
YEAR=1992,
ABSTRACT="This paper looks specifically at the issues of synchronization
in providing multimedia services for Open Distributed Processing (ODP).
Three forms of synchronization are identified: stream, continuous and
event based. In providing stream synchronization, an extension to the
ANSA model is proposed to provide the {\em stream} abstraction allowing
QoS parameters to be negotiated via a three-way handshake protocol, and
thus using a rate-based approach to stream synchronization. Continuous
synchroniazation is based on matching rates and start times among
streams, and the design relies on the existence of a global clock.
Real-time events for event based synchronization are implemented using a
technique called {\em bounded invocation}, and an algorithm for
providing related guarantees is given.",
ENTRYBY=Ao
}

@TECHREPORT{Coul92:Supporting,
AUTHOR="Coulson, G. and Blair, G. S. and Stefani, J. B. and Horn, F. and
Hazard, L.",
TITLE="Supporting the Real-time Requirements of Continuous
Media in Open Distributed Processing",
INSTITUTION="University of Lancaster",
NUMBER="MPG-92-35",
YEAR=1992,
ABSTRACT="This paper examines ways in which the current model for Open
Distributed Processing (ODP) can be extended to accomodate continuous
media. The model currently provides {\em discrete interfaces}, which are
inadequate for operations on continuous media objects which may, for
example, require periodic invocation. {\em Continuous interfaces} are
proposed, which provide for QoS specification via {\em annotation}, and
hide the underlying issues of allowing continuous operations. The impact
on ODP is considered with respect to both computational (language,
interface) issues and engineering (mechanism, implementation) issues.
Continuous interfaces, as well as {\em reactive objects} to allow
real-time resource allocation, synchronization and scheduling, address
the computational issues. The engineering issues are addressed in part
by {\em streams} and {\em QoS Managers}, which provide an abstraction
for handling QoS guarantees.",
ENTRYBY=Ao
}

@TECHREPORT{Camp93:Integrated,
AUTHOR="Campbell, A. and Coulson, G. and Garcia, F. and Hutchison, D.
and Leopold, H.",
TITLE="Integrated Quality of Service for Multimedia Communications",
INSTITUTION="Lancaster University",
NUMBER="MPG-93-17",
YEAR=1993,
ENTRYBY=Ao
}

%%%
%%% Synchronization protocols and properties
%%%

@ARTICLE{Li92:R,
AUTHOR="Li, Li and A. Karmouch, N. D. Georganas",
TITLE="Real-time Synchronization Control in Multimedia
Distributed Systems",
JOURNAL=ccr,
VOLUME=22,
NUMBER=3,
PAGES="79-86",
YEAR=1992,
ABSTRACT="This paper describes an approach to providing real-time
multimedia transmission (and synchronization) in a network environment.
The two main issues of {\em continuity} and {\em matching} are
addressed. A system design compatible with the OSI layered protocol
model is used, encorporating two subsystems (application and transport).
The Transport Sub-system is compatible with ATM networks. A Segment
Constraint Control (SCC) system is developed to handle synchronization
between various media. Media {\em groups} can be specified at the
application level, where strict time-matching or loose synchronization
can also be specified. Strict time-matching is handled by the SCC,
whereasa loose synchronization is handled at the application level (by
the Activity Constraint Control).",
ENTRYBY=Ao
}

%%%
%%% Scheduling
%%%

@TECHREPORT{Auds90:Real,
AUTHOR="Audsley, N. and Burns, A.",
TITLE="Real-time System Scheduling",
INSTITUTION="University of York",
NUMBER="YCS 134",
YEAR=1990,
ABSTRACT="This paper is a survey of issues involved in the scheduling of
real-time sytems. Major issues covered include dealing with periodic and
non-periodic processes, transient overloads, priority inversion,
blocking, calculation of execution times, etc., with respect to
uniprocessor, multiprocessor and distributed systems. A brief look at
the approach of some current real-time kernels to scheduling is given,
as well as a summary of the major problems which need to be addressed in
order to produce adequate real-time systems.",
NOTE="First year report, ESPRIT BRA Project (3092)",
URL="ftp://ftp.cs.york.ac.uk/pub/realtime/papers/YCS134.ps.Z",
ENTRYBY=Ao
}

%%%
%%% Multimedia file systems or storage server
%%%

@ARTICLE{Gemm9201:Principles,
AUTHOR="Gemmell, J. and Christodoulakis, S.",
TITLE="Principles of Delay Sensitive Multimedia Data Storage and
Retrieval",
JOURNAL=tois,
VOLUME=10,
NUMBER=1,
PAGES="51-90",
MONTH=jan,
YEAR=1992,
ABSTRACT="This paper outlines a theoretical framework for the retrieval
(from a dedicated storage device) of delay-sensitive data for the
purpose of playback. While delay-sensitive data includes digital audio,
animation and video, this paper examines only digital audio, claiming
that the results extend to the other data types as well. The principal
focus is on buffering requirements between transfer and playback,
including the effect of delays (both artificial and unavoidable) on
these requirements. Issues such as multiple channel playback, sharing
storage devices, and storage placement strategies (e.g., interleaved
placement) are also explored within the context of the given framework.",
ENTRYBY=Ao
}

@TECHREPORT{Hart9204:Zebra,
AUTHOR="Hartman, John H. and Ousterhout, John K.",
TITLE="Zebra: A Striped Network File System",
INSTITUTION="University of California, Berkeley",
NUMBER="UCB/CSD 92/683",
MONTH=apr,
YEAR=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.",
ENTRYBY=Ao
}

@ARTICLE{Rose9202:Implementation,
AUTHOR="Rosenblum, M. and Ousterhout, J. K.",
TITLE="The Design and Implementation of a Log-Structured File System",
JOURNAL=tocs,
VOLUME=10,
NUMBER=1,
PAGES="26-52",
MONTH=feb,
YEAR=1992,
ENTRYBY=Ao
}

@INPROCEEDINGS{Patt8806:C,
AUTHOR="Patterson, David A. and Gibson, Garth and Katz, Randy H.",
TITLE="A Case for Redundant Arrays of Inexpensive Disks
({RAID})",
BOOKTITLE="Proceedings of the 1988 ACM Conference on Management of Data
(SIGMOD)",
ADDRESS="Chicago, IL",
PAGES="109-116",
MONTH=jun,
YEAR=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.",
ENTRYBY=Ao
}

@TECHREPORT{Lee9202:RAID,
AUTHOR="Lee, Edward K. and Chen, Peter M. and Hartman, John H and
Chervenak Drapeau, Ann L. and Miller, Ethan L. and Katz, Randy H. and
Gibson, Garth A. and Patterson, David A.",
TITLE="{RAID-II}: A Scalable Storage Architecture for
High-Bandwidth Network File Service",
INSTITUTION="University of California, Berkeley",
NUMBER="UCB/CSD 92/672",
MONTH=feb,
YEAR=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.",
URL="ftp://tr-ftp.cs.berkeley.edu/pub/tech-reports/csd/csd-92-672/all.ps",
ENTRYBY=Ao
}

@TECHREPORT{Oust8810:C,
AUTHOR="Ousterhout, John and Douglis, Fred",
TITLE="Beating the {I/O} Bottleneck: A Case for
Log-Structured File Systems",
INSTITUTION="University of California, Berkeley",
NUMBER="UCB/CSD 88/467",
MONTH=oct,
YEAR=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.",
ENTRYBY=Ao
}

@MISC{Ciot9204:R,
AUTHOR="Ciotti, Bob",
TITLE="Session Reservable File Systems ({SRFS})",
HOWPUBLISHED="A report prepared for the CRAY User Group Meeting, Berlin
Germany",
MONTH=apr,
YEAR=1992,
ABSTRACT="Session Reservable File Systems (SRFS) is a mechanism for
integrated file space management for existing OSs such as UNIX or
UNICOS. It is intended primarily for a scientific application setting,
where jobs typically require a large percentage of the existing system
file space for temporary and/or output files. A user can reserve a
specific amount of file space and jobs run by that user will be
guaranteed access to that amount of file space. Any unused reserved file
space is returned to the system when the user logs out.",
ENTRYBY=Ao
}

@ARTICLE{Stae9303:C,
AUTHOR="Staehli, Richard and Walpole, Jonathon",
TITLE="Constrained-Latency Storage Access",
JOURNAL=ieeecomp,
PAGES="44-53",
MONTH=mar,
YEAR=1993,
ABSTRACT="A class of problems (Constrained-Latency Storage Access or
CLSA) is identified which require guarantees of the latency time of
retrieval of data from storage devices (archival devices, such as disks
are of primary interest here). The approach used for reducing latency is
data prefetching (i.e., having the {\em fetch} event precede the {\em
demand} event for a particular storage access). This requires some
degree of predictability of data requirements. Applications can be
categorized on the basis of how they determine their demand events as
{\em sequential}, {\em scripted}, or {\em dynamic}. An alternate
categorization, based on the information needed to predict demand
events, is {\em periodic}, {\em predictable}, or {\em unpredictable}.
The various categories of applications are examined to see whether and
how constrained-latency guarantees can be met.",
ENTRYBY=Ao
}

@INPROCEEDINGS{Sale8602:,
AUTHOR="Salem, Kenneth and Garcia-Molina, Hector",
TITLE="Disk Striping",
BOOKTITLE="Proceedings of the 2$^{nd}$ International Conference on Data
Engineering",
ORGANIZATION="ACM",
PAGES="336-342",
MONTH=feb,
YEAR=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.",
ENTRYBY=Ao
}

@ARTICLE{Cabr91:R,
AUTHOR="Cabrera, Luis-Felipe",
TITLE="Swift: Using Distributed Disk Striping to Provide
High {I/O} Data Rates",
JOURNAL="Computer Systems",
VOLUME=4,
NUMBER=4,
PAGES="405-436",
MONTH="Fall",
YEAR=1991,
ABSTRACT="Swift is an I/O architecture designed to exploit the idea of
disk striping in distributed systems (i.e., using distributed file
servers) rather than in disk arrays (although disk arrays may be
components of the system). The goal of Swift is to make storage transfer
rates scalable with the number of storage sites, which may be composed
of slower storage technologies, up to the limits imposed by CPU and
network capacities (which are presumably growing at a much faster rate
than storage device tranfer rates). Swift makes use of three major
components: the {\em storage agents}, which manage the actual secondary
storage devices; the {\em distribution agent}, which stripes data over
the storage agents; the {\em storage mediator}, which provides access
and security services and passes a {\em transfer plan} to the
distribution agent. The results of two studies are presented here. One
is a simple Ethernet based prototype to demonstrate the validity of the
distributed disk striping design used by Swift. The other is a
simulation used to examine the scalabilty of the system and determine
the effects of CPU MIPS and network protocol overheads.",
ENTRYBY=Ao
}

%%%
%%% Multimedia operating systems and architectures.
%%%

@ARTICLE{Berr9005:Architecture,
AUTHOR="Berra, P. B. and Chen, C. Y. R. and Ghafoor, A. and Lin, C. C.
and Little, T. D. C. and Shin, D.",
TITLE="Architecture for Distributed Multimedia Database Systems",
JOURNAL=comcom,
VOLUME=13,
NUMBER=4,
PAGES="217-231",
MONTH=may,
YEAR=1990,
ENTRYBY=Ao
}

@ARTICLE{Stan8907:Spring,
AUTHOR="Stankovic, John A. and Ramamritham, Krithi",
TITLE="The Spring Kernel: A New Paradigm for Real-time
Operating Systems",
JOURNAL="Operating Systems Review",
VOLUME=27,
NUMBER=3,
PAGES="54-71",
MONTH=jul,
YEAR=1989,
ABSTRACT="This paper asserts that other current RTOSs (real-time
operating systems) are not predictable enough and are not scalable
because they use the wrong paradigm. I.e., they use the conventional OS
paradigm, only try to make it faster (fast being a relative term).
Fairness and average response time should not be a considerstion in a
RTOS, but rather that critical tasks are guaranteed to complete within
their deadline, and that as many other tasks as possible do so as well.
In the Spring kernel, tasks are categorized as {\em critical}, {\em
essential} and {\em non-essential}. Critical tasks are given an a priori
guarantee and resources are reserved for them, whereas essential tasks
are guaranteed on-line. The Spring kernel is implemented as a network of
multiprocessors, called the SpringNet.",
ENTRYBY=Ao
}

%%%
%%% Multimedia applications
%%%

@ARTICLE{Mack8907:Virtual,
AUTHOR="Mackay, W. E. and Davenport, G.",
TITLE="Virtual Video Editing in Interactive Multimedia Applications",
JOURNAL=cacm,
VOLUME=32,
NUMBER=7,
PAGES="802-810",
MONTH=jul,
YEAR=1989,
ENTRYBY=Ao
}

@ARTICLE{Dann93:Tactus,
AUTHOR="Dannenberg, R. B. and Neuendorffer, T. and Newcomer, J. M. and
Rubine, D. and Anderson, D.",
TITLE="Tactus: Toolkit-level Support for Synchronized Interactive
Multimedia",
JOURNAL="Multimedia Systems Journal",
NOTE="(to appear)",
YEAR=1993,
ENTRYBY=Ao
}

@MISC{Rowe93:V,
AUTHOR="Rowe, Lawrence A. and Larson, Ray A.",
TITLE="A Video-on-Demand System (proposal)",
YEAR=1993,
ABSTRACT="This is a proposal for a video-on-demand (VOD) system to
provide a large (1000 video) storage base for videos to be accessed via
network. The project has two major components: the design of a video
file server, archive server, and client program; the design of a
meta-data database and video browser to access information about the
videos. The archive server is introduced into the system due to the
expectation that secondary storage (i.e., disk arrays) will not be able
to provide cost effective capacity to archive a large number of videos,
and as such tertiary storage devices (e.g., tape jukeboxes) will be
required (with frequently used or expected video data cached on
secondary storage video file servers). The video browser/meta-data
database component of the project entails indexing the video data on the
basis of {\em bibliographic data}, {\em structural data} and {\em
content data}. User interface issues in accessing this data will also be
examined.",
ENTRYBY=Ao
}

@ARTICLE{Fish9301:V,
AUTHOR="Fish, Robert S. and Kraut, Robert E. and Root, Robert W. and
Rice, Ronald E.",
TITLE="Video as an Architecture for Informal Communication",
JOURNAL=cacm,
VOLUME=36,
NUMBER=1,
PAGES="48-61",
MONTH=jan,
YEAR=1993,
ABSTRACT="An experiment exploring the use of video telephony as a means
for informal communication (in a work place setting) is described. The
goal is to use social science to derive a better design for such a
system. Various paradigms seeking to simulate different modes of
face-to-face communication were implemented and available for use by the
subjects of the study (23 volunteers in a research work setting).
Feedback from the participants was used to evaluate the design, in terms
of issues such as usefulness, privacy and intrussiveness. As well, novel
uses for the technology were noted.",
ENTRYBY=Ao
}

@INPROCEEDINGS{Koeg9308:MMIS,
AUTHOR="Koegel, John F. and Rutledge, Lloyd W. and Keskin, Can",
TITLE="{HyOctane}: A {HyTime} Engine for an {MMIS}",
BOOKTITLE="Proceedings ACM Multimedia '93",
ADDRESS="Anaheim, CA",
MONTH=aug,
YEAR=1993,
ABSTRACT="This paper introduces the HyTime international standard for
multimedia data interchange, and briefly describes HyOctane, an
experimental HyOctane engine. HyTime is an extension to SGML (Standard
Generalized Markup Language), a language for specifying the structure of
text documents, allowing the specification of hypermedia documents.
HDTDs (HyTime document type definitions) can be specified using
hyperlinks. HyTime engines such as HyOctane process HyTime specified
documents into a form which can be used by applications.",
ENTRYBY=Ao
}

@ARTICLE{Kret9205:,
AUTHOR="Kretz, F. and Colaitis, F.",
TITLE="Standardizing Hypermedia Information Objects",
JOURNAL=ieeecm,
VOLUME=30,
NUMBER=5,
PAGES="60-70",
MONTH=may,
YEAR=1992,
ABSTRACT="A standard is introduced here in an attempt to provide a
generic, object-oriented abstraction for the use of multimedia data by
various applications. The data itself is modelled as multimedia and
hypermedia information objects (MH objects). In addition, the standard
is to provide for an audiovisual interactive scriptware (AVIS) layer for
specifying scripts to manipulate the MH objects. {\em Content objects}
contain the raw encoded data for a single medium (including information
required for decoding and presentation). A {\em projector object}
contains presentation attributes (e.g., volume, mono/stereo) for a
content object (or a composite object, which specifies relations between
objects in time and space). This paper describes only the first phase of
the development of the MHEG standard, which attempts to establish a
methodology for specifying MH objects and scripts.",
ENTRYBY=Ao
}

@INPROCEEDINGS{Lamp9111:X,
AUTHOR="Lamparter, B. and Effelsberg, W.",
TITLE="{X-MOVIE}: Transmission and Presentation of Digital
Movies under {X}",
BOOKTITLE="Second International Workshop on Network and
Operating System Support for Digital Audio and Video,
Heidelberg",
PUBLISHER="Springer-Verlag, Heidelberg",
MONTH=nov,
YEAR=1991,
ABSTRACT="This report gives a brief description of a system for storing,
transmitting and presenting digital motion picture data making use of an
extension to the X Window System. A key fgeature of this system is that
it requires no special hardware. The movie presentation system, called
X-MOVIE, has three key components: an {\em XClient}, which is an
application making use of a modified {\bf Xlib} library to use X-MOVIE
functions; a {\em Movie Client}, which is really an extension to the X
Server, implementing functions to maninpulate movie data and movie
windows; a {\em Movie Server}, which retrieves movie data from storage
and transmits it to the Movie Client. Issues of movie data formats and
transmission protocols are also touched upon. A new algorithm is
presented for motion video compression using gradual adaptation of the
color lookup table (CLUT).",
ENTRYBY=Ao
}

@INPROCEEDINGS{Lamp9204:MTP,
AUTHOR="Lamparter, B. and Effelsberg, W. and Michl, N.",
TITLE="{MTP}: A Movie Transmission Protocol for Multimedia
Applications",
BOOKTITLE="Multimedia92, 4th IEEE ComSoc International Workshop on
Multimedia Communications, Monterey, California",
MONTH=apr,
YEAR=1992,
ABSTRACT="One method for video compression is the the dithering of
colours to use a CLUT (colour lookup table). Thus a 24 bit
representation of colours can be reduced to 8 bits, with the restriction
that only 256 different colours can be displayed at once. Loading an
entirely new CLUT for each frame can cause the problem that false
colours will be momentarily displayed for the previous frame, a visually
disturbing effect. An algorithm is presented here to reduce this effect
by gradually updated the CLUT during movie transmission, limiting the
number of new colours which can be introduced to the CLUT between
successive images (and also reducing somewhat the number of colours
which can be simultaneously displayed). This was implemented in the
context of the XMovie project.",
ENTRYBY=Ao
}

@INPROCEEDINGS{Pric9308:MHEG,
AUTHOR="Price, Roger",
TITLE="{MHEG}: An Introduction to the future International
Standard for Hypermedia Object Interchange",
BOOKTITLE="Proceedings ACM Multimedia '93",
ADDRESS="Anaheim, CA",
PAGES="121-128",
MONTH=aug,
YEAR=1993,
ABSTRACT="This paper provides a brief introduction to the future MHEG
standard for specifying the coded representation of final form
multimedia and hypermedia objects for interchange between services and
applications. High-level descriptions are given concepts such as {\em
containers} and {\em presentations}. Also described (briefly) are
content data, and standards used for specification thereof, the
inheritence tree for MHEG objects, specification of time and space,
links and synchronization between objects, and the behaviour of objects
in terms of actions and interactions.",
ENTRYBY=Ao
}

@INPROCEEDINGS{Gain9308:Open,
AUTHOR="Gaines, Brian R. and Shaw, Mildred L. G.",
TITLE="Open Architecture Multimedia Documents",
BOOKTITLE="Proceedings ACM Multimedia '93",
ADDRESS="Anaheim, CA",
MONTH=aug,
YEAR=1993,
ENTRYBY=Ao
}

@ARTICLE{Hill9301:R,
AUTHOR="Hill, R. D. and Brink, T. and Patterson, J. F. and Rohall, S. L.
and Winter, W. T.",
TITLE="The Rendezvous Language and Architecture",
JOURNAL=cacm,
VOLUME=36,
NUMBER=1,
PAGES="62-67",
MONTH=jan,
YEAR=1993,
ABSTRACT="The {\em Rendezvous} system is a software toolkit and
architecture for building multiuser applications for use in
conversational settings. Multiuser applications are viewed as providing
distributed {\em conversational props}. A converasational prop is an
artifact used by the participants in the conversation to facilitate or
augment the conversation. The toolkit provides an API for building
multiuser applications, based on the X Window System. A widget set
designed especially for multiuser applications is provided.",
ENTRYBY=Ao
}


@TECHREPORT{Tind93:Scheduling,
AUTHOR="Tindell, K. and Burns, A.",
TITLE="Scheduling hard real-time multi-media disk traffic",
INSTITUTION="University of York",
ADDRESS="York, Great Britain",
TYPE="Technical Report",
YEAR=1993,
NUMBER="YCS 93-204",
ABSTRACT="In this report we show how existing real-time scheduling
theory, developed to analyse the scheduling of processors and network
communication, can also be applied to the problem of guaranteeing the
performance of multi-media information streams read from a disk drive.
We develop simple analysis that is independent of disk layout
information; a small example is analysed, using a simple tool
embodying the analysis. The analysis is then extended to take account
of disk layout information; the example is then re-analyzed using this
less pessimistic model. One of the advantages of the proposed approach
is that it gives a precise means of determining the impact of buffer
size on disk utilisation.",
KEYWORDS="real-time systems; disk scheduling; multimedia operating
systems; performance evaluation",
REFERENCES=13,
URL="ftp://ftp.cs.york.ac.uk/reports/YCS-93-204.ps.Z",
ENTRYBY=Sc
}

%%
%% University of California, Berkeley
%%    Plateau Multimedia Project
%%
%%   ftp site for Plateau Multimedia Project:
%%     s2k-ftp.cs.berkeley.edu:/pub/multimedia/papers
%%

%% Carnegie-Mellon University
%%
%% ftp site:
%%   reports.adm.cs.cmu.edu

@INPROCEEDINGS{Kita9312:Integrated,
AUTHOR="Kitayama, T. and Nakajima, T. and Arakawa, H. and Tokuda, H.",
TITLE="Integrated Management of Priority Inversion in Real-Time
{M}ach",
BOOKTITLE="IEEE Real-Time Systems Symposium",
MONTH=dec,
YEAR=1993,
URL="ftp://reports.adm.cs.cmu.edu/",
ENTRYBY=Ao
}

@TECHREPORT{Merc9305:Processor,
AUTHOR="Mercer, C. W. and Savage, S. and Tokuda, H.",
TITLE="Processor Capacity Reserves for Multimedia Operating
Systems",
INSTITUTION="School of Computer Science, Carnegie Mellon University",
NUMBER="CMU-CS-93-157",
MONTH=may,
YEAR=1993,
ENTRYBY=Ao
}

%%%%%% End of CMU
%% 
%% IBM's European Networking Center, Germany
%%
%% ftp site:
%%   ibminet.awdpa.ibm.com:/pub/heipapers
%%

@INPROCEEDINGS{Hehm9111:Implementing,
AUTHOR={Hehmann, D. and Herrtwich, R. G. and Schultz, W. and
Schütt, T. and Steinmetz, R.},
TITLE="Implementing {HeiTS}: Architecture and Implementation Strategy
of the {Heidelberg} High-Speed Transport System",
BOOKTITLE="Proc. Second International Workshop on Network and
Operating System Support for Digital Audio and Video",
ADDRESS="Heidelberg, Germany",
PAGES="33--44",
MONTH=nov,
YEAR=1991,
URL="ftp://ibminet.awdpa.ibm.com/pub/heipapers/",
ENTRYBY=Ao
}

@INPROCEEDINGS{Herr9209:System,
AUTHOR="Herrtwich, R. G. and Wolf, L.",
TITLE="A System Software Structure for Distributed Multimedia
Systems",
BOOKTITLE="5th ACM SIGOPS European Workshop",
ADDRESS="Le Mont Saint-Michel, France",
MONTH=sep,
YEAR=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",
URL="ftp://ibminet.awdpa.ibm.com/pub/heipapers/stmichel.ps.Z",
ENTRYBY=Sc
}

%%%%%% End of IBM European Networking Center
%%
%% Interactive Multimedia Association (IMA)
%%
%% ftp site:
%%   ibminet.awdpa.ibm.com:/pub/ima/{response, rft}
%%

@TECHREPORT{IMA93:Multimedia,
AUTHOR="{IMA Compatibility Project}",
TITLE="Multimedia System Services Specification 1.0",
INSTITUTION="Interactive Multimedia Association",
YEAR=1993,
URL="ftp://ibminet.awdpa.ibm.com:/pub/ima/",
ENTRYBY=Ao
}

@TECHREPORT{IMA92:Request,
AUTHOR="{IMA Compatibility Project}",
TITLE="Request for Technology: Multimedia System Services",
INSTITUTION="Interactive Multimedia Association",
YEAR=1992,
URL="ftp://ibminet.awdpa.ibm.com:/pub/ima/",
ENTRYBY=Ao
}