NOSSDAV'95 Paper Abstracts
Kurt Rothermel and Tobias Helbig,
An Adaptive Stream Synchronization Protocol,"
Proc. 5th Intl. Workshop on Network and Operating System Support
for Digital Audio and Video, Durham, New Hampshire, April 18-21,
1995, pp. 189-202.
Abstract:
Protocols for synchronizing data streams should be highly adaptive
with regard to both changing network conditions as well as to
individual user needs. The Adaptive Synchronization Protocol we are
going to describe in this paper supports any type of distribution
of the stream group to be synchronized. It incorporates buffer level
control mechanisms allowing an immediate reaction on overflow or
underflow situations. Moreover, the proposed mechanism is flexible
enough to support a variety of synchronization policies and allows
to switch them dynamically during presentation. Since control messages
are only exchanged when the network conditions actually change,
the message overhead of the protocol is very low.
Keywords:
Communication protocols, stream synchronization, quality of service,
buffer control algorithm.
S.W. Lau and John C.S. Lui,
"A Novel Video-On-Demand Storage Architecture for Supporting Constant
"Frame Rate with Variable Bit Rate Retrieval,"
Proc. 5th Intl. Workshop on Network and Operating System Support
for Digital Audio and Video, Durham, New Hampshire, April 18-21,
1995, 315-326.
Abstract: One of the quality of service (QOS) factors in
video-on-demand (VOD) applications is to provide high resolution
quality to end users. One way to achieve this is to provide a
constant display frame rate (e.g., 30 frames/sec) at the display
station. However, due to the nature of video files and compression
technique applied, video frame sizes vary significantly from frame to
frame. Therefore, although the display frame rate is fixed, data
retrieval is a variable bit rate process. Conventional VOD storage
servers assume a peak rate retrieval of video files. Therefore, the
number of concurrent requests to the VOD server cannot be maximized.
In this paper, we consider a VOD storage the number of concurrent
requests to the VOD server cannot be maximized. In this paper, we
consider a VOD storage server which can support a fixed frame rate at
the display and at the same time, variable bit rates retrieval of
compressed video files. We describe (1) video files layout strategy,
(2) request scheduling algorithm, (3) buffering issues and, (4)
various VCR features support such that the number
of concurrent requests can be maximized.
Keywords:
Multimedia storage architecture, scheduling, variable
bit rate retrieval.
G. Neufeld, D. Makaroff and N. Hutchinson,
"The Design of a Variable Bit Rate Continuous Media Server,"
Proc. 5th Intl. Workshop on Network and Operating System Support
for Digital Audio and Video, Durham, New Hampshire, April 18-21,
1995, pp. 375-378.
Abstract:
This paper describes the design and implementation of a file server
for variable bit rate continuous media. Most continuous media file servers
have been designed for constant bit rate streams. We address the problem
of building a server where each stream may have a different bit rate
and, more importantly, where the bit rate within a single stream may
vary considerably. Such servers will be come increasingly more important
because of Variable Bit Rate (VBR) compression standards such as MPEG-2.
Keywords:
S. Lam and G. Xie,
"Burst Scheduling Networks: Flow Specification and Performance Guarantees,"
Proc. 5th Intl. Workshop on Network and Operating System Support
for Digital Audio and Video, Durham, New Hampshire, April 18-21,
1995, pp. 303-306.
Abstract:
We present a class of packet switching networks, called Burst Scheduling
Networks, designed to provide throughput, delay, and delay jitter guarantees.
These performance guarantees are derived from the delay guarantee of a VC
server, and a new traffic model called Flow Specification.
The delay guarantee of a VC server has several desirable properties, including
the following firewall property: The guarantee to a flow is unaffected by
the behavior of other flows sharing the same server. There is no assumption
that sources are flow-controlled or well-behaved.
Each guaranteed flow is modeled as a sequence of bursts, each of which is a
sequence of packets. Bursts are needed to specify two types of jitter bounds:
over the delays of packets in a burst, and over the delays of bursts in a flow.
For video flows, each encoded picture is naturally modeled by a burst. The
model is also appropriate for audio and data flows that require delay and
delay jitter guarantees.
With the new traffic model, a flow can be partitioned into intervals (bursts)
that have substantially different average rates; the first packet of a burst
carries information on the size and average rate of the burst. Switches are
designed to process flows efficiently in bursts.
Keywords: Packet switching, delay guarantee, delay jitter guarantee,
throughput guarantee, firewall property, virtual clock, Burst
Scheduling, video.
K. Almeroth and M. Ammar,
"The Role of Multicast Communication in the Provision of Scalable and
Interactive Video-On-Demand Service,"
Proc. 5th Intl. Workshop on Network and Operating System Support
for Digital Audio and Video, Durham, New Hampshire, April 18-21,
1995, pp. 267-270.
Abstract:
Multicast delivery can improve the scalability of Video-On-Demand
(VOD) systems by allowing multiple customers to share one set of video server
and network resources. In such a system, providing customer interactivity
can be difficult. In this paper we overview our work which addresses how
interactivity may be accomplished in a multicast delivery environment and
analyzes the performance of such systems.
Keywords: Video-On-Demand, video, multicast, interactive, scalable
E. Biersack and J. Nonnenmacher,
"WAVE: A New Multicast Routing Algorithm for Static and Dynamic
Multicast Groups,"
Proc. 5th Intl. Workshop on Network and Operating System Support
for Digital Audio and Video, Durham, New Hampshire, April 18-21,
1995, pp. 243-254.
Abstract:
We present a new multicast algorithm called WAVE for establishing
source-specific multicast trees. WAVE meets multiple quality of
service requirements (constraints) such as delay, cost, and available
bandwidth, simultaneously. Simulation results show that WAVE performs
very good in terms of delay and cost for both, static and dynamic
multi- cast groups, when compared to the best multicast algorithms
known.
Keywords:
L. Wolf, L. Delgrossi, R. Steinmetz, S. Schaller and H. Wittig,
"Issues of Reserving Resources in Advance,"
Proc. 5th Intl. Workshop on Network and Operating System Support
for Digital Audio and Video, Durham, New Hampshire, April 18-21,
1995, pp. 27-37.
Abstract:
Resource management offers Quality-of-Service reliability for
time-critical continuous-media applications. Currently existing resource
management systems provide only means to reserve resources starting with
the reservation attempt and lasting for an unspecified duration. However,
for several applications such as video conferencing the ability to reserve
the required resources in advance is needed. This paper explains a model
for resource reservation in advance. We identify and discuss issues which
must be resolved in resource reservation in advance systems. Some of the
possible scenarios to be considered are described and we show how the
resource reservation in advance scheme can be embedded in a general
architecture.
Keywords:
A. Mauthe and G. Coulson,
"Scheduling and Admission Testing for Jitter Constrained Periodic Threads,"
Proc. 5th Intl. Workshop on Network and Operating System Support
for Digital Audio and Video, Durham, New Hampshire, April 18-21,
1995, pp. 219-222.
Abstract:
We present new admission tests for periodic real-time threads with
explicitly stated deadlines. In standard earliest-deadline-first
scheduling, the deadline of a periodic thread is conventionally the
end of the current period. In contrast, our tests supports periodic
threads in which the deadline may be earlier than the end of the
current period. In the extreme case, the deadline may be specified as
identical to the execution time, which results in perfectly
isochronous periodic threads. The provision of such threads, which
we refer to as jitter constrained threads, helps end-systems to honour
jitter- as well as throughput-related QoS parameters in distributed
multimedia systems. In addition, such threads can reduce end-to-end
delay and buffer memory requirements as less buffering is needed to
smooth excessive delay jitter. Our admission tests are primarily
designed around the earliest deadline first (EDF) scheduling
algorithm but the tests can also be applied to priority based
scheduling schemes, such as rate monotonic, with minimal
modification.
Keywords: Admission Testing, CPU-Scheduling
K. Fall, J. Pasquale and S. McCanne,
"Workstation Video Playback Performance with Competitive Process Load,"
Proc. 5th Intl. Workshop on Network and Operating System Support
for Digital Audio and Video, Durham, New Hampshire, April 18-21,
1995, pp. 179-182.
Abstract:
While many researchers believe that multimedia applications are best
managed with hard, real-time scheduling mechanisms, models based on
application-level adaptation with relaxed scheduling constraints are
gaining acceptance. We analyze an existing video conferencing
application, which was designed without explicit support for CPU
resource management, and propose modifications to its architecture to
support CPU load adaptation. Inter-frame display times (IDT) for a 30
frames/sec video segment are analyzed for increasing multiprocessing
loads. As expected, the IDT variance (and mean) increases markedly
with load, especially beyond saturation. We show that this display
jitter can be significantly reduced by gracefully adapting the
application's load requirements to match the available CPU resources.
Keywords:
V. Bansal, R.J. Siracusa, J.P. Hearn, G. Ramamurthy and D. Raycaudhuri,
"Adaptive QoS-Based API for ATM Networking,"
Proc. 5th Intl. Workshop on Network and Operating System Support
for Digital Audio and Video, Durham, New Hampshire, April 18-21,
1995, pp. 299-302.
Abstract:
This paper describes work-in-progress on a new adaptive QoS-based
application programming interface (API) for ATM networking. This
ATM API, referred to as the "ATM Service Manager (ASM)" is motivated
by the observation that current transport interfaces do not provide
QoS features necessary for multimedia applications to achieve desired
performance/cost objectives on multiservice ATM networks. In general,
an application connected to ATM has a choice of transport protocols
(e.g. TCP, UDP, alternative multimedia stream protocols), ATM service
class (e.g. ABR, VBR, CBR) and QoS parameters (e.g. peak/average
bandwidth, delay jitter, cell loss) for each media type. It is our
view that application software should be shielded from the complexity
of QoS-based service management by an adaptive API which is responsible
for mapping the application requirements to ATM (and vice versa).
Keywords: ATM networks, QoS, API, Multimedia network interfaces
M. Grossglauser, S. Keshav and D. Tse,
"The Case Against Variable Bit Rate Service,"
Proc. 5th Intl. Workshop on Network and Operating System Support
for Digital Audio and Video, Durham, New Hampshire, April 18-21,
1995, pp. 307-310.
Abstract:
Keywords:
![](line1.gif)
K. Kawachiya, M. Ogata, N. Nishio and Hideyuki Tokuda,
"Evaluation of QOS-Control Servers on Real-Time Mach,"
Proc. 5th Intl. Workshop on Network and Operating System Support
for Digital Audio and Video, Durham, New Hampshire, April 18-21,
1995, pp. 123-126.
Abstract:
The Keio-MMP (MultiMedia Platform) project has as its aim the
development of a common software platform for distributed multimedia
computing. Its base software is Real-Time Mach 3.0 (RT-Mach). We are
extending RT-Mach and constructing servers for multimedia processing
on our extension. One goal of our project is to incorporate the
concept of Quality of Service (QOS), and to manage system resources
over multiple continuous-media sessions on the basis of this concept.
However, it is difficult for multimedia applications to manage their
QOS appropriately by themselves without system support. There should
be some framework for QOS management at the system-software level. To
provide such support, we adopted a QOS Manager-based scheme, and
developed an experimental QOS-Control Server on RT-Mach. This paper
gives the results of several experiments with the QOS-Control Server
and describes the extensions of RT-Mach needed for efficient QOS
control. The thread-model extension makes it possible to create a
thread suitable for continuous-media processing, and the
CPU-reservation mechanism gives more efficient QOS control. Based on
these experiments, we are now designing a ``QOS-Ticket'' model as a
new QOS-control scheme.
Keywords: QOS control, resource management, operating system,
multimedia, continuous media
H. Kitamura, K. Taniguchi, H. Sakamoto and T. Nishidam,
"A New OS Architecture for High Performance Communication
over ATM Networks -- Zero-copy architecture --
,"
Proc. 5th Intl. Workshop on Network and Operating System Support
for Digital Audio and Video, Durham, New Hampshire, April 18-21,
1995. pp. 87-90.
Abstract:
A new OS architecture, referred to as Zero-copy architecture,
for high performance communication is proposed. This architecture
dissolved memory copy bottleneck that is a major overhead in
protocol processing. This can reduce CPU processing overhead, in
addition to realizing high speed data communication. This
architecture is shown to be suitable for large volume of data
communication like video and image transfer.
Keywords: Communication Protocol, Performance Evaluation, ATM,
Multimedia Communication, Workstation, UNIX
C. Lindblad,
"VuSystem Performance Measurements,"
Proc. 5th Intl. Workshop on Network and Operating System Support
for Digital Audio and Video, Durham, New Hampshire, April 18-21,
1995, pp. 79-82.
Abstract:
In this paper I discuss some performance measurements made of the
VuSystem, a programming system for the software-based processing of
audio and video data. The VuSystem is designed to run on ordinary
Unix workstations with no specific support for the manipulation of
multimedia data. Measurements made of processing times of
representative filter modules demonstrate the viability of the
approach.
Keywords:
S. Cen, C. Pu, R. Staehli, C. Cowan and J. Walpole,
" A Distributed Real-Time MPEG Video Audio Player,"
Proc. 5th Intl. Workshop on Network and Operating System Support
for Digital Audio and Video, Durham, New Hampshire, April 18-21,
1995, pp. 151-162.
Abstract:
This paper presents the design, implementation and experimental
analysis of a distributed, real-time MPEG video and audio player. The
player is designed for use across the Internet, a shared environment
with variable traffic and with great diversity in network bandwidth and
host processing speed. We use a novel toolkit approach to build
software feedback mechanisms for client/server synchronization, dynamic
Quality-of-Service control, and system adaptiveness. Our experimental
results show that the feedback mechanisms are effective, and that the
player performs very well in the Internet environment.
Keywords: multimedia, real-time system, software feedback, synchronization, Quality-of-Service control, adaptiveness
M. Degermark, T. Kohler, S. Pink and O. Schelen,
"Advance Reservations for Predictive Service,"
Proc. 5th Intl. Workshop on Network and Operating System Support
for Digital Audio and Video, Durham, New Hampshire, April 18-21,
1995, pp. 3-14.
Abstract:
We extend a measurement-based admission control algorithm
suggested for predictive service to provide advance
reservations for guaranteed and predictive service
while keeping the attractive features of predictive service.
The admission decision for advance reservations is based on
information about flows that overlap in time.
For flows that have not yet started, the
requested values are used, and for those that have already started
measurements are used. This allows us to estimate the network load
accurately for the near future. To provide advance reservations we
ask users to include durations in their requests.
We provide simulation results to show that predictive service with
advance reservations provides utilization levels significantly higher
than those for guaranteed service.
Keywords: Predictive Service, Admission Control, Advance Reservations,
Quality-of-Service, Networking
M. Jones, P. Leach, R. Draves and J. Barrera, III,
"Support for User-Centric Modular Real-Time
Resource Management in the Rialto Operating System,"
Proc. 5th Intl. Workshop on Network and Operating System Support
for Digital Audio and Video, Durham, New Hampshire, April 18-21,
1995, pp. 55-66.
Abstract:
This paper describes ongoing investigations into algorithms for user-centric
modular distributed real-time resource management. These investigations are
being conducted in the context of the Rialto operating system -- an
object-based real-time kernel and programming environment currently being
developed within Microsoft Research.
A primary goal of this research is to develop appropriate real-time
programming abstractions to allow multiple independent real-time programs to
dynamically coexist and share resources on the same hardware platforms. Use
of these abstractions is intended both to allow individual applications to
reason about their own resource requirements and for per-machine system
resource planner applications to reason about and control resource allocations
between potentially competing applications. The set of resources being
managed is dynamically extensible, and may include remote resources in
distributed environments. The local planner conducts resource negotiations
with individual applications on behalf of the user, with the goal of
maximizing the user's perceived utility of the set of running applications
with respect to resource allocations for those applications.
Keywords:
H. Zhang and E. Knightly,
" A New Approach to Support Delay-Sensitive VBR Video in Packet-Switched Networks,"
Proc. 5th Intl. Workshop on Network and Operating System Support
for Digital Audio and Video, Durham, New Hampshire, April 18-21,
1995, pp. 275-286.
Abstract:
Previous approaches to supporting video on packet-switched networks
include deterministic service, statistical service, predicted service,
and feedback-based schemes. These schemes represent different tradeoffs
in quality of service (QOS), achievable network utilization, and method
of dealing with overload. In this paper, we propose a new service that
attempts to strike an efficient balance with the above tradeoffs. The
approach is based on deterministic guarantees with client controlled
renegotiation of QOS parameters and graceful adaptation during overload
periods. We evaluate the scheme using two traces of MPEG-compressed video
and show that, even with simple renegotiation polices and relatively low
renegotiation frequencies, high network utilization in the range of 50%
to 80% can be achieved. For traffic that is bursty over long intervals,
this represents a 100% to 150% improvement in network utilization compared
to deterministic service. Compared to statistical and predicted service, our
approach allows more graceful and client-controlled QOS degradation during
overload period.
Keywords: VBR video, guaranteed QOS, resource management, multimedia
communication
R. Bettati, D. Ferrari, A. Gupta, W. Heffner, W. Howe, M. Moran, Q. Nguyen and R. Yavatkar,
" Connection Establishment for Multi-Party Real-Time Communication,"
Proc. 5th Intl. Workshop on Network and Operating System Support
for Digital Audio and Video, Durham, New Hampshire, April 18-21,
1995, pp. 255-266.
Abstract:
There is considerable interest in the network community
in supporting real-time multi-party applications, such
as video conferencing. The Tenet Group at UC Berkeley
and ICSI has designed and implemented protocols that
provide quality of service (QoS) guarantees for real-
time traffic on packet switching networks. Suite 2 of
the Tenet protocols provides scalable, flexible and
efficient network support for real-time multi-party
connections. We outline our method of connection
establishment and describe the design issues and alter-
natives, and our decisions. Preliminary measurements
confirm the viability of our approach for real-time
multicast connection establishment.
Keywords: Multicast connection establishment; real-time communication;
multimedia networking
J. Bolot, H. Crepin and A. Garcia,
" Analysis of Audio Packet Loss in the Internet,"
Proc. 5th Intl. Workshop on Network and Operating System Support
for Digital Audio and Video, Durham, New Hampshire, April 18-21,
1995, pp. 163-174.
Abstract:
We consider the problem of distributing audio data over networks
such as the Internet that do not provide support for real-time
applications. Experiments with such networks indicate that audio
quality is mediocre in large part because of excessive audio
packet losses. In this paper, we show using measurements over
the Internet as well as analytic modeling that the number of
consecutively lost audio packets is small unless the network
load is very high. This indicates that open loop error control
mechanisms based on forward error correction would be adequate
to reconstruct most lost audio packets.
Keywords: Packet audio, loss process, Internet, forward error correction
N. Chaddha, G. Wall and B. Schmidt,
"An End to End Software Only Scalable Video Delivery System,"
Proc. 5th Intl. Workshop on Network and Operating System Support
for Digital Audio and Video, Durham, New Hampshire, April 18-21,
1995, pp. 139-150.
Abstract:
Precompressed video delivery systems commonly operate at fixed data rates.
However, variations in the availability of network bandwidth and processor
cycles are common in dynamic general purpose computing environments.
Variability arises from the outright lack of resources (e.g. network bandwidth
and cpu cycles), contention for available resources due to congestion, or a
user's unwillingness to allocate needed resources to the task. Users of a
scalable video delivery system have greater flexibility and therefore, the
system can more effectively deliver video in the presence of system resource
scarcity. This paper describes an end-to-end system combining a new scalable
video compression algorithm, video delivery software, a software video decoder,
and a market-based mechanism for the resolution of conflicts in providing
video to the user.
Keywords: Scalable Video, Software Only, End-to-End Video Delivery System
C. Cranor and G. Parulkar,
"Design of Universal Continuous Media I/O[*],"
Proc. 5th Intl. Workshop on Network and Operating System Support
for Digital Audio and Video, Durham, New Hampshire, April 18-21,
1995, pp. 83-86.
Abstract:
The problem this research addresses is how to modify an existing
operating system's I/O subsystem to support new high-speed networks
and high-bandwidth multimedia applications that will play an important
role in future computing environments. The proposed I/O subsystem is
called universal continuous media I/O (UCM I/O).
Keywords: I/O, networking, operating systems, multimedia
D. Ferrari, A. Gupta and G. Ventre,
"Distributed advance reservation of real-time connections,"
Proc. 5th Intl. Workshop on Network and Operating System Support
for Digital Audio and Video, Durham, New Hampshire, April 18-21,
1995, pp. 15-26.
Abstract:
The ability to reserve real-time connections in advance is essential in
all distributed multi-party applications (i.e., applications involving
multiple human beings) using a network that
controls admissions to provide good quality of service. This paper
discusses the requirements of the clients of an advance reservation
service, and a distributed design for such a service. The design
is described within the context of the Tenet Real-Time Protocol
Suite 2, a suite being developed for multi-party communication,
which will offer advance reservation capabilities to its clients
based on the principles and the mechanisms proposed in the paper.
Some simulation results about the performance
of these mechanisms are also presented.
Keywords: Resource management/reservation, admission control, communication
protocols, advance reservations
P. Goyal, S. Lam and H. Vin,
"Determining End-to-End Delay Bounds In Heterogeneous Networks,"
Proc. 5th Intl. Workshop on Network and Operating System Support
for Digital Audio and Video, Durham, New Hampshire, April 18-21,
1995, pp. 287-298.
Abstract:
We define a class of Guaranteed Rate (GR)
scheduling algorithms. The GR class includes Virtual Clock,
Packet-by-Packet Generalized Processor Sharing and Self Clocked Fair
Queuing. For networks that employ scheduling algorithms belonging to
GR, we present a method for determining an upper bound on end-to-end
delay. The method facilitates determination of end-to-end delay
bounds for a variety of sources. We illustrate the method by
determining end-to-end delay bounds for sources conforming to Leaky
Bucket and Exponentially Bounded Burstiness.
Keywords:
K. Jeffay and D. Bennett,
" A Rate-Based Execution Abstraction For Multimedia Computing,"
Proc. 5th Intl. Workshop on Network and Operating System Support
for Digital Audio and Video, Durham, New Hampshire, April 18-21,
1995, pp. 67-78.
Abstract:
Process models for multimedia computing must allow applications
to adapt their pattern of execution as resources become scarce or abundant.
As processes adapt, they should be able to express their desired performance
in terms of application-defined data units or events. We propose a process
model wherein processes execute according to a general rate specification of
x process executions every y time units. In addition, a separate parameter
is used to specify the desired response time for the completion of each
execution. In all cases the real-time performance of a rate-based process
is predictable. The model is general enough to encompass or extend many of
the existing models proposed for multimedia systems. Our model of rate-based
execution is described along with an implementation that detects when processes
should adapt their execution rate and minimizes latency in interprocess
communication.
Keywords:
H. Lauer, C. Shen, R. Osborne, J. Howard and Q. Zheng, M, Takegaki, H. Shimakawa and I. Mizunuma,
"Digital Audio and Video in Industrial Systems,"
Proc. 5th Intl. Workshop on Network and Operating System Support
for Digital Audio and Video, Durham, New Hampshire, April 18-21,
1995, pp. 175-178.
Abstract:
This is a position paper discussing the requirements of networks in industrial
environments, especially with respect to digital audio and video. Topics
include resource allocation, issues surrounding switching, scheduling, and
priorities, end system interface requirements, and traffic characteristics
and their implications on flow control.
Keywords:
P. Leydekkers, V. Gay and L. Franken,
"A Computational and Engineering View on Open Distributed Real-time MultiMedia Exchange,"
Proc. 5th Intl. Workshop on Network and Operating System Support
for Digital Audio and Video, Durham, New Hampshire, April 18-21,
1995, pp. 43-54.
Abstract:
An important requirement for distributed multimedia applications is
the support of real-time communication and the means to specify real-time
aspects. The aim of this paper is to extend RM-ODP and TINA-C computational
and engineering views on distributed systems for the specification and
support of real-time communication. It is expected that these bodies have a
major impact in the area of distributed processing. However, concepts and
mechanisms to support real-time communication are not yet fully included or
detailed in these standards. In particular this paper addresses Quality of
Service (QoS) specifications for continuous dataflows. These QoS
specifications are described from the ODP computational and engineering
viewpoint and the repercussions of these QoS specifications for functions
located in both the computing and telecommunications environment are
discussed.
Keywords: Distributed Multimedia Architectures, QoS, ODP, TINA-DPE
J. Nieh and M. Lam,
"Integrated Processor Scheduling for Multimedia,"
Proc. 5th Intl. Workshop on Network and Operating System Support
for Digital Audio and Video, Durham, New Hampshire, April 18-21,
1995, 215-218.
Abstract:
The advent of multimedia ushers forth a growing class of applications
that must manipulate digital audio and video within well-defined
timeliness requirements. Existing processor schedulers are inadequate
in supporting these requirements. They fail to allow the integration
of these continuous media computations with conventional interactive
and batch activities. By observing that all tasks do in fact have
timeliness requirements, we have created a new scheduler that provides
integrated processor allocation for all classes of computational
activities. By exploiting information on the relative importance of
different activities to the user, the scheduler seeks to maximize the
value the system delivers to the user, and gracefully degrades
activities in order of importance when all timeliness requirements can
not be met. The solution is unique in the degree to which it allows
users control over the allocation of processing resources, and in that
it provides an integrated scheduling solution in supporting multimedia
applications. We have implemented the scheduler in the Solaris
operating system to demonstrate its effectiveness for real
applications on real systems.
Keywords: processor scheduling, operating systems, multimedia, time constraints, overload management
L. Schreier and M. Davis,
" System-Level Resource Management for Network-Based Multimedia
Applications,"
Proc. 5th Intl. Workshop on Network and Operating System Support
for Digital Audio and Video, Durham, New Hampshire, April 18-21,
1995, 127-130.
Abstract:
SRI International (SRI) has developed a model for system-level resource
management in distributed systems, and applied this model in the context
of a multimedia conferencing application. The model considers user
objectives, resource constraints, and adaptable execution techniques.
User objectives are specified by means of benefit functions. We have
implemented a prototype of a distributed multimedia display application
that demonstrates key aspects of the model, including adaptation to
a changing execution environment.
Keywords: resource management, adaptable systems, quality of service,
benefit functions
H. Schulzrinne,
"When can we unplug the radio and telephone?"
Proc. 5th Intl. Workshop on Network and Operating System Support
for Digital Audio and Video, Durham, New Hampshire, April 18-21,
1995, pp. 183-184.
Abstract:
Despite high-performance workstations, special-purpose communication
devices like telephones and radio still dominate. Reasons include
inadequate system and network support, both in terms of performance
and the ability to write applications without specialized signal
processing knowledge. The interaction of applications with the network
will depend largely on the future tariff structures rather than
protocol issues.
Keywords: audio, video, research issues, ATM, multimedia, programming
abstractions
H. Schulzrinne,
"Dynamic Configuration of Conferencing Applications using Pattern-Matching Multicast,"
Proc. 5th Intl. Workshop on Network and Operating System Support
for Digital Audio and Video, Durham, New Hampshire, April 18-21,
1995, pp. 231-242.
Abstract:
Multimedia conferencing systems are usually large, complex software
systems. We describe a local control architecture and communication
protocols that allow to tie together media agents, controllers and
auxiliary applications such as media recorders and management proxies
into a single conference application. Unlike other systems, control
of a single conference can be shared between several controllers.
Each media can be handled by one or more independent media agents.
Parts of the system have been implemented using an IP-multicast-based
audio conferencing tool NeVoT. The communicating applications
disseminate state and control information through a replicator. The
replicator mainly limits distribution of messages based on expressed
interest of other applications, thus implementing an
application-level, receiver-driven local multicast. It also
automatically starts applications as needed.
Keywords: conference control, multicast, RPC, CSCW
S. Paek, P. Bocheck and S. Chang,
"Scalable MPEG2 Video Servers with Heterogeneous QoS on Parallel Disk Arrays,"
Proc. 5th Intl. Workshop on Network and Operating System Support
for Digital Audio and Video, Durham, New Hampshire, April 18-21,
1995, 363-374.
Abstract:
In this paper we focus on the video storage unit of a video server. We
present a new, flexible data placement strategy for independent parallel disk
arrays. The trade-off between utilization efficiency and interactive delay is
investigated for this data placement strategy. Based on this trade-off, we
show the advantage of video servers supporting a range of interactivity QoS.
For our data placement strategy, we show that using scalable video
improves the utilization and interactivity performance of a video server. We
use three-layer, scalable MPEG2 digital video to support resolution QoS at a
video server. Finally, we show that the data placement strategy reduces the
complexity of admission control at the video server to that of a single disk
system.
Keywords:
S. Stoller and J. DeTreville,
"Storage Replication and Layout in Video-on-Demand Servers,"
Proc. 5th Intl. Workshop on Network and Operating System Support
for Digital Audio and Video, Durham, New Hampshire, April 18-21,
1995, 351-363.
Abstract:
We propose and analyze an architecture for storage servers in large
Video on Demand (VoD) systems. We describe a method for distributing
the collection of titles among the levels of the storage hierarchy,
based on estimates of the mean demand for each title. The resulting
distribution minimizes cost for a given level of performance. Since
high availability is desirable in VoD systems, we consider the use of
mirroring or parity-based redundancy (a la RAID) and estimate the effect
on the system's cost and availability. In the very-large-scale storage
systems needed for VoD, the placement of disk arrays on the pool of
computers must be chosen carefully to provide high availability for the
least cost. We propose a strategy for arranging disk arrays on a pool of
computers; our strategy is inspired by Holland and Gibson's work on
parity declustering for RAID. To estimate the availability benefit of
redundancy, we developed an analytical availability model that takes
into account the system's tolerance to all kinds of failures (e.g.,
software bugs as well as disk crashes).
Keywords: video-on-demand, RAID, availability, layout, storage
A. Chaney, I. Wilson and A. Hopper,
"The Design and Implementation of a
RAID-3 Multimedia File Server,"
Proc. 5th Intl. Workshop on Network and Operating System Support
for Digital Audio and Video, Durham, New Hampshire, April 18-21,
1995, pp. 327-338.
Abstract:
The Olivetti Research Laboratory has developed an experimental system based on
intelligent peripherals connected directly to an ATM network. As well as
multimedia modules (e.g. audio and video) the system also includes a directly
connected RAID-3 storage server called the "Disc Brick". This paper describes
the architecture of the Disc Brick, and discusses some of the hardware and
software issues raised by its design. It also presents measurements taken
from a Disc Brick in operation, and discusses how the observations relate to
the original design objectives. Finally, the paper attempts to evaluate the
Disc Brick as part of ORL's family of directly connected peripherals.
Keywords:
R. Yavatkar and K. Lakshman,
"A CPU Scheduling Algorithm for Continuous Media Applications,"
Proc. 5th Intl. Workshop on Network and Operating System Support
for Digital Audio and Video, Durham, New Hampshire, April 18-21,
1995. pp. 223-226.
Abstract:
We address the problem of designing a CPU scheduling algorithm that
can guarantee a steady, average rate of execution to multimedia (MM)
applications. In particular, our goal is to provide dynamic QOS
control in the presence of varying compute times, arrival and
departure of processes, and CPU overloads. We have designed and
evaluated a new CPU management algorithm called Rate-based Adjustable
Priority Scheduling (RAP). that makes the following new
contributions. First, we do not assume a priori knowledge of compute
times for the MM applications. RAP includes an admission control
algorithm that estimates and allocates available capacity to new
processes. Second, we consider only average computing time needed by a
process and admission control algorithm provides guarantees on average
rate of execution. Third, our algorithm includes a rate adaptation
mechanism and an application-level PLL (Phase Locked Loop) to adapt to
CPU overload and variations in compute times.
Keywords:
M.M. Buddhikot and G. Parulkar,
" Efficient Data Layout, Scheduling and Playout Control in MARS,"
Proc. 5th Intl. Workshop on Network and Operating System Support
for Digital Audio and Video, Durham, New Hampshire, April 18-21,
1995, pp. 339-350.
Abstract:
Large scale on-demand multimedia servers, that can provide independent
and interactive access to a vast amount of multimedia information to a
large number of concurrent clients, will be required for a wide spread
deployment of exciting multimedia applications. Our project, called
Massively-parallel And Real-time Storage (MARS), is aimed at prototype
development of such a large scale server. This paper primarily focuses
on the distributed data layout and scheduling techniques developed in
this project. These techniques support a high degree of parallelism
and concurrency, and efficiently implement various playout control
operations, such as fast forward, rewind, pause, resume, frame advance
and random access.
Keywords: Large scale multimedia storage servers,
Interactive playout control,
Desk Area Networks, Storage systems
A. Campbell, D. Hutchinson and C. Aurrecoechea,
"Dynamic QoS Management for Scalable Video Flows,"
Proc. 5th Intl. Workshop on Network and Operating System Support
for Digital Audio and Video, Durham, New Hampshire, April 18-21,
1995, pp. 107-118.
Abstract:
We introduce the concept of Dynamic QoS Management (DQM) for
control and management of hierarchically coded flows operating
in heterogeneous multimedia networking environments. The
motivation that underpins our scheme is to bridge the
heterogeneity gap that exists between applications, end-
systems and networks. QoS adaptors, QoS filters and QoS
groups are key scalable objects used in resolving quality
of service capability mismatch. QoS filters manipulate
hierarchically coded flows as they progress through the communications
system, QoS adaptors scale flows at the end-systems based on the
flow's measured performance and user supplied QoS scaling policy,
and QoS groups provide baseline
quality of service for multicast communication.The focus of
the work is driven by a) the special features of scalable
video flows-in particular MPEG2, b) the needs of both scalable and
single-layer video for transmission over multimedia
networks such as ATM. A novel adaptive network service is
proposed for the transmission of multi-layer coded flows
that offers "hard" guarantees to the base layer, and "fairness"
guarantees to the enhancement layers based on a new
bandwidth allocation technique called Weighted Fair Sharing
(WFS).
Keywords: MPEG-2, Adaptive Service, Scalable Flows,
Dynamic QoS Management, End-to-end QoS
D. Venkatesh and T.D.C. Little,
"Dynamic Service Aggregation for Efficient
Use of Resources in Interactive Video Delivery,"
Proc. 5th Intl. Workshop on Network and Operating System Support
for Digital Audio and Video, Durham, New Hampshire, April 18-21,
1995, pp. 119-122.
Abstract:
To support future interactive information delivery services there is a
need to balance individual interactivity with the desire to maximize
the number of supported sessions. Currently, few techniques have
demonstrated the ability to renegotiate and scale service parameters
per session in progress as required to adapt to differing terminal
equipment characteristics and network congestion. This paper
addresses this problem through the definition of decomposable service
groups that permit aggregation of interactivity, terminal
characteristics, and levels of service scaling. The proposed approach
applies the characteristics of end applications and data storage
requirements to the design of a data scaling mechanism.
Keywords: interactive multimedia services, service scaling, multicast,
dynamic service aggregation, video-on-demand
Alexandros Eleftheriadis and Dimitris Anastassiou,
" Meeting Arbitrary QoS Constraints Using Dynamic Rate Shaping of Coded Digital Video,"
Proc. 5th Intl. Workshop on Network and Operating System Support
for Digital Audio and Video, Durham, New Hampshire, April 18-21,
1995, pp. 95-106.
Abstract:
We introduce the concept of {\em Dynamic Rate Shaping}, a technique to adapt
the rate of compressed video bitstreams (MPEG-1, MPEG-2, H.261, as well as
JPEG) to dynamically varying rate (and delay) constraints. The approach
provides an interface (or filter) between the encoder and the network, with
which the encoder's output can be perfectly matched to the network's quality
of service characteristics. Since the presented algorithms do not require
interaction with the encoder, they are fully applicable to precoded, stored
video (as in, for example, video-on-demand systems). By providing decoupling
of the encoder and the network, universal interoperability can be achieved. A
set of low-complexity algorithms for dynamic rate shaping is presented, and
both optimal and extremely fast designs are discussed. The latter are simple
enough to allow software-based implementation. Experimental results are
provided using actual MPEG-2 bitstreams.
Keywords:
B. Schmidt, J. Northcutt, and M. Lam
"A Method and Apparatus for Measuring Media Synchronization,"
Proc. 5th Intl. Workshop on Network and Operating System Support
for Digital Audio and Video, Durham, New Hampshire, April 18-21,
1995, pp. 203-214.
Abstract:
Media synchronization is widely regarded as a fundamental problem in the
field of multimedia. While much work has been conducted in this area, and
many different solutions have been proposed, no method for obtaining a
repeatable, objective measure of synchronization performance exists. Thus,
there has been no means for determining the effectiveness of potential media
synchronization solutions. In this paper we present an experimental
methodology for quantitatively measuring the performance of different media
synchronization schemes. We describe a complete (hardware and software)
test environment for measuring audio/video synchronization quality of
various media players, and we also present empirical performance
measurements of an example media player. The results show that external
observation is necessary for accurate assessments of synchronization
performance. This test and evaluation methodology is applicable to other
media delivery systems and can serve as the first step in isolating and
quantifying the effects of individual components of a media delivery system.
Keywords: Multimedia, Synchronization, Measurement, Clocks