No full-text available
To read the full-text of this research,
you can request a copy directly from the authors.
StarWorks-a video applications server
Abstract
The authors describe StarWorks, a video applications server
software designed to support a wide range of digital video applications.
The StarWorks system will allow existing network applications, such as
databases and groupware applications, to add video support. These
applications can take advantage of the video application services to
support the real-time demands of streaming data. Attention is given to
the StarWorks operating environment, client software, and server
... VOD server design issues have engendered many research activities during the past few years 1, 10,11,18,20,21,32,33,13,12,5,23,30,31,6,36,14,15]. Related design issues such as real-time support for delay-sensitive retrieval, disk layout strategies for continuous media, admission control for new requests, and disk scheduling schemes have been studied by the researchers. ...
... Therefore, in order to provide a large number of concurrent accesses, the large-scale VOD server should be equipped with a mass storage system with multiple disk systems. Some VOD server design based on the multiple-disk systems have been reported in 23,18,30,31,6,36,14,15] In these studies, multiple disks are usually connected to the same SCSI bus, or 2 SCSI buses in 18]. However, most of the studies are based on analytical models and only veri ed by simulations. ...
... Another alternative solution to design the mass storage system of a video server is to utilize a RAID (Redundant Arrays of Inexpensive Disks) architecture. There are only few reported paper in 31,30] that investigated this solution. However, this study only covered the basic stripping method which only represented one solution based on RAID architecture. ...
Advancements in storage technology along with the fast deployment of high-speed networks has allowed the storage, transmission and manipulation of multimedia information such as text, graphics, still images, video and audio to be feasible. Our study focused on, the performance of the mass storage system for a large-scale video-on-demand server. Different video file striping schemes, such as application level striping and device driver level striping, were examined in order to study scalability and performance issues. To study the impact of different concurrent access patterns on the performance of a server, experimental results were obtained on group access on a single video file and multiple group accesses on multiple video files
... Starlight's StarWork TM 34,35] | StarWork TM is a digital video networking software capable of supporting a wide range of video applications and tens to hundreds of concurrent users. It uses software striping in which no special RAID hardware is required. ...
... Video servers | The video servers store a number of movies (characterized by their duration, popularity and streaming data rate) accessible by the users. Each server has nite storage and streaming capacities 35,34,32,33,29,30,31,26,15,37,38,50,49]. Such resources are considered to be always available and in a sense already paid for. ...
Video-on-demand (VOD) refers to video services in which users can request any video program from a server at any time. VOD has important applications in entertainment, education, information, and adverstising, such as movie-on-demand, distance learning, home shopping, interactive news, etc. In order to provide VOD services accommodating a large number of video titles and concurrent users, a VOD system has to be scalable --- scalable in storage and scalable in streaming capacity. Our goal is to design such a system with low cost, low complexity, and offering high level of service quality (in terms of, for example, user delay experienced or user loss rate). Storage scalability is achieved by using a hierarchical storage system, in which video files are stored in tertiary libraries or jukeboxes and transferred to a secondary level (of magnetic or optical disks) for display. We address the design of such a system by specifying the required architectural parameters (the bandwidth and storag...
... During 1990s, many media players, e.g. Windows Media Player 6.4 of Microsoft [128] and QuickTime of Apple [125], were developed after StarWorks [109], which is the first commercial streaming product, appeared in 1992. Around 2002, the launch of Adobe Flash [121] which is a single and unified video streaming platform largely promoted the development of the field of video streaming. ...
With the growth of the Internet, video streaming services have been rapidly and widely deployed in the recent years. Different video applications are developed to satisfy heterogeneous users’ demands and preferences. For example, 360 degree video streaming provides users with immersive experience and freedom to explore the 360 degree environment. In addition, to meet the needs of real-time content delivery, ultra low latency live video streaming services focus on achieving sub-second playback latency. The specialties of these applications can definitely provide ultimate video experience to the users. However, at the same time, they also generate special challenges. To guarantee high quality of experience (QoE) for 360
degree video streaming, instead of buffer control and video rate adaptation, accurate user Field-of-View (FoV) prediction and optimized rate allocation become necessary. In ultra low latency live video streaming systems, the “live” property makes the playback latency and buffer length necessarily short. Correspondingly, it leads to higher risk of running into video freeze under dynamic network environment. In order to provide high quality and low latency experience to the users simultaneously, the video rate and playback speed must be adapted in a coordinated fashion.
In this thesis, we conduct in-depth research of video streaming technologies in two types of video applications: 360 degree video streaming and low latency live video streaming. In the work of 360 degree video streaming, we propose a novel coding/streaming strategy, namely two tier structure, in which the 360 degree video is streamed in both base tier and enhancement tier. The buffer control and rate allocation between the two tiers are also optimized so that user QoE can be further improved. Through experiments driven by real bandwidth and user FoV traces, we demonstrate that the proposed two-tier systems substantially outperform the regular DASH streaming and the single tier system. In addition, we also propose the concept of “flock” which is defined as a group of users watching the same live 360 degree event with different latency. A collaborative FoV prediction scheme is proposed in which the actual FoV information of users in the front of the flock is utilized to predict the FoV of the users behind them. By establishing the
“streaming flock”, the effective video rate is almost doubled when compared with the benchmark using single user self-prediction. Finally, we propose both model-based and model-free solutions to adapt video rate and playback speed jointly for low-latency live streaming. Our joint adaptation solutions achieve close-to-optimal performance in a wide range of dynamic network conditions.
... Data is read from disk into the RAM buffer in relatively large chunks (in order to reduce disk-access overhead), and data for multiple video streams is then streamed out onto the distribution network in small units, such as ATM cells. While in the server, data may also be operated upon for purposes such as error correction, encryption and contentcustomization. Interesting papers on various issues pertaining to real VOD servers include [4] and [5]. ...
A video-on-demand (VOD) storage server is a parallel, storage-centric system used for playing a large number of relatively slow streams of compressed digitized video and audio concurrently. Data is read from disks in relatively large chunks, and is then "streamed" out onto a distribution network. The primary design goal is to maximize the ratio of the number of concurrent streams to system cost while guaranteeing glitch-free operation. This paper focuses on load-balancing for the purpose of providing throughput that is independent of viewing choices. At the interdisk level, data striping is the obvious solution, but may lead to a quadratic growth of RAM buffer requirements with system size. At the intradisk level multizone recording results in variable disk throughput. Deterministic schemes for solving each problem are discussed, as well as their joint operation. Finally, efficient staging of data from tertiary storage devices to disk is shown to be possible.
... A conventional network based file server, such as NFS, is not well suited to function as a video server. The main reason is that conventional file servers are designed for small data files whose usage and semantics are fundamentally different from those of real-time continuous media objects [22, 21, 12]. Real-time audio and video streams possess temporal characteristics not found in conventional data. ...
This paper presents scheduling and admission control algorithms for access to processor and storage resources of a video file server. The scheduling algorithms support multiple classes of tasks with diverse performance requirements, allow for the co-existence of guaranteed real-time requests with sporadic, and unsolicited requests, and ensure system stability during overloads. Performance guarantees are maintained for real-time streams in the presence of unpredictably varying non real-time traffic. A prototype video file server was implemented on an Intel 486 platform. Performance results show that a large number of streams can be supported, while maintaining efficient utilization of system resources.
... A conventional network based file server, such as NFS, is not well suited to function as a video server. The main reason is that conventional file servers are designed for small data files whose usage and semantics are fundamentally different from those of real-time continuous media objects [22, 21, 12]. Real-time audio and video streams possess temporal characteristics not found in conventional data. ...
This paper describes the design and implementation of a continuous media file server intended for use in emerging video-on-demand applications. The main focus and contribution of the paper is in scheduling and admission-control algorithms for accessing the server's processor and storage resources. The scheduling algorithms support multiple classes of tasks with diverse performance requirements and allow for the co-existence of guaranteed real-time requests with sporadic, and unsolicited requests. The scheduler maintains performance guarantees for real-time streams in the presence of unpredictably varying non-real-time traffic while ensuring system stability even during overloads. A prototype video file server was implemented on an Intel 486 platform. Performance results show that a large number of streams can be supported, while maintaining efficient utilization of system resources.
... A conventional network based file server, such as NFS, is not well suited to function as a video server. The main reason is that conventional file servers are designed for small data files whose usage and semantics are fundamentally different from those of real-time continuous media objects [22, 21, 12]. Real-time audio and video streams possess temporal characteristics not found in conventional data. ...
... O NE common architecture shared by most existing videoon-demand (VoD) systems is that they are based on a single server. The video server can range from a standard PC for small-scale systems [1], [2] to massively parallel supercomputers with thousands of processors for large-scale systems [3], [4]. However, the price/performance ratio escalates quickly for high-end hardware, and ultimately the capacity of a single server is still limited. ...
Most existing commercial video servers are designed for a single
server. Consequently, the capacity of the system in terms of maximum
sustainable concurrent sessions is limited by the performance of the
video server hardware. This paper proposes and analyzes the performance
of a novel parallel video server architecture where video data are
striped across an array of autonomous servers. The architecture allows
one to build incrementally scalable video servers without video data
replication. The proposed concurrent-push scheduling algorithm allows
the system to integrate with quality of service guarantees provided by
today's switching networks. In this paper, the striping policy, the
service model, and the concurrent-push scheduling algorithm are
presented. A system model is constructed to quantify three performance
metrics, namely, server buffer requirement, client buffer requirement,
and system response time. Results show that a simple extension of the
server-push service model does not perform well under the parallel video
server architecture. To improve system performance, a novel extension of
the grouped sweeping scheme called the asynchronous grouped sweeping
scheme (AGSS) is introduced. To further increase the scalability of the
architecture, a new subschedule striping scheme (SSS) is introduced.
With the proposed AGSS and SSS, our parallel video server architecture
can be scaled up to more than 10000 concurrent users
... Consequently, they may fail to deliver data on time, which will result in unacceptable playback quality. Real-time file servers, on the other hand, address the problem of real-time delivery over local area networks by carefully scheduling I/O operations to meet the consumption constraints of their clients [14,25,22,5,24,10]. ...
The design of a distributed video-on-demand system that is suitable for large video libraries is described. The system is designed to store 1000s of hours of video material on tertiary storage devices. A video that a user wants to view is loaded onto a video file server close to the users desktop from where it can be played. The system manages the distributed cache of videos on the file servers and schedules load requests to the tertiary storage devices. The system also includes a metadata database, described in a companion paper, that the user can query to locate video material of interest. This paper describes the software architecture, storage organization, application protocols for locating and loading videos, and distributed cache management algorithm used by the system.
... To support the above capabilities for application layer protocols, we designed two protocol suites for data and continuous stream services having the following features: 1. Fault-detection capability -the application protocol must be able to detect server failures efficiently and take appropriate actions to mask the failure (see below); 2. Fault-tolerance transparency -retrieval of an object during server failure should be transparent to the user unless the failure exceeds the fault-tolerance capability for the object; 3. Quality-of-service -the QOS requirement (if any) of a service must be maintained. While there are many studies on multimedia protocols [5][6][7], their emphasis are on the third point, i.e. QOS requirement for individual services. ...
In [1] we reported the design and implementation of a video-on-demand (VOD) system using a server array architecture. In this paper, we extend the work into a full multimedia server array system for world-wide-web services. The system allows stream-type services such as video and audio guaranteed continuous service irrespective of the background data traffic. The server array approach has the benefits of (1) more system capacity, as individual server has individual CPU, disk, and network channels, (2) scalable, as the system capacity increases with the number of servers without the need for data duplication, and (3) fault tolerant, as server-level fault-tolerant schemes can be devised for sustaining service even if a server fails. We describe the design and implementation of our WebArray system - a server array software system developed for multimedia world-wide-web services. Our implementation employs integrated scheduling at the disk and network subsystems to ensure the continuous delivery of video and audio. 1.
... These systems are designed to deliver movies to a set-top device connected to a television in a home or to a multimedia desktop computer. Developers of these systems have focussed on the network and file system performance problems of servicing many concurrent users [7][9] [20][23] [29]. They have not addressed the problem of finding a video in a large collection because these VOD applications require access to relatively few distinct titles. ...
Video-on-Demand systems have received a good deal of attention recently. Few studies, however, have addressed the problem of locating a video of interest in a large video database. This paper describes the design and implementation of a metadata database and query interface that attempts to solve this information retrieval problem. Sample queries were collected by interviewing a variety of users. These queries guided the design of indexes that can be used to answer the types of queries users want to ask. Three types of indexes were identified: 1) bibliographic (e.g., title, abstract, subject keywords, genre, director, cast, etc.), 2) structural (e.g., segments, scenes, shots, transitions, sound effects, etc.), and 3) content (e.g., keyframe sequences, scripts, objects and actors in scenes, etc.). A mixed-mode query interface was built that allows a user to select a set of videos and/or frame sequences and incrementally modify the answer set. The interface includes relational, hierarchi...
This ground-breaking study explores transformations in the TV industry under the impact of globalizing forces and digital technologies. Chalaby investigates the making of a digital value chain and the distinct value-adding segments which form the new video ecosystem. He provides a full account of the industry's global shift from the development of TV formats and transnational networks to the emergence of tech giants and streaming platforms. The author takes a deep dive into the infrastructure (communication satellites, subsea cable networks, data centres) and technology (cloud computing, machine learning and artificial intelligence) underpinning this ecosystem through the prism of global value chain theory. The book combines empirical data garnered over 20 years of researching the industry and offers unique insights from television and tech executives.
Computer communication demands for higher bandwidth and smaller delays are increasing rapidly as the march into the twenty-first century gains momentum. These demands are generated by visualization applications which model complex real time phenomena in visual form, electronic document imaging and manipulation, concurrent engineering, on-line databases and multimedia applications which integrate audio, video and data. The convergence of the computer and video worlds is leading to the emergence of a distributed multimedia environment. The initial step in providing multimedia services over computer networks is to ensure bandwidth availability for these services. In this paper a computational framework is developed in the context of the bandwidth needs for traffic generated in a distributed multimedia environment. Thereafter, this model is applied to the real-time problem of designing a backbone for a distributed multimedia environment at the NASA Classroom of the Future Program. The network incorporates legacy LANs and the latest high speed switching technologies. Performance studies have been conducted with different network topologies for various multimedia application scenarios to establish benchmarks for the operation of the network. In these performance studies it has been observed that network topologies play an important role in ensuring that sufficient bandwidth is available for multimedia traffic.
To cope with the increasing demand of customers, more magnetic disks may need to be added to a multimedia server to support a higher number of concurrent accesses. Due to the rapid improvement in the performance of magnetic disks, the newly added disks usually have higher sustained I/O bandwidth and bigger storage capacity. The traditional uniform striping in an environment with heterogeneous disks in the same striping group cannot fully utilize the I/O bandwidth and storage capacity of the newly added, higher performance disks. In this paper, we propose weighted striping techniques to fully utilize the I/O bandwidth and storage capacity of higher performance disks in a multimedia server. Our results show that weighted striping techniques greatly improve the performance of multimedia servers.
Today, vast amounts of text, images, graphics, animation, and even Java applets are being hosted and delivered by the WWW. With rapid advances in audio and video technologies, more and more contents will be encoded and delivered by means of audio/video in addition to texts and images. However, conventional Web servers are designed for data services and hence have no provisions for delivering continuous media such as audio and video. While one can still provide continuous-media services by a separate server in addition to the Web server, extra hardware cost and management complexity are incurred. Moreover, unused capacity in the Web server cannot be utilized even if the continuous-media server is overloaded, and vice versa. This paper presents a design and implementation of a server which integrates continuous-media services with traditional Web services. To resolve disk and network contentions, a simple yet effective fixed-priority scheduler is employed. Experimental results show that the proposed scheduler performs well with non-real-time hardware and operating system platforms.
This paper describes a multimedia surveillance system, the Media Controller. The Media Controller is designed for industrial surveillance and monitoring applications, and it provides the realtime recording and retrieval of digital video data. Considering the characteristics of industrial surveillance applications, the distributed recording of video data is shown to be necessary for surveillance systems. A method of managing video data in Media Controller is also described. It can effectively manage surveillance video data which is continuous (realtime), distributed, temporal, and partial. We have implemented a prototype, which confirmed the effectiveness of the Media Controller system.
Future Video-On-Demand (VOD) servers will need to support many existing and emerging video data types. These data types include 15-fps (frames per second) animation, 30- fps NTSC (National Television Systems Committee) quality video and 60-fps HDTV (High Definition Television) video. The different display speeds and frame sizes of these various video types impose a major constraint on the design of VOD storage systems. This paper presents the results of an experimental study, conducted on a Silicon Graphics Inc. Onyx computer system, that investigated the impact of these video types on the design of a VOD storage system. The key issues involved in supporting these different video types in a VOD environment are: (1) the video allocation method, and (2) the proper block size (a block is a basic unit of several contiguous video frames that will be accessed from several disks each time a request is made) to use for data striping and retrieval. Two allocation schemes, logical volume striping and application level striping, along with varying frame and block sizes for each of the three different video data types are examined in this paper. The focus of our study is to determine the maximum number of concurrent accesses that can be supported with a guaranteed quality of service. The degree of scalability (i.e., striping data over more disk arrays) of the experimental VOD system used is also studied.
Stony Brook Video Server (SBVS) is an Ethernet based video server
that employs only off the shelf PC components and is capable of
guaranteeing real time delivery of digital video streams from the
server's disk subsystem, through a shared local area network, to an end
user's display. SBVS integrates a real time software based disk array
with a Real Time Ethernet Protocol, RETHER to provide end to end
bandwidth guarantee. The paper reports the results of a detailed
performance evaluation study of SBVS, specifically the efficiency with
which SBVS utilizes the hardware capabilities of Fast Ethernet and SCSI
disks in the context of video servers, and the overhead of integrating
real time I/O and network components into a complete system. Two sets of
experiments were performed: (1) experiments that measure the maximum
performance of individual I/O and network hardware when operating based
on the algorithms used in SBVS; and (2) experiments that evaluate SBVS
as a whole in terms of the maximum number of video streams that can be
supported under varying stream rates, disk retrieval sizes, and number
of network interfaces. The major results are that SBVS-2, a Pentium-90
PC with a four-drive disk array and a single 100 Mbps Ethernet, can
support up to a maximum of 45 simultaneous MPEG-1 (1.5 Mbits/sec)
streams, and that using a six-drive array and serving two 100 Mbps
Ethernet segments, SBVS can support up to a maximum of 65 MPEG-1
streams, with CPU being the performance bottleneck
The video-on-demand (VOD) service is widely notable and the
technology to realize the VOD service is regarded as important. In this
paper, we propose a video server based on an ATM backbone network. This
video server supports VOD service. This video server is composed of an
ATM backbone network that links video segment file (VSF) servers storing
video data, and sequence control (SC) brokers controlling the reading of
video data from VSF servers and supplying video data streams to users.
In this system, video data is divided into VSFs and the VSFs are stored
on the disks in the VSF servers at random. With this architecture, it is
possible to construct various scales of video servers if the number of
VSF servers, the number of SC brokers, the number of disk drives per VSF
server and the switching capacity are chosen to offer to users the
demanded simultaneous accesses at the demanded video data rate. The
performance was evaluated by simulation and we clarified that it is
possible to support 32 simultaneous accesses at 3 Mbit/s video data.
Streams by using one SC broker and one VSF server which has 16 disk
drives of 3 MByte/s output rate, and 500 simultaneous accesses is able
to be offered by a system which has 16 SC brokers, 16 VSF servers and a
16×16 ATM switch. As a result, we confirmed that a video server of
this architecture is realizable from small scale to large scale
Video-on-demand (VOD) systems can support multiple concurrent video accesses with guaranteed user-controlled quality of service (QoS). The growth of VOD systems is driven by the increasing need for distributed multimedia applications such as company training, distance learning, telemedicine, and home entertainment. We propose and demonstrate an ATM-based (asynchronous transfer mode) VOD system (including VOD servers, clients and connection networks) based on a Macintosh platform. ATM is a promising network standard which may satisfy various requirements of VOD systems. In this VOD system, time-based data are encoded and decoded using the MPEG (Moving Picture Experts Group) standard. Issues related to VOD system design, including the multimedia programming environment, the required system modification to support remote MPEG playback, ATM networking, process scheduling, and buffer management for both VOD servers and clients, are discussed and analyzed based on the Macintosh platform. A prototype VOD system is used to explore the design issues proposed. The Macintosh platform provided an efficient and jitter-free VOD system supporting MPEG movie playback through ATM networks
In this paper, the design and implementation of a video conference
recorder on an inexpensive PC is described. The recorder can
simultaneously record several video and audio (A/V) streams in Unix file
format. The objectives of the file system design are high throughput and
low loss rate. A hard disk device driver and a file system are modified
and tuned for A/V recording. The new file system can have up to 60%
performance improvement over the traditional file system. Performance
results from our experiments are presented and discussed in the paper
Integrating multimedia into real-time control applications requires careful consideration. The impact of the system support needed to meet the soft real-time constraints of multimedia on meeting the hard deadlines of the control application is of paramount concern. We propose that integrated resource models and reflective real-time kernels can provide the basis for the needed solutions. Features such as quality of service and call admission would then span a spectrum of requirements from the soft real-time requirements of multimedia to the hard real-time constraints of the control application. The confluence of multimedia, real-time control and real-time active databases is also discussed
A video-on-demand (VOD) storage server is a parallel, storage-centric system used for playing a large number of relatively slow streams of compressed digitized video and audio concurrently. Data is read from disks in relatively large chunks, and is then “streamed” out onto a distribution network. The primary design goal is to maximize the ratio of the number of concurrent streams to system cost while guaranteeing glitch-free operation. This paper characterizes the VOD applications and then identifies several important issues along with an outline of possible approaches to dealing with them. Issues include the accommodation of unlimited demand for concurrent “private” viewing of the few hottest movies with limited resources, multi-zone recording and the resulting variable disk transfer rate, as well the interplay between fault-tolerance, load balancing, the size of RAM buffers and the organization of the storage subsystem
Video-on-demand is a new entertainment service that will soon be widely available. A small amount of material is very popular, while large amounts of material are viewed less frequently. This skew can be exploited by using a storage hierarchy, storing the less frequently viewed videos in lower-cost tertiary storage. This paper studies the use of tertiary storage for videos. Tertiary storage devices such as optical disks and magnetic tapes can be used to a) deliver data directly to viewers, or b) to stage data to disk for viewing. Analysis of these modes yields guidelines for server design. Examining device characteristics, workload characteristics, and cost, the two modes are compared to each other and to playing from disk. The data placement decision depends on the fraction of time a stream of a video is active. At current costs, videos having an active stream less than a third of the time should reside on tertiary storage. When a tertiary library has a much higher data rate than the video rate, videos should be staged disk for playing. Otherwise, they should be played directly from tertiary store
The design issues for a scalable video server is described and,
according to these issues, the system architecture of the scalable video
servers are proposed. The system uses ATM switching technology to scale
up the dimension of the service. There are four major components within
the proposed system, namely the server headend control (SHC) unit, the
stream router, the video pumping unit (VPU) and distributed storage
devices. The functions of each of the components are described and the
key components in the system are the stream router and the VPU.
Operations and performance limitations on the VPU are given and the
result shows that a single storage device can serve only a limited
number of video streams. Finally, the paper focuses on the design of the
stream router as it is used to connect multiple VPUs for producing
guaranteed throughput and bounded delay variance for each video stream
There are many problems to be solved for realizing
full-interactive on-demand service. It is essential to increase the
number of simultaneously accessible streams, achieve time control of the
delivering streams, and be able to deliver requested titles and serve
trick plays with a quick response. A method to solve each of these
problems is proposed, in terms of the video file system, disk
scheduling, and the network. A prototypical VOD system is constructed,
which includes an interactive network-interface, application protocols
and title management. It is confirmed that these proposals are feasible
Most video-on-demand (VOD) systems use a single, powerful server
to deliver video streams to users. We consider a novel server array
approach for delivering video services on networks. Such an approach has
the benefits of (l) greater system capacity, as individual server has
individual disk and network channel, (2) scalability, as more clients
can be supported by adding more servers without data duplication, and
(3) fault tolerance, as server-level fault-tolerant and fault-recovery
schemes can be devised. We describe our experiences and results in the
implementation of a server-array-based VOD system. Our system now has
four P5-90 servers serving 40 PC-486 stations using a 10 Mbps Ethernet
switch. The system can deliver 40 simultaneous and independent 1.2 Mbps,
30 fps, full TV size, MPEG-1 video streams. Our results demonstrate that
with careful protocol design, Ethernet can deliver continuous video and
audio services
Multi-zone recording increases disk capacity by approximating
fixed linear recording density. With fixed rotation speed, however,
transfer rate varies with track location, as does the number of video
streams that can be played concurrently. This paper proposes Track
Pairing, a deterministic scheme for static intra-disk data layout. By
recording each movie alternately on “outer” tracks and their
“inner” counterparts, a disk's throughput becomes
independent of the viewers' choices and its guaranteed streaming
capacity is maximized. With a 1.8:1 ratio of outermost to innermost
track capacities, guaranteed streaming capacity is increased by 40
percent, and is merely 22 percent below the streaming capacity of the
outermost track! Temporal overhead is modest, and required buffer sizes
are smaller than those with the Logical Track scheme, which also
maximizes guaranteed throughout. Track-Pairing has been implemented
under Microsoft's Windows NT, and can be extended to multiple disk
drives
The expansion of multimedia networks and systems depends on
real-time support for media streams and interactive multimedia services.
Multimedia data are essentially continuous, heterogeneous, and
isochronous, three characteristics with strong real-time implications
when combined. At the same time, some multimedia services, like
video-on-demand or distributed simulation, are real-time applications
with sophisticated temporal functionalities in their user interface. We
analyze the main problems in building such real-time multimedia systems,
and we discuss-under an architectural prospect-some technological
solutions especially those regarding determinism and efficient
synchronization in the storage, processing, and communication of audio
and video data
The use of emerging multimedia technologies in education will
cause a major shift in the educational services paradigm that promises
major advantages over present analog distance learning systems. However,
the difficulties encountered with the simultaneous introduction of new
services and new technologies often impede the accomplishment of set
goals. A more prudent and feasible first step might well be to use the
new technologies while following existing paradigms. Such a step will
then facilitate the transition to new services that capitalize on the
new technologies. Abiding by this principle, researchers at Stanford are
developing a prototype on-line distance learning system based on recent
advances in digital video compression, video servers, and multimedia
networking to deliver live and stored video lectures and seminars to
students at their desktops. For the sake of concreteness, we are
designing this system with the Stanford University Instructional
Television Network in mind, as a first step beyond existing video-based
instructional services. The prototype system and extensions, however,
are general enough to apply to other environments and application
domains
Previous studies on disk scheduling for video services were
usually based on computer simulation. We present analysis on disk
scheduling for video services. This paper first gives a short review on
the various disk scheduling algorithms. It then concentrates on the
analysis of two major disk scheduling algorithms, namely, CLOOK and
LOOK. The purpose of the analysis is to obtain the maximum number of
simultaneous users supported by systems using these two algorithms. The
results of the analysis show that the CLOOK algorithm performs slightly
better than LOOK algorithm in video applications. Then, this paper
discusses disk scheduling for supporting special user functions such as
“forward search” and “reverse search”. It is
shown that the maximum number of user streams supported drops
dramatically when such user functions are used. A technique called
redundancy for special user functions (RESUF) is then studied. Analysis
of the technique shows that it can keep the I/O demands to almost
constant under all user request conditions
In conventional video-on-demand systems, video data are stored in
a video server for delivery to multiple receivers over a communications
network. The video server's hardware limits the maximum storage capacity
as well as the maximum number of video sessions that can concurrently be
delivered. Clearly, these limits will eventually be exceeded by the
growing need for better video quality and larger user population. This
paper studies a parallel video server architecture that exploits server
parallelism to achieve incremental scalability. First, unlike data
partition and replication, the architecture employs data striping at the
server level to achieve fine-grain load balancing across multiple
servers. Second, a client-pull service model is employed to eliminate
the need for interserver synchronization. Third, an admission-scheduling
algorithm is proposed to further control the instantaneous load at each
server so that linear scalability can be achieved. This paper analyzes
the performance of the architecture by deriving bounds for server
service delay, client buffer requirement, prefetch delay, and scheduling
delay. These performance metrics and design tradeoffs are further
evaluated using numerical examples. Our results show that the proposed
parallel video server architecture can be linearly scaled up to more
concurrent users simply by adding more servers and redistributing the
video data among the servers
Stony Brook Video Server (SBVS) is an Ethernet-based video server
from the Experimental Computer Systems Laboratory, SUNY at Stony Brook,
and has been fully operational for over a year. The SBVS employs only
off-the-shelf PC components and is capable of guaranteeing real-time
delivery of digital video streams from the server's disk subsystem,
through a shared LAN, to an end user's display. The SBVS integrates a
real-time software-based disk array with the Real-Time Ethernet Protocol
to provide end-to-end bandwidth guarantee. This article describes in
detail how the real-time I/O subsystem interacts with the RETHER
subsystem, and the performance cost of this integration. The most recent
version of SBVS, a Pentium-90 PC with a six-drive disk array, can
support up to 45 simultaneous MPEG-1 (1.5 Mb/s) streams with a single
100 Mb/s Ethernet link, and up to 65 streams with two 100 Mb/s Ethernet
links
Future video-on-demand (VoD) servers will need to support many
existing and emerging video data types. These data types include 15-fps
(frames per second) animation, 30-fps National Television Systems
Committee (NTSC) quality video, and 60-fps high definition television
(HDTV) video. The different display speeds and frame sizes of these
various video types impose a major constraint on the design of VoD
storage systems. This paper presents the results of an experimental
study, conducted on a Silicon Graphics Inc. (SGI) Onyx computer system,
that investigated the impact of these video types on the design of a VoD
storage system. The key issues involved in supporting these different
video types in a VoD environment are as follows: (1) the video
allocation method and (2) the proper block size (a “block”
is a basic unit of several contiguous video frames that will be accessed
from several disks each time a request is made) to use for data striping
and retrieval. Two allocation schemes, logical volume striping and
application level striping, along with varying frame and block sizes for
each of the three different video data types are examined. The focus of
our study is to determine the maximum number of concurrent accesses that
can be supported with a guaranteed quality-of-service (QoS). The degree
of scalability (i.e., striping data over more disk arrays) of the
experimental VoD system used is also studied. Based on our experimental
results, application level striping demands smaller block sizes for all
three video types, and more concurrent accesses can be distributed over
the storage devices. The experimental results demonstrate that
application level striping has excellent scalability for animation and
NTSC videos
This paper presents the design of a generic video access device
called the stream pumping engine (SPE) for various video applications
such as scalable video-on-demand systems, or voluminous digital video
storage. The SPE integrates both the storage and network controllers in
one board and serves as the basic building block of digital multimedia
server systems. Storage devices such as SCSI hard disk drives, or
digital tapes, as well as network devices like Ethernet, and EtherSwitch
can therefore be connected together in a more efficient way. This paper
first outlines the design of the SPE, and then illustrates the
interconnection structure of multiple SPEs and concludes with the
applications of the SPE
One of the most important requirements to bring the interactive
video-on-demand services to residential consumer use is the scalability
of the video server. The design issues for a scalable video server is
described and, according to these issues, the system architecture of the
scalable video servers are proposed. The system uses ATM switching
technology to scale up the dimension of the service. There are four
major components within the proposed system, namely the server headend
control (SHC) unit, the stream router, the video pumping unit (VPU) and
distributed storage devices. The functions of each components is
described and the key components in the system are the stream router and
the VPU. Operations and performance limitations on the video pumping
unit are given and the result shows that a single storage device can
serve only a limited number of video streams. Finally the paper focuses
on the stream router as it is used to connect multiple VPUs.
Experimental results show that the proposed stream router can produce
guaranteed throughput and bounded delay variance for video applications
The design and implementation of a video conference recorder on an
inexpensive PC is described. The recorder can simultaneously record
several video and audio (A/V) streams in Unix file format. The
objectives of the file system design are high throughput and low loss
rate. A hard disk device driver and a file system are modified and tuned
for A/V recording. The new file system can have up to 60% performance
improvement over the traditional file system. Performance results from
our experiments are presented and discussed
With the advent of multimedia computing, there is an emerging need for systems that can support digital continuous media without requiring special adaptation logic on the part of application programs and that be implemented on existing network infrastructures. In this paper the system architecture of the Stony Brook Video Server (SBVS) is described. To guarantee real-time endto -end performance, SBVS uses a real-time network access protocol, RETHER, that uses existing Ethernet hardware. SBVS tightly integrates the bandwidth guarantee mechanisms between network transport and disk I/O. SBVS's stream-by-stream disk scheduling scheme optimizes the effective disk bandwidth without incurring scheduling overhead every cycle. In addition, SBVS implements multi-resolution video coding to reduce network and I/O bandwidth demands in normal viewing mode, while supporting fast forward/backward without requiring extra bandwidth.
The introduction of computers can revolutionize distance learning in much the same way the introduction of the blackboard revolutionized classrooms in the 1860's. Computers allow the geographic dispersion of the instructors and the students, support active interaction among students and instructors in the learning process, provide a media-rich environment for presentation and viewing of class material, accommodate each individual's pace of learning, and eliminate the unnecessary note-taking usually associated with the old classroom model. In this paper, we present four basic requirements for a distance learning environment, motivate our vision of an ideal distance learning environment through demonstrative scenarios, and identify learner's educational activities and match them with their corresponding system capabilities. 2 Introduction Distance learning refers to the concept of students who are physically remote from the base location of their education. It is a mechanism...
We describe an experimental system supporting the creation of highly interactive simulations which are rich in media and structure. Strengths include author-modifiable schema, author-reconfigurable front ends, and relatively straightforward extensibility to new media types and search methods. The framework of meta-media distributed databases (MMDD) manages associations of arbitrary renderable objects (such as structured text, 3D-graphics, musical scorefiles, video, certain executables) using a hybrid object-oriented, relational database abstraction. Key features of the architecture include (1) atoms that are not traditional documents but abstract entities, which endows the MMDD with unusual flexibility, and (2) object-oriented core mediating classes which provide rational, extensible meta-data- and content-based services. The MMDD was designed around the projected needs of faculty and their assistant authors, based on extensive experience with creating multimedia simulations. It curr...
On-demand video servers based on hierarchical storage systems are able to offer highcapacity and low-cost video storage. In such a system, video files are stored in the tertiary level and transferred to the secondary level to be displayed. The design of video servers allowing user interaction with the playbacked video is of great interest. We have conducted a comprehensive study on the architecture and operation of such servers based on hierarchical storage systems for interactive video-on-demand. Our objective is to understand its performance characteristics, so as to design a video server to meet specific application requirements. The applications of interest are many: distance-learning, movie-on-demand, interactive news, home-shopping, etc. The design of such a server actually involves many design choices pertaining to both architecture and operational procedures. As far as architecture is concerned, we need to consider such system parameters as bandwidth, storage capacity, and the...
ResearchGate has not been able to resolve any references for this publication.