Buddhikot, Milind M.; Parulkar, Gurudatta M.; Kumar, SriHari; and Rangan, P. Venkat "Design of Storage Servers and Storage Hierarchies" To appear in "Handbook of Multimedia Systems," Prentice Hall.
-
Rapid advances in communication, high speed packet switching, mobile data communications, media compression and processor and memory design will make a distributed multimedia computing infrastructure akin to that shown in figure 1 feasible in the near future. Exciting interactive applications such as multimedia mail, orchestrated presentations, high quality on-demand audio and video, collaborative multimedia document editing, large digital libraries or multimedia archives, and virtual reality environments will become widespread in this infrastructure. A user may want to access these applications from a variety of end-systems such as Personal Digital Assistant (PDA), a set top box with a television set, a high end personal computer at home or a workstation in the workplace. The interactive nature of these applications requires that the user be able to control playout or content of multimedia streams such as video, audio and graphics. Specifically, a user may perform playout control commands such as fast-forward, rewind, pause/resume, fast-search and random access search or may request media editing functions such as clip, enhance, filter, restore etc. Also, most of these applications handle multimedia information in stored or static form. Therefore, the performance of such applications directly depends on the performance of storage servers that store such multimedia information.
Gong, Fengmin; Parulkar, Gurudatta M. ``An Application-oriented Error Control Scheme for A High Speed Transport Protocol,'' To appear in ACM/IEEE Transactions on Networking December, 1996.
- Many new network applications demand interprocess communication (IPC) services that are not supported by existing transport protocol mechanisms. Large bandwidth-delay products of high-speed networks also render some of the existing error and flow control mechanisms less efficient. In particular, new error control schemes that can provide variable degrees of error recovery accorking to the application's requirements are desirable. This paper presents the design, evaluation, and implementation of an application-oriented error control scheme that is aimed at supporting efficient IPC in high-speed networking environments. Our results indicate that the proposed error control scheme allows effective control of trade-off between the amount of error an application can tolerate and the amount of delay it has to suffer.
- Gokhale, Aniruddha; Harrison, Tim; Schmidt, Douglas C.; and Parulkar, Gurudatta M. "Operating System Support for High-Performance, Real-time CORBA," Proceeding to the 5th International Workshop on Object-Orientation in Operating Systems, IEEE, October 1996.
- A broad range of applications (such as avionics, telecommunication systems, and multimedia on demand) require various types of real-time guarantees from the underlying middleware, operating systems, and networks to achieve their quality of service (QoS). In addition to providing real-time guarantees and end-to-end QoS, the underlying services used by these applications must be reliable, flexible, and reusable. Requirements for reliability, flexibility and reusability motivate the use of object-oriented middleware like the Common Object Request Broker Architecture (CORBA). However, the performance of current CORBA implementations is not suitable for latency-sensitive real-time systems (e.g., avionics), and constrained latency systems (e.g., teleconferencing).
This paper describes key changes that must be made to the CORBA specifications, existing CORBA implementations, and the underlying operating system to develop real-time ORBS (RT ORBs). RT ORBS must deliver real-time guarantees and end-to-end QoS to latency-sensitive applications. While many operating systems now support real-time scheduling, they do not provide integrated solutions. The main thesis of this paper is that advances in real-time distributed object can be achieved only by simultaneously integrating techniques and tools that simplify application development; optimize application, I/O subsystems, and network performance; and systematically measure performance to pinpoint and alleviate bottlenecks.
Adiseshu, Hari; Parulkar, Gurudatta M.; Varghese, George. "A Reliable and Scalable Striping Protocol," To appear in ACM SIGCOMM'96, Stanford University, 1996.
- Link striping algorithms are often used to overcome transmission bottlenecks on computer networks. Traditional striping algorithms suffer from two major disadvantages. They provide inadequate load sharing on the presence of variable length packets, and may result in non-FIFO delivery of data. We describe a new family of link striping algorithms that solves both problems. Our scheme applies to any layer that can provide multiple FIFO channels.
- Meijler, Theo D.; and Engel, Robert. "Making Design Patterns explicit in FACE, a Framework Adaptive Composition Environment," submitted to Europlop 1996.
-
Creating applications using object-oriented frameworks is done at a relatively low abstraction level, leaving a large gap with the high abstraction level of a design. This makes the use of a framework difficult, and allows design and realization to diverge. Design patterns are more specific elements of design, and thus reduce this gap. We even bridge this gap by making design patterns and the classes that play a role within them into special purpose software components. System realization becomes a matter of composing special purpose class-components. We also introduce a system, FACE, which supports the visual composition of such specifications.
Blaine, G. James; Cox, Jerome R., Jr.; Jost, R.G. "Networks for Electronic Radiology," Radiology Clinics of North America, May 1996, Vol. 34, No. 3, pp. 505-524.
- We provide a brief tutorial on relevant network principles and products, a perspective on the evolution of this field, a simple view of the requirements posed by electronic radiology, and finally we offer a somewhat biased view of networking developments that can be anticipated over the next few years.
Gopalakrishnan, R.; Parulkar, Gurudatta M. "Bringing Real-time Scheduling Theory and Practice Closer for Multimedia Computing", Proceedings of the ACM SIGMETRICS Conference on Measurement and Modeling of Computer Systems, Philadelphia, May 1996.
- This paper seeks to bridge the gap between theory and practice of real-time scheduling in the domain of high speed multimedia networking. We show that the strict preemptive nature of real-time scheduling leads to more context switching, and requires system calls for concurrency control. We present our scheduling scheme called rate-monotonic with delayed preemption (RMDP) and show how it reduces both these overheads. We then develop analytical framework to analyze RMDP and other scheduling schemes that lie in the region between strict (immediate) preemption. Our idealized scheduler simulation methodology accounts for the blocking introduced by these schemes under the usual assumption that the time for context switching and preemption is zero. We derive simpler schedulabilty tests for non-preemptive scheduling and prove a variant of rate-monotonic scheduling, and prove a variant of rate-monotonic scheduling that has fewer preemptions. Our measurements on Sparc and Pentium platforms, show that for the workload we considered, RMDP increases useful utilization by as much as 8%. Thus our scheduling policies have the potential to improve performance over existing methods.
Buddhikot, Milind M.; Parulkar, Gurudatta M.; Gopalakrishnan, R. "Scalable Multimedia-On-Demand via World-Wide-Web (WWW) with QOS Guarantees" Sixth International Workshop on Network and Operating System Support for Digital Audio and Video, NOSSDAV96, Zushi, Japan, April 23-26, 1996.
- This paper focuses on our prototyping effort presently underway at the Washington University in St. Louis under the NSF's National Challenge Award (NCA) grant aimed at deploying a scalable Multimedia-On-Demand (MOD) server. This server will offer a scalable MOS service that supports end-to-end Quality-of-Service (QOS) guarantees, high concurrency, and a web based access that provides complete playout control functions as random search, ff rw, pause, slow-play and content based searches. Rest of this abstract describes our approach to achieving these objectives.
Dittia, Zubin D.; Parulkar, Gurudatta M.; Cox, Jerome R., Jr. "Design and Implementation of a Versatile Multimedia Network Interface and I/O Chip," Proceedings of the 6th International Workshop on Network and Operating System Support for Digital Audio and Video (NOSSDAV), April 1996.
- We present an I/O architecture that is particularly well suited to the transport and delivery of continuous media data to devices distributed in the desk or home area. Our designs make use of the APIC chip, which is a high performance (1.2 GB/s) research prototype of an ATM network interface that we have developed at Washington University. Both hardware and software issues are considered. Due to space constraints, this extended abstract will only touch on aspects of our architecture that we consider important. More details will be presented in the full-length version of the paper.
Papadopoulos, Christos; Parulkar, Gurudatta M.; "Retransmission-Based Error Control for Continuous Media Applications," Proc. 6th Intl. Workshop on Network and Operating System Support for Digital Audio and Video (NOSSDAV), April 1996.
- We have designed and implemented a transmission-based error control scheme for CM applications, which aims to provide the best possibilities at a minimal cost, without violating the application's timing constraints. We have enhanced selective repeat retransmission with: (1) playout buffering to increase the time available for recovery, (2) gap-based rather than timer-based loss detection to minimize loss detection latency, (3) implicit expiration of sender retransmission buffers to eliminate acknowledgments, (4) conditional retransmission requests to avoid triggering
late, unnecessary retransmissions, and (5) data integrity information delivery to the application to
aid in concealment. Experimental results show that the mechanism significantly reduces observed
loss without violating the application's delay constraints.
Schmidt, Douglas C.; Parulkar, Gurudatta M.; Cranor, Charles D. "Gigabit CORBA - High-Performance Distributed Object Computing,'' Gigabit Networking Workshop (GBN'96), 24 March 1996, San Francisco, in conjunction with INFOCOM `96.
Wu, Dakang. "An Efficient Signaling Structure for ATM Networks,'' Proceeding of IEEE INFOCOM `96, March 26, 1996.
- As ATM becomes widely accepted as the communication standard for high speed networks, the signaling system structure and protocols that support ATM become more and more important. To support existing, future and unknown applications, the signaling system has to be very flexible and efficient. In this paper we define the signaling problem, present several possible signaling system structures, compare the advantages and disadvantages of these systems, and then we propose a new signaling system structure. The fundamental idea of the new signaling system is the logical separation of the signaling system structure from the underlying communication network, even though they may be built on the same physical network. The proposed signaling system structure shows very promising performance in terms of signaling latency and scalability.
Turner, Jonathan S. ``Maintaining High Throughput During Overload in ATM Switches,'' Proceedings of Infocom, 3/96.
- This paper analyzes two popular heuristics for ensuring packet integrity in ATM switching systems. In particular, we analyze the behavior of packet tail discarding, in order to understand how the packet level link efficiency is dependent on the rates of individual virtual circuits and the degree of the imposed overload. In addition, we study early packet discard and show that the queue capacity needed to achieve high efficiency under worst-case conditions grows with the number of virtual circuits and we determine the efficiency obtainable with more limited queue capacities. Using the insights from these analyses, extensions to early packet discard are proposed which achieve high efficiency with dramatically smaller queue capacities (independent of the number of virtual circuits).
Return to the top