R. West, K. Schwan, and C. Poellabauer. Scalable scheduling support for loss and delay constrained media streams. In Proceedings of the 5th IEEE Real-Time Technology and Applications Symposium. IEEE, June 1999.  Home/Search   Document Details and Download   Summary   Related Articles   Check

....notion of pfairness was introduced in [3] to minimize the length of time during which a task is denied service. This fairness notion enforces scheduling each periodic task proportional to its temporal property, which is the ratio of its execution time requirement to its period. A more recent study [20] addressed real time scheduling problem taking fair degradation into account. This study used a finite range (window) to keep track of packet losses of multimedia streams and employed a scheduling policy that mainly worked on the greatest degradation (loss) first basis. While a fairness issue with ....

R. West, K. Schwan, and C. Poellabauer, Scalable Scheduling Support for Loss and Delay Constrained Media Streams, IEEE Real-Time Technology and Application Symposium, 1998. 9

....QoS guarantees. At the network level, QoS support is either guaranteed by establishing real time channel [3] between the server and the client or attempted in a best effort manner [4] Similarly, at the server level, QoS support is either guaranteed [5, 6] or attempted in a best effort manner [7]. In this paper, we assume a particular type of real time traffic (known as a real time stream) from the server to a client, which is defined as a sequence of related packets that are transmitted at a regular interval (period) with certain delay constraints (deadline) 4] From the perspective of ....

....the perceived QoS of the client is reduced by the difference in usefulness between an on time packet and the tardy packet. Thus, a late packet can be viewed in the same category as a lost packet since late packets will also reduce the QoS perceived by the client. In this paper, loss rate [2, 7] can be defined as the fraction of packets that are discarded (dropped) or transmitted after their deadline. Although loss rate can be used as an overall QoS metric, a more applicable QoS metric can be used for streams that can tolerate occasional packet losses. Several streams in the media ....

[Article contains additional citation context not shown here]

R. West, K. Schwan, C. Poellabauer, "Scalable Scheduling Support for Loss and Delay Constrained Media Streams," RTAS'99.

....locations at which the losses occur can have a significant e#ect on the QoS perceived by the user. For instance, in Voice over IP (VoIP) a loss of several consecutive packets in a row may cut out an entire talkspurt whereas the same loss rate distributed over time may be imperceptible to the user [4, 5, 6]. Fault tolerance: For extremely critical communications, it may be not only a requirement to have zero loss but also that loss cannot be tolerated at any time, regardless of faults in the network. Examples of such applications might include remote surgery or process control. In such cases, it ....

R. West, K. Schwan, and C. Poellabauer, "Scalable Scheduling Support for Loss and Delay Constrained Media Streams," in Proc. IEEE Real-time Technology and Application Symposium (RTAS), 1999.

....QoS guarantees. At the network level, QoS support is either guaranteed by establishing real time channel [3] between the server and the client or attempted in a best effort manner [4] Similarly, at the server level, QoS support is either guaranteed [5, 6] or attempted in a best effort manner [7]. In this paper, we assume a particular type of real time traffic (known as a real time stream) from the server to a client, which is defined as a sequence of related packets that are transmitted at a regular interval (period) with certain delay constraints (deadline) 4] From the perspective of ....

....the perceived QoS of the client is reduced by the difference in usefulness between an on time packet and the tardy packet. Thus, a late packet can be viewed in the same category as a lost packet since late packets will also reduce the QoS perceived by the client. In this paper, loss rate [2, 7] can be defined as the fraction of packets that are discarded (dropped) or transmitted after their deadline. Although loss rate can be used as an overall QoS metric, a more applicable QoS metric can be used for streams that can tolerate occasional packet losses. Several streams in the media ....

[Article contains additional citation context not shown here]

R. West, K. Schwan, C. Poellabauer, "Scalable Scheduling Support for Loss and Delay Constrained Media Streams," RTAS'99.

No context found.

R. West, K. Schwan, and C. Poellabauer. Scalable scheduling support for loss and delay constrained media streams. In Proceedings of the 5th IEEE Real-Time Technology and Applications Symposium. IEEE, June 1999.

No context found.

R. West, K. Schwan, and C. Poellabauer. Scalable scheduling support for loss and delay constrained media streams. In Proceedings of the 5th IEEE Real-Time Technology and Applications Symposium. IEEE, June 1999.

No context found.

R. West, K. Schwan, and C. Poellabauer. Scalable scheduling support for loss and delay constrained media streams. In Proceedings of the 5th IEEE Real-Time Technology and Applications Symposium. IEEE, June 1999.

No context found.

R. West, K. Schwan, and C. Poellabauer. Scalable scheduling support for loss and delay constrained media streams. In Proceedings of the 5th IEEE Real-Time Technology and Applications Symposium. IEEE, June 1999.

No context found.

Richard West, Karsten Schwan, and Christian Poellabauer. Scalable scheduling support for loss and delay constrained media streams. Technical Report GIT-CC-98-29, Georgia Institute of Technology, 1998.

No context found.

R. West, K. Schwan, and C. Poellabauer. Scalable Scheduling Support for Loss and Delay Constrained Media Streams. In Proc. 5th Real-Time Technology and Applications Symposium, Vancouver, Canada, 1999.

No context found.

R. West, K. Schwan, and C. Poellabauer. Scalable Scheduling Support for Loss and Delay Constrained Media Streams. In Proc. 5th Real-Time Technology and Applications Symposium, Vancouver, Canada, 1999.

No context found.

R. West, K. Schwan, and C. Poellabauer. Scalable scheduling support for loss and delay constrained media streams. In Proceedings of the 5th IEEE Real-Time Technology and Applications Symposium. IEEE, June 1999.

No context found.

R. West, K. Schwan, and C. Poellabauer. Scalable scheduling support for loss and delay constrained media streams. In Proceedings of the 5th IEEE Real-Time Technology and Applications Symposium. IEEE, June 1999.

No context found.

R. West, K. Schwan, and C. Poellabauer. Scalable scheduling support for loss and delay constrained media streams. In Proceedings of the 5th IEEE Real-Time Technology and Applications Symposium. IEEE, June 1999.

No context found.

R. West, K. Schwan, and C. Poellabauer. Scalable scheduling support for loss and delay constrained media streams. In Proceedings of the 5th IEEE Real-Time Technology and Applications Symposium. IEEE, June 1999.

No context found.

R. West, K. Schwan, and C. Poellabauer. Scalable scheduling support for loss and delay constrained media streams. In Proceedings of 5th IEEE Real-Time and Embedded Technology and Applications Symposium (RTAS 1999), June 1999.

....Boston University Boston, MA 02215 richwest cs.bu. edu Ivan Ganev and Karsten Schwan College of Computing Georgia Institute of Technology Atlanta, GA 30332 fganev,schwang cc.gatech.edu Abstract This paper describes our experience using Dynamic Window Constrained Scheduling (DWCS) [13, 14, 12, 11] to schedule processes (and threads) on available CPUs in a Linux system. We describe the implementation of a kernel loadable module that replaces the default Linux scheduler. Each process scheduled using DWCS has a request period of T time units and a window constraint , x=y. The end of one ....

....is exceeded it may be impossible to interpret the received video sequence at the client. Likewise, if frames are generated too fast, playout bu ers at the client might over ow and some frames will again be lost. We have developed an algorithm called Dynamic Window Constrained Scheduling (DWCS) [13, 14, 12] to support real time and multimedia applications requiring rate based service constraints like those described above. DWCS was originally designed as a packet scheduler to provide (m; k) rm deadline guarantees [5] and fair queueing [2, 15, 3, 1, 4, 9, 10] for loss and delay constrained trac ....

[Article contains additional citation context not shown here]

R. West, K. Schwan, and C. Poellabauer. Scalable scheduling support for loss and delay constrained media streams. In Proceedings of the 5th IEEE Real-Time Technology and Applications Symposium. IEEE, June 1999.

....of the current window constraint are decremented and there fore the window constraint is relaxed. On the other hand, if a process misses to be scheduled within its period T, only the numerator is decremented, resulting in a tighter window constraint. Details about DWCS can be found in [32, 34]. The original DWCS algorithm works according to the following (simplified) rules: The process with the closest deadline (i.e. the time until its current period T expires) will be selected. If several processes have the same deadline, the process with the tightest current window constraint x y ....

R. West, K. Schwan, and C. Poellabauer. Scalable Scheduling Support for Loss and Delay Constrained Media Streams. In Proc. 5th Real-Time Technology and Applications Symposium, Vancouver, Canada, 1999.

....means that customized scheduling solutions can be provided, with the flexibility of software at hardware speeds. This ensures matching the needs of the constantly changing landscape of network services and protocols. Our previous experiences with host based and embedded software schedulers [11], has shown that meeting packet time requirements of multi gigabit links is difficult with software only realizations of scheduling disciplines. Scheduling disciplines must be able to make a decision within a packet ( packet length(in bits) time line speed(bps) tO maintain high link ....

....division for streams. In fact, DWCS can be con figured to operate as an EDF scheduler (loss tolerances are o ) static priority scheduler (infinite deadline, static priority is original loss tolerance of streams) and fair scheduler (WFQ weights can be set using deadlines and loss tolerances) [10, 11]. The next Section describes the ShareStreams architecture, the software architecture for managing packet queues along with a hardware scheduling discipline realization to schedule streams using the ShareStreams architecture. 3 The ShareStreams Hardware Software Ar chitecture Certain ....

[Article contains additional citation context not shown here]

R. West, K. Schwan, and C. Poellabauer. Scalable scheduling support for loss and delay constrained media streams. In IEEE Real Time Technology and Applications Symposium, pages 24 , 1999.

....Karsten Schwan Georgia Institute of Technology, College of Computing, 801 Atlantic Drive Atlanta, GA, 30332 0280 Shiw, xt2000, schwan cc.gatech.edu indrani ece.gatech.edu Abstract. This paper describes the design and implementation of the Dynamic Window Constrained Scheduling (DWCS) 1] 2][3] algorithm to schedule packets on network processors. The DWCS algorithm characterizes multimedia streams with diverse Quality of Service (QoS) requirements. Earlier implementations of DWCS on Linux and Solaris machines use a heap based implementation, which requires O(n) time to find the next ....

....by the QoS requirement of each stream. Packets of less priority not able to meet their QoS may be dropped to save computing power or bandwidth for more urgent packets. This paper presents an efficient implementation of the Dynamic Window Constrained Scheduling Algorithm (DWCS) 1] 2][3] on high speed programmable network processors. DWCS is designed to guarantee the QoS requirements of media streams with different performance objectives. Using only two attributes, deadline and loss tolerance, DWCS can maximize the network bandwidth usage and limit the number of packets that have ....

[Article contains additional citation context not shown here]

Richard West, Karsten Schwan and Christian Poell abauer, "Scalable Scheduling Support for Loss and Delay Constrained Media Streams", in Proceedings of the 5th IEEE Real-Time Technology and Applications Symposium (RTAS) , 1999.

....individual messages can be shaped according to the needs of the application. While flexible, using such methods can be ine#cient and does not scale well for high system loads, especially when each stream and individual messages have to be shaped (e.g. scheduled) according to the application s needs [9]. 1.2 Contributions We provided a comprehensive API for creating and attaching socket streams to end system QoS attributes. The API called QSockets is based on the BSD socket API and follows the same semantics. We argue that using this API for end system QoS does not impose any substantial ....

....the e#ciency of system level packet scheduling. Interesting aspects of the GNU Linux QSockets implementation include the following. First, our experience is that kernel level packet scheduling performs at levels of granularity and predictability that far outpaces previous user level approaches [9]. This is particularly important for interactive media applications that have smaller packet sizes, like VOIP. Second, it turned out to be awkward to use the existing GNU Linux QoS infrastructure for the applicationlevel tra#c scheduling required by our multimedia codes. This is because the Linux ....

[Article contains additional citation context not shown here]

R. West, K. Schwan, and C. Poellabauer. Scalable Scheduling Support for Loss and Delay Constrained Media Streams. In Proc. 5th RealTime Technology and Applications Symposium, Vancouver, Canada, 1999.

.... in this paper, with two machines running vic, where the first one is used as a video source (server vic) and the second as a video sink (client vic) Both hosts use a CPU resource manager (managing a round robin scheduler) and the vic server also uses a network resource manager (managing a DWCS [20] packet scheduler) NETWORK INTERFACE NETWORK INTERFACE DATA PATH CONTROL PATH VIC (VIDEO CONFERENCING) VIC (VIDEO CONFERENCING) Manager CPU Manager CPU Network KERNEL KERNEL CPU CPU Figure 3: Experimental setup. 3.2 Communication Setup A user level application initiates a data ....

....component, that is, only quality events and no monitoring events are being issued. Applications are assigned a default priority class (e.g. 50 in the following experiments) and can be modified in the range of 1 and 99. Resource 2: Network. The Dynamic Window Constraint Scheduler or DWCS [20] is a real time scheduler based on three attributes: a period T , a window constraint or loss rate x y, and a run time c, where DWCS guarantees an activity c time units of service within a period T . However, this guarantee is relaxed by the loss rate, which indicates that x service invocations ....

R. West, K. Schwan, and C. Poellabauer. Scalable Scheduling Support for Loss and Delay Constrained Media Streams. In Proc. of the 5th Real-Time Technology and Applications Symposium, Vancouver, Canada, 1999.

....denominator of the current window constraint are decremented and therefore the window constraint is relaxed. On the other hand, if a process misses to be scheduled within its period T, only the numerator is decremented, resulting in a tighter windowconstraint. Details about DWCS can be found in [32, 34]. The original DWCS algorithm works according to the following (simpli ed) rules: The process with the closest deadline (i.e. the time until its current period T expires) will be selected. If several processes have the same deadline, the process with the tightest current window constraint x y ....

R. West, K. Schwan, and C. Poellabauer. Scalable Scheduling Support for Loss and Delay Constrained Media Streams. In Proc. 5th Real-Time Technology and Applications Symposium, Vancouver, Canada, 1999.

No context found.

West, R., Schwan, K., and Poellabauer, C., 1999. Scalable Scheduling Support for Loss and Delay Constrained Media Streams, Proc. IEEE Real-time Technology and Application Symposium (RTAS). 24

No context found.

R. West, K. Schwan, and C. Poellabauer, Scalable Scheduling Support for Loss and Delay Constrained Media Streams, Proc. of IEEE Real-Time Technology and Applications Symposium, 1999

First 50 documents

Online articles have much greater impact   More about CiteSeer.IST   Add search form to your site   Submit documents   Feedback  

CiteSeer.IST - Copyright Penn State and NEC