S. Pejhan, M. Schwartz, and D. Anastassiou. Error control using retransmission schemes in multicast transport. IEEE/ACM Transactions on Networking, 4(3):413-27, June 1996.  Home/Search   Document Details and Download   Summary   Related Articles   Check

....e#ectively address bandwidth heterogeneity. On the error control side, a large body of work has focused on providing partial or full reliability to all receivers using retransmissions [44, 97] or type II hybridARQ [147] schemes employing both error control codes across packets and retransmissions [90, 98, 67, 109]. Since source based retransmission schemes are well known to be non scalable to large groups of receivers, elaborate schemes have been employed to organize the receivers into repair trees to e#ect hierarchical retransmission [97, 67, 150] or local groups to facilitate local repair [44] However, ....

....Furthermore, FEC packets are only transmitted to branches that lead to subscribers requiring FEC, thereby reducing overall network load. In contrast, traditional FEC schemes for multicast often compute a set of FEC packets which are multicast to all receivers regardless of whether they are needed [90, 98]. Given a menu of data and FEC layers, each receiver in a layered FEC implementation needs to perform two basic functions. First, it needs to determine the amount of available bandwidth; this is discussed in Section 5.6. Second, it needs to choose from the menu the optimal set of data and FEC ....

S. Pejhan, M. Schwartz, and D. Anastassiou. Error control using retransmission schemes in multicast transport protocols for real-time media. ACM/IEEE Trans. Networking, 4(3), pp. 413--27, June 1996. 176

....protocols with NS 2 simulation studies of SRM [2] RMTP [3] and TMTP [4] variations. The packet delivery delay is an important issue for multimedia applications. For example real time applications like interactive distributed simulations, distributed games, or the delivery of MPEG I frames [5] benefit from guaranteed reliability and low delays. Besides analyzing the delay between sender and receiver we determine the round trip delay between sending a data packet and receiving the last corresponding control packet at the sender. The round trip delay determines the time after a data ....

Pejhan, S., Schwartz, M., Anastassiou, D.: Error control using retransmission schemes in multicast transport protocols for real-time media. IEEE/ACM Transactions on Networking 4 (1996) 413--427

....deal with the worstcase packet loss rate scenario. This leads to inefficiencies. The overhead for error correction also increases total network load. Thus the entire network is taxed due to the worse performing route [23, 12] The alternative is to use a closed loop approach. Close loop approaches [25, 22, 7, 29] , where the receivers request the retransmission of lost packets, have the drawback of higher latency and are difficult to scale [6, 4] Additionally, since packet loses generally occur during congestion, these requests and subsequent retransmissions can make matters worse. The algorithm we ....

Sassan Pejhan, Mischa Schwartz, and Dimitris Anastassiou. Error control using retransmission schemes in multicast transport protocols for realtime media. IEEE/ACM Transactions on Networking, 4(3):413--427, June 1996. 11

....full TG is received, just as with protocol 1. IV. BANDWIDTH ANALYSIS Table I summarizes the variables and notation that will be used for the bandwidth analysis. We define the bandwidth B as the bandwidth consumed by a multicast packet per link, averaged over all links in the multicast tree [20]. The bandwidth of a multicast packet in a multicast group i is the product of the number M i of transmissions per packet (original and retransmissions) and the number H i of links traversed. Given H = R w b Delta G 1 links in total, where w b is the number of physical hops in a backbone link as ....

Sassan Pejhan, Mischa Schwartz, and Dimitris Anastassiou, "Error control using retransmission schemes in multicast transport protocols for real-time media," IEEE/ACM Transactions on Networking, vol. 4(3), pp. 413--427, June 1996.

....wasted bandwidth, and high processing requirements. Feedback implosion imposes high requirements to the mechanism for feedback implosion avoidance. Several solutions exist for implosion avoidance based on hierarchies, timers, tokens, and probing, see section 7 on related work. Very little work [2, 3, 4] was done on the analysis of timer based schemes for multicast feedback. We give an analytical foundation of timer based feedback where the timer choice, the sender receiver delays, and the delays between receivers can be modeled by arbitrary distributions. The analysis allows to compute: ffl The ....

....of subgroups and can not be employed in a scenario like satellite distribution with unicast backward channels. Approaches based on MAC protocols suffer from scalability problems. Tokens lead to high feedback latencies and random timers in [3, 4] are based on a uniform distribution. The analysis in [2] compares multicast feedback with random uniform timers to unicast feedback with respect to the cost in terms of number of control packets per link. The authors conclude that unicast control packets outperform multicast control packets for a small number of receivers. SRM [3] exploits ....

Sassan Pejhan, Mischa Schwartz, and Dimitris Anastassiou, "Error control using retransmission schemes in multicast transport protocols for realtime media," IEEE/ACM Transactions on Networking, vol. 4(3), pp. 413--427, June 1996.

....modified reformation process is described in section 5.3. Finally, the messages that are lost by different receivers on a multicast tree are correlated. When a message is lost or corrupted on a link of the tree, all of the downstream receivers lose the message and there may be a NACK implosion [13 , 8 , 14 , 15]. NACK implosion is reduced in our multiple loop architecture by having multiple token sites and limiting the number of receivers that recover messages from any particular site, as described in section 4. In section 5.2 we show how to further reduce the number of NACK s in RMP. 4. ....

S. Pejhan, M. Schwartz, D. Anastassiou, "Error Control Using Retransmission Schemes in Multicast Transport Protocols for Real-Time Media", IEEE/ACM Trans. on Networking, vo. 4, no. 3, June 1996, pp. 413-427.

....of problems. Unicast 18 retransmissions trickle down the tree from retransmitter to retransmitter (e.g. the STORM protocol unicasts retransmissions despite its emphasis on deadline oriented data) multicast retransmissions have been shown to lower average packet delay for reliable transmission [43]. Furthermore, unicast retransmissions are a waste of bandwidth; sending multiple copies of the same data is inefficient, especially since a multicast routing tree already connects the receivers. To make up for the delay from unicast retransmission, STORM allows nodes to change parents if the ....

....if an application does not require 100 of the data, such as in audio or video services where some loss is tolerable, the emphasis of a reliable protocol may be placed on getting as much of the data as possible before a deadline passes. Multicast retransmissions can lower average packet delay [43], which is desirable for the real time streaming applications that resilient multicast protocols are designed to support. 2 RMTP does not specify aggregate Acks explicitly, but the Reliable Multicast File Transfer Protocol (RMFTP) an application based on RMTP, does already use aggregate Acks. ....

S. Pejhan, M. Schwartz, and Anastassiou D. Error control using retransmission schemes in multicast transport. IEEE/ACM Transactions on Networking, 4(3):413--27, June 1996.

....In this paper we present a delay analysis of tree based reliable multicast protocols. The message delivery delay is an important issue for multimedia applications. For example, real time applications like interactive distributed simulations, distributed games, or the delivery of MPEG I frames [1] benefit from guaranteed reliability and low delays. Besides time constraints of some applications, low delays are vital for providing high throughput with a window based sending scheme [2] In contrast to previous delay analysis we assume a more realistic system model as explained later in ....

S. Pejhan, M. Schwartz, and D. Anastassiou, "Error control using retransmission schemes in multicast transport protocols for real-time media," IEEE/ACM Transactions on Networking, vol. 4, no. 3, pp. 413--427, 1996.

....our analytical model. I. INTRODUCTION If a reliable multicast protocol is used for real time applications, the resulting message delivery delay is an important issue. For example real time applications like interactive distributed simulations, distributed games, or the delivery of MPEG I frames [1] benefit from guaranteed reliability. Besides time constraints of some applications, low delays are vital for providing high throughput with a window based sending scheme [2] Known multicast protocols try to achieve reliable data delivery by forward error correction or retransmission schemes or ....

S. Pejhan, M. Schwartz, and D. Anastassiou, "Error control using retransmission schemes in multicast transport protocols for real-time media," IEEE/ACM Transactions on Networking, vol. 4, no. 3, pp. 413--427, 1996.

.... the first multicast group, the receivers are downstream from the source router via different links, and so no link savings are realizable (k # 1) On the other hand, the receivers in the second multicast group lie in a common distribution path, and significant link savings 1 As pointed out by [19], results based on hop based metrics are generalizable to both source based shortestpath trees and minimal spanning trees. Our results will not be significantly different even if we adopt a hop distance hybrid metric (using a rule of thumb [16] that 100 kilometers of link distance have equivalent ....

S. Pejhan, M. Schwartz and D. Anastassiou, Error control using retransmission schemes in multicast transport protocols for real-time media, IEEE/ACM Transactions on Networking 4(3) (1996) 413-- 427.

....e#ectively address bandwidth heterogeneity. On the error control side, a large body of work has focused on providing partial or full reliability to all receivers using retransmissions [20, 21] or type II hybrid ARQ [22] schemes employing both error control codes across packets and retransmissions [23, 24, 25, 26]. Since source based retransmission schemes are well known to be non scalable to large groups of receivers, elaborate schemes have been employed to organize the receivers into repair trees to e#ect hierarchical retransmission [21, 25, 27] or local groups to facilitate local repair [20] However, ....

....Furthermore, FEC packets are only transmitted to branches that lead to subscribers requiring FEC, thereby reducing overall network load. In contrast, traditional FEC schemes for multicast often compute a set of FEC packets which are multicast to all receivers regardless of whether they are needed [23, 24]. Given a menu of data and FEC layers, each receiver in a layered FEC implementation needs to 6 perform two basic functions. First, it needs to determine the amount of available bandwidth; this is discussed in Section 5. Second, it needs to choose from the menu the optimal set of data and FEC ....

S. Pejhan, M. Schwartz, and D. Anastassiou, "Error control using retransmission schemes in multicast transport protocols for real-time media," ACM/IEEE Trans. Networking, vol. 4, no. 3, pp. 413--27, June 1996.

....In this paper we present a delay analysis of tree based reliable multicast protocols. The message delivery delay is an important issue for multimedia applications. For example real time applications like interactive distributed simulations, distributed games, or the delivery of MPEG Iframes [1] benefit from guaranteed reliability and low delays. Besides time constraints of some applications, low delays are vital for providing high throughput with a window based sending scheme [2] In contrast to previous delay analysis we assume a more realistic system model as explained later in ....

S. Pejhan, M. Schwartz, and D. Anastassiou, "Error control using retransmission schemes in multicast transport protocols for realtime media," IEEE/ACM Transactions on Networking, vol. 4, no. 3, pp. 413--427, 1996.

....that make it possible to retransmit lost packets in some applications. For instance, a few seconds of delay are not noticeable in a real time news report, but would be intolerable in an interactive voice conversation. Several techniques have been proposed to recover lost packets in video systems [3,4,5]. In schemes that require retransmissions there must be sufficient storage at receivers to store the signal while waiting to recover missing packets. In video applications storage has been an important concern, however, as explained in section 2.4, storage is not a major concern in MBone video. ....

....the link costs associated with multicast and point to point transmissions, then select the appropriate mode. However, most reliable multicast mechanisms, such as the white board[8] multicast retransmissions, and certain analytical results indicate that all retransmissions should be multicast[5]. As a practical matter, when packets are lost in the network, they are normally lost by several sites, making it likely that multicast is the least costly way to supply retransmissions. The main reason for not multicasting retransmissions to very large groups is that it is very likely that some ....

S. Pejhan, M. Schwartz, D. Anastassiou, "Error Control Using Retransmission Schemes in Multicast Transport Protocols for Real-Time Media", IEEE/ACM Trans. on Networking, vo. 4, no. 3, June 1996, pp. 413-427.

....ARQ, the sender explicitly retransmits the packets that are requested by the receiver. Although this introduces a delay of at least three one way trip times, previous research has demonstrated the effectiveness of ARQ for the transmission of real time media, especially in the case of multicasting [20]. For our application, ARQ has proven quite effective provided that network capacity was not saturated. Although we have so far demonstrated unicast alone, our system can be extended easily to support multicast, permitting its application to the simultaneous distribution of a performance to ....

S. Pejhan, M. Schwartz, D. Anastassiou. Error control using retransmission schemes in multicast transport protocols for real-time media. IEEE/ACM Transactions on Networking, Vol.4, No.3, June, 1996.

....Error control algorithms are necessary to ensure reliable data transmission. Three techniques are used for error control : 1) automatic repeat request (ARQ) which uses acknowledgments, time outs, and retransmissions, 2) forward error correction (FEC) and 3) error concealment (EC) at the receiver [1]. For continuous media that presents real time constraints error recovery through retransmission is usually not appropriate. Forward error correction schemes and special encoding techniques are proposed to enhance reliability for audio and video [2, 3, 4, 5] For discrete media, lost or corrupted ....

....this technique requires a multicast tree for each receiver if the underlying multicast routing mechanism uses source based trees. In applications, such as video on demand, where the receivers interact rarely with the transmitter this would place an extra burden on the network routers. Reference [1] provides both a qualitative and a quantitative analysis of the damping and slotting technique. Pejhan et al. show that this scheme can be counterproductive in some cases. They also conclude that the benefits of this technique are more significant for schemes using positive acknowledgments. The ....

S. Pejhan, M. Schwartz, and D. Anastassiou, "Error Control Using Retransmission Schemes in Multicast Transport Protocols for Real-Time Media," IEEE/ACM Transactions on Networking, vol. 4, pp. 413--427, June 1996.

....with the worst case packet loss rate scenario. This leads to inefficiencies. The overhead for error correction also increases total network load. Thus the entire network is taxed due to the worse performing route [26, 12] The alternative is to use a closed loop approach. Close loop approaches [28, 25, 7, 33] , where the receivers request the retransmission of lost packets, have the drawback of higher latency and are difficult to scale [6, 4] Additionally, since packet loses generally occur during congestion, these requests and subsequent retransmissions can make matters worse. Robustness to data ....

Sassan Pejhan, Mischa Schwartz, and Dimitris Anastassiou. Error control using retransmission schemes in multicast transport protocols for real-time media. IEEE/ACM Transactions on Networking, 4(3):413--427, June 1996.

....suffers from scalability problems because the sender is involved in the adaptation. 6.1. 3 Retransmission Based Reliability Proposals exist for integrating reliable multicast schemes into audio and video applications so that missing packets can be recovered from neighbours with better reception [Pejhan et al. 1996; Maxemchuk et al. 1997; Xu et al. 1997] This is achieved by trading off quality for delay, as any reliable multicast protocol has to request retransmission and wait for the repair. Although this may be acceptable in a real time lecturing scenario, it becomes less useful with interactive ....

S. Pejhan, M. Schwartz, and D. Anastassiou. Error control using retransmission schemes in multicast transport protocols for real-time media. IEEE/ACM Transcations on Networking, 4(3):413--327, June 1996.

....information of is shown. mation that the sender believes has been lost; 4 as a result they do not unnecessarily waste bandwidth when there is no packet loss, and they can easily adapt to changes in the loss rates. Thus studies have examined soft ARQ for both unicast [20, 11] and multicast [21, 31, 30] streaming multimedia. One way our work differs from all of these is that we assume there is an overall transmission rate limit, so that a retransmission of one message can come at the expense of the first transmission of another; these other works assume that enough bandwidth is available for any ....

....for a high probability of successful retransmissions of single packet losses. Unlike our work, the authors do not consider multiple retransmissions of frames that are lost multiple times. A general examination of NACK based retransmission based schemes for multicast real time media is given in [21]. The authors present an analysis which indicates that not only are retransmissions both useful and practical for real time media, but in many situations it is optimal for the source to immediately multicast a retransmission upon the reception of a NACK from any receiver. Although the average ....

S. Pejhan, M. Schwartz, and D. Anastassiou. Error control using retransmission schemes in multicast transport protocols for real-time media. IEEE/ACM Trans. Networking., 4(3):413--427, June 1996.

No context found.

S. Pejhan, M. Schwartz, and D. Anastassiou. Error control using retransmission schemes in multicast transport. IEEE/ACM Transactions on Networking, 4(3):413-27, June 1996.

No context found.

S. Pejhan, M. Schwartz, and D. Anastassiou. Error Control using Retransmission Schemes in Multicast Transport Protocols for Real-time Media. IEEE/ACM Transactions on Networking, 4(3):413--427, 1996.

No context found.

S. Pejhan, M. Schwartz, D. Anastassiou, "Error Control Using Retransmission Schemes in Multicast Transport Protocols for Real Time Media", in IEEE/ACM Transactions on Networking, Vol. 4, pp. 413-427, June 1996.

No context found.

#PSA96# S. Pejhan, M. Schwartz, and D. Anastassiou, #Error Control Using Retransmission Schemes in Multicast Transport Protocols for Real-Time Media", IEEE#ACM Transactions on Networking, vol. 4 no.

No context found.

S. Pejhan, M. Schwartz, D. Anastassiou, "Error Control Using Retransmission Schemes in Multicast Transport Protocols for Real-Time Media", submitted IEEE/ACM Transactions on Networking.

No context found.

S. Pejhan, M. Schwartz and D. Anastassious, Error Control Using Retransmission Schemes in Multicast Transport Protocols for Real-Time Media , IEEE/ACM Transactions on networking. Vol. 4, No. 3, June 1996, pp. 413427.

No context found.

S. Pejhan, M. Schwartz, D. Anastassiou, "Error Control Using Retransmission Schemes in Multicast Transport Protocols for Real-Time Media", IEEE/ACM Trans. on Networking, vo. 4, no. 3, June 1996, pp. 413-427.

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