B. N. Levine, D. Lavo, and J. J. Garcia-Luna-Aceves, "The case for concurrent reliable multicasting using shared ack trees," in Proc. ACM Multimedia, (Boston, MA), Nov. 1996.  Home/Search   Document Not in Database   Summary   Related Articles   Check

....Such a construction interconnects the data delivery tree with some cross edges and is responsible for fast data recovery in PRM under high failure rates of overlay nodes. Existing approaches for resilient and reliable multicast use either reactive retransmissions (e.g. RMTP [18] STORM [20] Lorax [13]) or proactive error correction codes (e.g. Digital Fountain [3] and can only recover from packet losses on the overlay links. Therefore the proactive randomized forwarding is a key difference between our approach and other well known existing approaches. We explain the specific details of ....

....an acknowledgment tree structure with the source as the root. This structure is scalable because the acknowledgments are aggregated along the tree in a bottom up fashion and also allows local recovery and repair of data losses. Protocols like RMTP [18] TMTP [23] STORM [20] LVMR [14] and Lorax [13] construct this structure using TTL scoped network layer multicast as a primitive. In contrast, LMS [17] uses an additional mechanism, called directed subcast, to construct its data recovery structure. Our work differs from of all these above approaches in two key aspects. First, unlike all these ....

[Article contains additional citation context not shown here]

B. Levine, D. Lavo, and J. Garcia-Luna-Aceves. The case for concurrent reliable multicasting using shared ack trees. In Proc. ACM Multimedia, Nov. 1996.

....treatment of feedback implosion to an accompanying protocol. Certain reliable multicast protocols use tree based approaches for dealing with feedback implosion [11] e.g. using methods for feedback fusion or feedback aggregation in hierarchies leading back to the source, for instance, ACK trees [18]) However, these methods increase feedback latency and recovery latency, since they require several hop by hop processing and aggregation of messages. In order to avoid feedback implosion resulting from spatially correlated losses, methods of probabilistic feedback suppression may be used to ....

B.N. Levine, David Lavo, and J.J. Garcia-Luna-Aceves, The Case for Concurrent Reliable Multicasting Using Shared Ack Trees, Proc. ACM Multimedia 1996.

.... to avoid multiple retransmissions from several nodes [13] Certain reliable multicast protocols use tree based approaches for dealing with feedback implosion [11] e.g. using methods for feedback fusion or feedback aggregation in hierarchies leading back to the source, for instance, ACK trees [18]) However, these methods increase feedback latency and recovery latency, since they require several hop by hop processing and aggregation of messages. In order to avoid feedback implosion resulting from spatially correlated losses, methods of probabilistic feedback suppression may be used to ....

B.N. Levine, David Lavo, and J.J. Garcia-Luna-Aceves, The Case for Concurrent Reliable Multicasting Using Shared Ack Trees, Proc. ACM Multimedia 1996.

....are spatially correlated. However, these techniques increase the expected delay and also fail to lower the worst case bounds. Tree based solutions that aggregate feedback from receivers as the feedback moves up the multicast tree can alleviate the feedback problem altogether (e.g. see [LPGLA98,LLGLA96] However, these solutions impose additional processing and state at intermediate nodes within the tree. By combining a tree based solution with proactive FEC techniques can reduce the expected amounts of processing and state that these intermediate nodes incur. Extensive simulation over a ....

B. Levine, D. Lavo, and J.J. Garcia-Luna-Aceves. The Case for Concurrent Reliable Multicasting Using Shared Ack Trees. In Proceedings of ACM Multimedia'96, Boston, MA, November 1996. 23

....in [5] to demonstrate the performance benefits of including waypoints: servers inside the network dedicated to improving protocol performance. In the application presented in that paper, waypoints act as repair servers, and reduce the repairing load on receivers in the session. A variety of works [12, 13, 14, 1, 5] present different algorithms for generating the repair hierarchy that can be used to provide reliable multicast. All of these approaches build their repair graphs dynamically using distributed algorithms. A dynamic algorithm to build the repair graph can customize the graph precisely to the ....

B. Levine, D. Lavo, and J.J. Garcia-Luna-Aceves. The Case for Concurrent Reliable Multicasting Using Shared Ack Trees. In Proceedings of ACM Multimedia'96, Boston, MA, November 1996.

.... early work has focused on distributed systems, providing primitives for constructing distributed applications, such as the ISIS system[30] and the V kernel[31] Other early work has focused on local area networks or broadcast links [32, 33, 34, 35] A good survey of the early work can be found in [27]. Here we focus on recent work on reliable multicast that aims to provide scalability to very large groups. The vast majority of recent reliable multicast protocols use receiver reliable recovery, shown by Pingali, Towsley, and Kurose to be superior to sender reliable recovery [11] We divide ....

B.N. Levine, David Lavo, and J.J. Garcia-Luna-Aceves, "The Case for Concurrent Reliable Multicasting Using Shared Ack Trees," Proc. ACM Multimedia

....loss rates on the links in this topology. What we do not provide (are unable to) is an algorithm for identifying the physical topology of a network. A more important motivation for this work is that knowledge of the multicast topology can be used by multicast applications. It has been shown in [9] that organizing a set of receivers in a bulk transfer application into a tree can substantially improve performance. Such an organization is central component of the widely used RMTP II protocol [20] The development of tree construction algorithms for the purpose of supporting reliable multicast ....

B.N. Levine, David Lavo , and J.J. Garcia-Luna-Aceves, "The Case for Concurrent Reliable Multicasting Using Shared Ack Trees," Proc. ACM Multimedia Boston, MA, November 18--22, 1996.

....require reliable unicast communication. TCP was designed to be a generic transport protocol for reliable unicast communication: it provides a simple ordered reliable data stream. Although several researchers have proposed similar generic transport protocols for reliable multicast communication [52, 35, 31], Floyd et al. 17] demonstrate how di erent multicast applications have widely di erent requirements for reliability, and hence propose a scheme for reliable multicast that can take into account these diverse application needs. Some applications may require ordered delivery while other may not ....

.... range of reliability mechanisms and each such region should be optimized by locally deploying the most suitable protocol, e.g. hop wise ARQ might be appropriate to e ectively accommodate the high loss rates of a series of radio links, while SRM [17] works well in a high bandwidth LAN, and Lorax [31] is better for a wide area topology arranged as a tree. To this end, the RMX framework allows us to seamlessly integrate a diverse set of protocols running across a disjoint set of network clouds. RMX supports two di erent variants of protocol conversion: transport level conversion and ....

Levine, B., Lavo, D., and Garcia-Luna-Aceves, J. The Case for Concurrent Reliable Multicasting Using Shared Ack Trees. In Proceedings of ACM Multimedia '96 (Boston, MA, Nov. 1996).

....multicast routing has yet to materialize and a reliable multicast transport protocol that offers congestion control and robust and scalable behavior remains a research problem. The failure of many and varied research efforts to bear truly viable end to end multicast transport protocols [10, 33, 17] or truly viable wide area, interdomain multicast routing protocols [9, 1, 15] brings into question whether the proposed multicast service model is in fact the appropriate core building block. None of the proposals for reliable multicast have satisfied the requirements for safe deployment on the ....

B. Levine, D. B. Lavo, and J. J. Garcia-Luna-Aceves. The Case for Concurrent Reliable Multicasting Using Shared Ack Trees. In Proceedings of ACM Multimedia, Boston, MA, Nov. 1996. ACM.

....are comparatively well understood and have proven extremely successful (e.g. TCP) the problem of defining a general purpose reliable multicast transport protocol has remained a formidable challenge for several years. Though a number of wide area reliable multicast protocols have been proposed [33, 47, 21, 32], none currently satisfies the requirements for safe deployment in the global Internet because no single existing scheme simultaneously provides (1) scalability to very large multicast sessions, 2) congestion control that accommodates constrained and potentially heterogeneous link bandwidths, ....

LEVINE, B., LAVO, D. B., AND GARCIA-LUNA-ACEVES, J. J. The Case for Concurrent Reliable Multicasting Using Shared Ack Trees. In Proceedings of ACM Multimedia (Boston, MA, Nov. 1996), ACM.

.... distributed systems, providing primitives for constructing distributed applications, such as the ISIS system[67] and the V kernel[68] Other early work has focused on local area networks or broadcast links [69, 70, 71, 72, 73] We will not cover the early work here; a good survey can be found in [64]. We will focus on recent work on reliable multicast that aims to provide scalability to very large groups. As we described earlier, the vast majority of reliable multicast protocols use receiver reliable recovery which was shown by Pingali, Towsley, and Kurose to be superior to sender reliable ....

B.N. Levine, David Lavo , and J.J. Garcia-Luna-Aceves, "The Case for Concurrent Reliable Multicasting Using Shared Ack Trees,"Proc. ACM Multimedia 1996 Boston, MA, November 18--22, 1996.

....transmission. Narada has striking differences from self configuring protocols developed in other contexts. First, Narada distinguishes itself from normal routing protocols in that it changes the very topology over which routing is performed. Second, most existing self configuring protocols [12, 13, 14, 22] assume native IP Multicast support. Narada attempts selfconfiguration in the absence of a lower level multicast service, and this is fundamentally more challenging. We explain the distributed algorithms that Narada uses to construct and maintain the mesh in Section 3.1. We present heuristics ....

....links to ensure connectivity of the virtual topology, and drop links it perceives as not useful. Self configuration has been proposed in other contexts. AMRoute [1] allows for robust IP Multicast in mobile adhoc networks by exploiting user multicast trees. Several reliable IP Multicast protocols [12, 13, 22] involve group members self organizing into structures that help in data recovery. Adaptive Web Caching [14] is a self organizing cache hierarchy. The key feature that distinguishes Narada from these protocols is that Narada does not assume a native multicast medium AMRoute assumes a native ....

B. N. Levine, D. B. Lavo, and J. J. Garcia-Luna-Aceves. The case for concurrent reliable multicasting using shared ACK trees. In Proceedings of ACM Multimedia'96, November 1996.

....if separated from the centralized decision point. In particular, we use a centralized algorithm to determine from which repair servers a given repair server should request repairs. Other works use distributed algorithms that use various probing techniques to identify appropriate repair servers [2, 3, 5, 6, 7], which complicates their deployment. 2. The complexity of the protocol at the repair servers is kept simple, and the amount of state that a repair server must maintain is kept low. 3. The system utilizes a simple model of the underlying network. A system designed to utilize a complex network ....

....the construction of the ring. 3.1.2 Two Trees A good repair graph efficiently distributes repairs from servers that are more likely to receive data in the initial transmission to those servers that are less likely to receive the data. Most often, a tree is constructed for this purpose [3, 4, 5, 7], since it is the most efficient means of producing a fully connected graph (a path exists from the root of the tree to each node) However, a node failure, unless lying at a leaf of the tree, would partition the tree and prevent propagation of repairs from nodes on one side of the partition to ....

[Article contains additional citation context not shown here]

B. Levine, D. Lavo, and J.J. Garcia-Luna-Aceves. The Case for Concurrent Reliable Multicasting Using Shared Ack Trees. In Proceedings of ACM Multimedia'96, Boston, MA, November 1996.

....in [6] to demonstrate the performance bene ts of including waypoints: servers inside the network dedicated to improving protocol performance. In the application presented in that paper, waypoints act as repair servers, and reduce the repairing load on receivers in the session. A variety of works [12, 13, 14, 2, 6] present di erent algorithms for generating the repair hierarchy that can be used to provide reliable multicast. All of these approaches build their repair graphs dynamically using distributed algorithms. A dynamic algorithm to build the repair graph can customize the graph precisely to the ....

B. Levine, D. Lavo, and J.J. Garcia-Luna-Aceves. The Case for Concurrent Reliable Multicasting Using Shared Ack Trees. In Proceedings of ACM Multimedia'96, Boston, MA, November 1996.

....of loss and therefore do not require RM. However, fast recovery of key segments (e.g. the I frames) in the presence of a replay buffer may still be preferred. niques addressing these problem typically trade off between latency and bandwidth. For example, delaying or restricting retransmissions [11, 15, 16] can reduce the recovery traffic but also increase the recovery latency. Forward Error Correction (FEC) 2] is an appealing approach to avoid this feedback implosion. Several FEC based protocols require no receiver feedback (RMDP [6] Digital Fountain [3] and Fcast [8] while some send ....

....if an identical one arrives. The delay is a randomized function of the round trip time (RTT) between the data source and the sender of the NAK (or repair) This suppression process is referred to as duplicate avoidance (DA) Another class of RM protocols employs tree based local recovery [15, 16, 18, 33]. A subgroup hierarchy is constructed among multicast receivers and rooted at the source. Members close to each other form a subgroup and elect a designated receiver (DR) for the subgroup. A DR may be a member of a higher level subgroup. Only DRs retransmit NAKs and repairs are forwarded only ....

[Article contains additional citation context not shown here]

B. Levine, D. Lavo, and J.J. Garcia-Luna-Aceves, "The Case for Concurrent Reliable Multicasting Using Shared Ack Trees", Proc. ACM Multimedia'96 Conference.

No context found.

B. N. Levine, D. Lavo, and J. J. Garcia-Luna-Aceves, "The case for concurrent reliable multicasting using shared ack trees," in Proc. ACM Multimedia, (Boston, MA), Nov. 1996.

No context found.

B. N. Levine, D. Lavo, and J. J. Garcia-Luna-Aceves. The case for concurrent reliable multicasting using shared ack trees. In Proc. ACM Multimedia, Boston, MA, Nov. 1996.

No context found.

B. N. Levine, D. Lavo, and J. J. Garcia-Luna-Aceves. The case for concurrent reliable multicasting using shared ack trees. In Proc. ACM Multimedia, Boston, MA, Nov. 1996.

....correlation of the control geometry with the actual underlying multicast routing tree, or the endto end reliable multicast tree. Hierarchical organization of stations also supports inter group collaboration, subgroup addressing, and allows more scalable and economic transmission of session data [21]. A tree based group coordination protocol representing STT is discussed in Section 5. Control messages in a STT protocol are passed along branches of the tree in a parent child relation reflecting multicast group membership. No specific node alone is hence burdened with the obligation to make ....

....FLOOR CONTROL 5.1. HGCP Description We outline the operation of the Hierarchical Group Coordination Protocol (HGCP) as an example for STT protocols. Although tree protocols have been the subject of related research, e.g. on mutual exclusion [28] channel access [5] or reliable multicast [21,26,42], to our knowledge no tree based group coordination protocol has been proposed in the literature. HGCP represents two innovations: floor control is inherently hierarchical, and the control tree for group coordination is correlated in its operation to an underlying tree based multicast service. ....

[Article contains additional citation context not shown here]

B. N. Levine, D. Lavo, and J. J. Garcia-Luna-Aceves. The case for concurrent reliable multicasting using shared ack trees. In Proc. ACM Multimedia, Boston, MA, Nov. 1996.

....schemes are also analyzed which shows that compared with RNKA, DNKA reduces the retransmission delay significantly. 1. Introduction The reliable multicast protocols proposed to date can be classified as: sender initiated protocols [3] receiver initiated protocols [3,12] tree based protocols [2,6,7,8,9,13], ring based broadcast protocols [4,11,18] ring based store forward protocols [5] ring media based protocols [17] and sequencer based protocols [16] Sender initiated protocols require the source to receive positive acknowledgments (ACKs) from all receivers, before it is allowed to release ....

....( ACK tree ) The source or an intermediate node in the ACK tree ( which we call hop ) is responsible for reliable packet delivery only for its immediate children in the tree. High throughput and scalibility can be achieved by limiting the number of children at each hop node and the source [2, 3]; the structure of the tree needs to be updated when a hop has too many children. From the above, it is clear that current proposals for reliable multicasting have limitations. This paper proposes new structures for reliable multicasting and is organized as follows. Section 2 introduces a ....

[Article contains additional citation context not shown here]

B.N. Levine, D.N. Lavo, and J.J. Garcia-Luna-Aceves, "The Case for Concurrent Reliable Multicasting Using Shared Ack Trees," Proc. ACM Multimedia'96, Boston, MA, Nov. 1996.

No context found.

B. N. Levine, D. Lavo, and J. J. Garcia-Luna-Aceves, "The case for concurrent reliable multicasting using shared ack trees," in Proc. ACM Multimedia, Boston, MA, Nov. 1996, pp. 365--376.

No context found.

B. Levine, D. B. Lavo, and J. J. Garcia-Luna-Aceves, "The Case for Concurrent Reliable Multicasting Using Shared Ack Trees," in Proceedings of ACM Multimedia, Boston, MA, Nov. 1996, ACM.

No context found.

B. Levine, D. Lavo, and J. Garcia-Luna-Aceves. The case for concurrent reliable multicasting using shared ack trees. In Proc. ACM Multimedia, Nov. 1996.

No context found.

B. Levine, D. Lavo, and J. Garcia-Luna-Aceves. The case for concurrent reliable multicasting using shared ack trees. In Proc. ACM Multimedia, Nov. 1996.

No context found.

B. Levine, D. Lavo, and J. Garcia-Luna-Aceves. The case for concurrent reliable multicasting using shared ack trees. In Proc. ACM Multimedia, Nov. 1996.

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