D. Rubenstein, S. Kasera, D. Towsley, and J. Kurose. Improving Reliable Multicast Using Active Parity Encoding Services (apes). Technical Report TR98-79, Computer Science Department, University of Massachusetts at Amherst, July 1998.  Home/Search   Document Details and Download   Summary   Related Articles   Check

....Taiwan, R.O.C. under Contract NSC 89 2213E007 127 and the program for promoting academic excellence of universities 89 E FA04 1 4. I. Introduction Using FEC (Forward Error Correction) for multicasting in the Internet has received a lot of attention recently (see e.g. 15] 4] 16] 20] [18], 12] Several important observations were made in the literature based on numerical comparisons of various protocol schemes that use FEC. For instance, it is noted in [15] that multicasting protocols with FEC scale well for a broad range of receivers and that the amount of redundancy should be ....

....of various protocol schemes that use FEC. For instance, it is noted in [15] that multicasting protocols with FEC scale well for a broad range of receivers and that the amount of redundancy should be matched to the size of data in order to obtain high throughput. Also, it is observed in [18] that the throughput for a multicasting problem with heterogeneous groups of receivers is dominated by the group with high loss. Motivated by providing rigorous justi cation and insightful explanations for these important observations, we take a theoretical approach to study a basic multicasting ....

[Article contains additional citation context not shown here]

D. Rubenstein, S. Kasera, D. Towsley and J. Kurose, \Improving reliable multicast using active parity encoding services (APES)," Proc. IEEE INFOCOM'99, pp. 1248-1255, New York, U.S.A., March 1999.

....poorly when the loss occurs at the source link. FEC is an appealing approach to avoid feedbe k implosion. Several FEC bOW( protocols require no receiver feedbL ks [14] 15] FEC has alsob een integrated withfeedbW k bHW reliab( multicast protocols, called hyb L ARQ [8] 9] 11] [13]. In reactive use of FEC [8] retransmissions are parity packets instead of individual lost data packets. For repair request, a NAK that indicates the numb er of losses not the sequence numb ers of lost packets is sentb y a receivers. This approach complicates the reliab# multicast ....

....packets. Inab ove allFEC b35( approaches, the network b(#HBO(4 is wastedb y redundant parity packets multicastb ecause FEC is applied to end to end, i.e. same amount of redundant parity packets are transmitted to all receivers that su#er heterogeneous packet loss prob533(4B# 5 APES [12] [13], achieves e#ciency in terms of networkb(B1 L5( due to additional network support called repair server(RS) is another reactive FEC approach. APES sends parity packets in place of retransmissions. Receivers reconstruct original data packets from received packets. RS ensures that each of its ....

[Article contains additional citation context not shown here]

D. Rubenstein, S. Kasera, D. Towsley, and J. Kurose, "Improving reliable multicast using active parity encoding

....more packets. Inab ove allFEC b35( approaches, the network b(#HBO(4 is wastedb y redundant parity packets multicastb ecause FEC is applied to end to end, i.e. same amount of redundant parity packets are transmitted to all receivers that su#er heterogeneous packet loss prob533(4B# 5 APES [12], 13] achieves e#ciency in terms of networkb(B1 L5( due to additional network support called repair server(RS) is another reactive FEC approach. APES sends parity packets in place of retransmissions. Receivers reconstruct original data packets from received packets. RS ensures that each of ....

....Source link loss which is likely to causefeedb5 k implosion, is one of the serious technical probWL1 forscalabH reliabi multicast. One promising way to overcome feedbW k implosion is FEC. Manyreliab( multicast protocolsb ased on merging FEC and ARQ techniques haveb een proposed [9] [12]. The main ideab ehind these hyb# ARQ FEC approaches is that the sender encodes data and transmits the original data along with some redundant data. If a receiver detects losses which cannotb e recovered from the data received from the sender, then the receiver requests the numb er of lost ....

D. Rubenstein, S. Kasera, D. Towsley, and J. Kurose, "Improving reliable multicast using active parity encoding services (APES)," Proc. IEEE INFOCOM'99, New York, April 1999.

....(both featuring parity transmission recovery) It is shown that for a fixed network topology with 113 receivers DER has superior bandwidth performance compared to CER. Latency issues as well as the influence of network and transmission mode parameters on all results are not considered in [10] In [13] CER and DER protocols with optimizations have been compared regarding buffer size requirements and bandwidth. The results presented there underline our results. However, latency is not considered. CER protocols are attractive since they are easier to deploy than DER protocols and require less ....

D. Rubenstein and D. Towsley, "Improving reliable multicast using active parity encoding services," in INFOCOM 99, March 1999.

....blocks. For each block, h FEC encoded packets are generated for suitably chosen h. Receivers can recover the original block by receiving any B packets out of the B h ones. When combined with an appropriate ARQ technique, the FEC technique incurs very low network overhead [14] 15] 12] 21] [20], 11] The FEC approach can be employed in any of the protocols we have described so far such as SRM, tree based protocols etc. A noteworthy example is the SHARQFEC protocol [11] that combines hierarchical scoping and hybrid FEC ARQ. It breaks the entire multicast group into hierarchically ....

....We study the scalability of reliable multicast protocols. Some of the fundamental known techniques for devising scalable reliable multicast protocols include receiver centric repair [5] scoping [11] 6] suppressing unwanted traffic by using timers [5] use of FEC packets [14] 7] 11] 21] [20] etc. In this paper, we offer an additional tool, namely, an efficient protocol (LMR) to proactively transmit FEC repair packets in layers with minimal redundant repair traffic. We presented simulation results that show that using our LMR protocol with known reactive multicast protocols makes them ....

D. Rubenstein, S. Kasera, J. Kurose, and D. Towsley. Improving reliable multicast using active parity encoding services(apes). In IEEE Infocom '99.

....spectrum layered multicast for each source approaches Figure 1: Illustration of the Buffet concept. ffl Layering Techniques component: This provides tools to help application to layer its data. We plan to utilize publicly available implementations such as Progressive JPEG [23] and FEC techniques [39, 35], while also introducing a Degrees of Importance layering technique suitable for periodic data types, and a 2D Layered Organization that combines layering according to receiver interest with layering according to bandwidth resources. ffl Subscription Policies component: This provides policies for ....

....at the expense 24 of processing and receiving more data at the receivers. 4.4 Layering Techniques This component provides the application at hand a range of techniques to layer its data. We plan to utilize publicly available implementations such as Progressive JPEG [23] and FEC techniques [39, 35]. Here, we describe a Degrees of Importance layering technique suitable for periodic data types, and a 2 D Layered Organization that combines layering for receivers priorities on the data and bandwidth capacities in the network. 4.4.1 2 D Layered Organization Vicisano et al. 38, 39] describe ....

Rubenstein, D., Kasera, S., Towsley, D., and Kurose, J. Improving Reliable Multicast Using Active Parity Encoding Services (APES). In Proceedings of IEEE INFOCOM 99 (New York, NY, March 1999).

....of B packets, called blocks. For each block, h FEC encoded packets are generated for suitably chosen h. Receivers can recover the original block by receiving any B packets out of the B h ones. When combined with an appropriate ARQ technique, the FEC technique incurs very low network overhead [11, 12, 13, 14, 15, 3]. The FEC approach can be employed in any of the protocols we have described so far such as SRM, tree based protocols etc. A noteworthy example is the SHARQFEC protocol [3] that combines hierarchical scoping and hybrid FEC ARQ. It breaks the entire multicast group into hierarchically nested scopes ....

....estimation. 7 Conclusion We study the scalability of reliable multicast protocols. Some of the fundamental known techniques for devising scalable reliable multicast protocols include receiver centric repair [1] scoping [3, 16] suppressing unwanted traffic by using timers [1] use of FEC packets [11, 18, 3, 14, 15] etc. In this paper, we offer additional tools, namely, efficient algorithms to transmit FEC repair packets in layers so as to minimize the total excess repair traffic. We employ these on existing protocols and present simulation results that show the combined protocols to be substantially more ....

D. Rubenstein, S. Kasera, J. Kurose, and D. Towsley. Improving reliable multicast using active parity encoding services(apes). In To appear in IEEE Infocom '99.

....of B packets, called blocks. For each block, h FEC encoded packets are generated for suitably chosen h. Receivers can recover the original block by receiving any B packets out of the B h ones. When combined with an appropriate ARQ technique, the FEC technique incurs very low network overhead [11, 12, 13, 14, 15, 3]. The FEC approach can be employed in any of the protocols we have described so far such as SRM, tree based protocols etc. A noteworthy example is the SHARQFEC protocol [3] that combines hierarchical scoping and hybrid FEC ARQ. It breaks the entire multicast group into hierarchically nested scopes ....

D. Rubenstein, S. Kasera, J. Kurose, and D. Towsley. Improving reliable multicast using active parity encoding services(apes). In To appear in IEEE Infocom '99.

No context found.

D. Rubenstein, S. Kasera, D. Towsley, and J. Kurose. Improving Reliable Multicast Using Active Parity Encoding Services (APES). Technical report, University of Massachusetts, CMPSCI 98-79, July 1998.

....also utilized by many other sessions. We model link loss as a two state Markov process, which has been shown to be an effective means for modeling bursty loss [YMKT99] We do not consider hybrid techniques that combine proactive FEC transmission with local recovery techniques [Ker98] Our work in [RKTK99] demonstrates that in certain multicast session con gurations, a hybrid protocol can further reduce bandwidth requirements of reliable multicast protocols. In contrast, Nonnenmacher et al., 98] demonstrate that in many scenarios, there is no additional bene t to adding repair servers. It has ....

.... of providing reliable multicast in a scalable manner to a large numbers of receivers using randomly delayed transmissions [FJL 97] retransmission hierarchies [PSLS97,KKT98,PPV98,SBE 98] representatives [DO97] parity encoding (FEC) NBT98] and FEC hierarchy combinations [Ker98,NLJ 98,RKTK99] The primary concern of these works is to reduce bandwidth consumption and feedback implosion of a reliable multicast session. Often, the mechanism that provides this scalability also reduces the expected latency of reliable delivery. However, it is often dicult to determine a bound on the ....

D. Rubenstein, S. Kasera, D. Towsley, and J. Kurose. Improving Reliable Multicast Using Active Parity Encoding Services (APES). In Proceedings of IEEE INFOCOM'99, New York, NY, March 1999. An extended version appears as UMass CMPSCI technical report 98-79. 24

....while, at the same time, are reasonable candidates for implementation in a router. Furthermore, GMTS objects are lightweight and contain only a small amount of state. This is in contrast to recently proposed active repair services that have been proposed by the active networking community, [2, 3, 8], which require the caching of packets for the purpose of providing retransmissions. The paper is organized as follows. In the next section, we present a simple example of how GMTS can be used the context of a reliable multicast transport protocol. Section 3 introduces the generic transport ....

D. Rubenstein, S. Kasera, D. Towsley, J. Kurose. \Improving Reliable Multicast Using Active Parity Encoding Services (APES)", Proc. INFOCOM'99.

....Repair Server. SDBR BRSR GRSR Encodes Yes Yes No Decodes Yes No No Buffers Data Repairs Repairs Fwds repairs fr. src Never Sometimes Always scribed protocols to each block of data being transferred. Due to lack of space, specific feedback mechanisms are not presented here, but appear in [11]. The Store Data Build Repairs Protocol (SDBR) This protocol is similar to previous proposals [5] 9] Once a repair server reliably obtains k source packets, it reproduces (via FEC decoding) the k original data packets, which it subsequently buffers. Whenever an additional repair is required ....

....3 The decoding is not a requirement of the protocol. However, previous work assumes the capability is available. If decoding is not performed, then SDBR will require receivers to perform the recursive decoding algorithm described in [11]. APPEARS IN IEEE INFOCOM 1999 3 receive these repairs, there is never a need to buffer more than k repairs per block. This means that these protocols will never buffer more packets per block than SDBR. They also do not perform decoding at repair servers. BRSR and GRSR differ in that BRSR ....

[Article contains additional citation context not shown here]

Dan Rubenstein, Sneha Kasera, Don Towsley, and Jim Kurose, Improving Reliable Multicast Using Active Parity Encoding Services (APES), UMass CMPSCI Technical Report 98-79.

No context found.

D. Rubenstein, S. Kasera, D. Towsley, and J. Kurose. Improving Reliable Multicast Using Active Parity Encoding Services (apes). Technical Report TR98-79, Computer Science Department, University of Massachusetts at Amherst, July 1998.

No context found.

D. Rubenstein, S. Kasera, D. Towsley, and J. Kurose. Improving reliable multicast using active parity encoding services (apes). In IEEE INFOCOM'99, New York, April 1999.

No context found.

D. Rubenstein, S. Kasera, D. Towsley, and J. Kurose. Improving Reliable Multicast Using Active Parity Encoding Services (apes). Technical Report TR98-79, Computer Science Department, University of Massachusetts at Amherst, July 1998.

No context found.

D. Rubenstein, S. Kasera, D. Towsley, and J. Kurose. Improving reliable multicast using active parity encoding services (APES). In Proc. Infocom, 1999.

No context found.

D. Rubenstein, S. Kasera, D. Towsley, and J. Kurose. Improving reliable multicast using active parity encoding services (APES). In Proc. Infocom, 1999.

No context found.

D. Rubenstein, S. Kasera, D. Towsley, and J. Kurose, \Improving Reliable Multicast Using Active Parity Encoding Services," in Proc. of IEEE INFOCOM '99, March 1999, pp. 1248-1255.

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