Suchitra Raman

Scalable data naming for application level framing in reliable multicast

The Application Levd Raining (ALF) protocol architecture [2] encourag~ apphcation control over mech* that traditiontiy W tithin the %ansport layer”, e.g., loss detection md recovery. ~aditiond ARQ-based rtiable protocok for unicast (e.g., TCP) as w~ as mtdticast (e.g., Horus [30], ~ITP [15], etc.) number data units sequenti~y to detect 10SSS. Unfortunatdy, these hmport-levd sequence nuntbers do not permit receivers to fiatibly t~or their r~abti~ semantics. Achieving receiver-driven r&abti~ is cutnbersome in the e%mg ~ayeredn artitecture of the netiork protocol stack whine the receiving application has no knowledge of how apphcation-levd objects map onto transport levd sequence numbers. b this paper, we propose a new data naming scheme that e\Toss the structure of appEcation data to the transport layer, thereby mcing the ~\TrXbti@ of= apphcations’ r&abti@ and ordering se mantiw. IfTe apply this data naming scheme to a rtiable mtitiwt protocol framework to achieve receiver-t~ored re fiabii~ that enhancw its sdabii~. To demonstrate the ~caey of our scheme, we have designed and implemented our scalable naming and announcement protocol (SNAP) in the mtitimedia application too~t MASH [17] as a reusable protocol modtie.

Asymptotic behavior of global recovery in SRM

The development and deployment of a large-scale, wide-area multicast infrastructure in the Internet has enabled a new family of multi-party, collaborative applications. Several of these applications, such as multimedia slide shows, shared whiteboards, and large-scale multi-player games, require reliable multicast transport, yet the underlying multicast infrastructure provides only a best-effort delivery service. A difficult challenge in the design of efficient protocols that provide reliable service on top of the best-effort multicast service is to maintain acceptable performance as the protocol scales to very large session sizes distributed across the wide area. The Scalable, Reliable Multicast (SRM) protocol [6] is a receiver-driven scheme based on negative acknowledgments (NACKs) reliable multicast protocol that uses randomized timers to limit the amount of protocol overhead in the face of large multicast groups, but the behavior of SRM at extremely large scales is not well-understood.In this paper, we use analysis and simulation to investigate the scaling behavior of global loss recovery in SRM. We study the protocols control-traffic overhead as a function of group size for various topologies and protocol parameters, on a set of simple, representative topologies --- the cone (a variant of a clique), the linear chain, and the binary tree. We find that this overhead, as a function of group size, depends strongly on the topology: for the cone, it is always linear; for the chain, it is between constant and logarithmic; and for the tree, it is between constant and linear.