The proliferation of applications that must reliably distribute bulk data to a large number of autonomous clients motivates the design of new multicast and broadcast prot.ocols. We describe an ideal, fully scalable protocol for these applications that we call a digital fountain. A digital fountain allows any number of heterogeneous clients to acquire bulk data with optimal efficiency at times of their choosing. Moreover, no feedback channels are needed to ensure reliable delivery, even in the face of high loss rates. We develop a protocol that closely approximates a digital fountain using a new class of erasure codes that for large block sizes are orders of magnitude faster than standard erasure codes. We provide performance measurements that demonstrate the feasibility of our approach and discuss the design, implementation and performance of an experimental system. 1

Abstract—The proliferation of applications that must reliably distribute large, rich content to a vast number of autonomous receivers motivates the design of new multicast and broadcast protocols. We describe an ideal, fully scalable protocol for these applications that we call a digital fountain. A digital fountain allows any number of heterogeneous receivers to acquire content with optimal efficiency at times of their choosing. Moreover, no feedback channels are needed to ensure reliable delivery, even in the face of high loss rates. We develop a protocol that closely approximates a digital fountain using two new classes of erasure codes that for large block sizes are orders of magnitude faster than standard erasure codes. We provide performance measurements that demonstrate the feasibility of our approach and discuss the design, implementation, and performance of an experimental system. Index Terms—Content delivery, erasure codes, forward error correction, reliable multicast, scalability. I.

Previous scalable on-demand streaming protocols do not allow clients to recover from packet loss. This paper develops new protocols that (1) have a tunably short latency for the client to begin playing the media, (2) allow heterogeneous clients to recover lost packets without jitter as long as each client's cumulative loss rate is within a tunable threshold, and (3) assume a tunable upper bound on the transmission rate to each client that can be as small as a fraction (e.g., 25%) greater than the media play rate. Models are developed to compute the minimum required server bandwidth for a given loss rate and playback latency. The results of the models are used to develop the new protocols and assess their performance. The new protocols, Reliable Periodic Broadcast (RPB) and Reliable Bandwidth Skimming (RBS), are simple to implement and achieve nearly the best possible scalability and efficiency for a given set of client characteristics and desirable/feasible media quality. Furthermore, the results show that the new reliable protocols that transmit to each client at only twice the media play rate have similar performance to previous protocols that require clients to receive at many times the play rate.

The use of scalable video with layered multicast has been shown to be an effective method to achieve rate control in heterogeneous networks. In this paper, we propose the use of layered FEC as an error control mechanism in a layered multicast framework. By organizing FEC into multiple lay-ers, receivers can obtain different levels of protection commensurate with their respective channel conditions. Efficient network utilization is achieved as FEC streams are multicast, and only to receivers that need them. Furthermore, FEC is used without overall rate expansion by selectively dropping data layers to make room for FEC layers. Effects of bursty losses are amortized by stag-gering the FEC streams in time, giving rise to a trade-off between delay and quality. For rate control at the receivers, we propose an equation-based approach that computes network usage as a function of measured network characteristics. We show that equation-based rate control achieves more fair bandwidth sharing amongst competing sessions as compared to existing multicast rate con-trol schemes such as RLM and RLC. Fairness is achieved since competing sessions sharing a path will measure similar network characteristics. Simulations and actual MBONE experiments are per-formed using error-resilient, scalable video compression. We find that video quality is significantly improved at the same communication rate when layered FEC is used. 1

In secure group communications, users of a group share a common group key. A key server sends the group key to authorized new users as well as performs group rekeying for group users whenever the key changes. In this paper, we investigate scalability issues of reliable group rekeying, and provide a performance analysis of our group key management system (called keygem) based upon the use of key trees. Instead of rekeying after each join or leave, we use periodic batch rekeying to improve scalability and alleviate out-of-sync problems among rekey messages as well as between rekey and data messages. Our analyses show that batch rekeying can achieve large performance gains. We then investigate reliable multicast of rekey messages using proactive FEC. We observe that rekey transport has an eventual reliability and a soft real-time requirement, and that the rekey workload has a sparseness property, that is, each group user only needs to receive a small fraction of the packets that carry a rekey message sent by the key server.

Real-Time reliable multicast over a best-effort service network remains a challenging research problem. Most protocols for reliable multicast use repair techniques that result in significant and variable delay, which can lead to missed deadlines in real-time scenarios. This paper presents a repair technique that combines forward error correction (FEC) with automatic repeat request (ARQ). The novel aspect of the technique is its ability to reduce delay in reliable multicast delivery by sending repairs proactively (i.e., before they are required). The technique requires minimal state at senders and receivers, and no additional active router functionality beyond what is required by the current multicast service model. Furthermore, the technique uses only end-to-end mechanisms, where all data and repairs are transmitted by the data-originating source, leaving receivers free from any burden of sending repairs. We simulate a simple round-based version of a protocol embodying this technique to show its effectiveness in preventing repair request implosion, reducing the expected time of reliable delivery of data, and keeping bandwidth usage for repairs low. We show how a protocol using the technique can be adapted to provide delivery that is reliable before a real-time deadline with probabilities extremely close to one. Finally, we develop several variations of the protocol that use the technique in various fashions for high rate data streaming applications, and present results from additional simulations that examine performance in a variety of Internet-like heterogeneous networks. 1

Abstract — We propose and evaluate novel reliable multicast protocols that combine active repair service (a.k.a. local recovery) and parity encoding (a.k.a. forward error correction or FEC) techniques. We show that, compared to other repair service protocols, our protocols require less buffer inside the network, maintain the low bandwidth requirements of previously proposed repair service / FEC combination protocols, and reduce the amount of FEC processing at repair servers, moving more of this processing to the end-hosts. We also examine repair service / FEC combination protocols in an environment where loss rates differ across domains within the network. We find that repair services are more effective than FEC at reducing bandwidth utilization in such environments. Furthermore, adding FEC to a repair services protocol not only reduces buffer requirements at repair servers, but also reduces bandwidth utilization in domains with high loss, or in domains with large populations of receivers. I.

We present the design and specification of a protocol for scalable and reliable group rekeying together with performance evaluation results. The protocol is based upon the use of key trees for secure groups and periodic batch rekeying. At the beginning of each rekey interval, the key server sends a rekey message to all users consisting of encrypted new keys (encryptions, in short) carried in a sequence of packets. We present a scheme for identifying keys, encryptions, and users, and a key assignment algorithm that ensures that the encryptions needed by a user are in the same packet. Our protocol provides reliable delivery of new keys to all users eventually. It also attempts to deliver new keys to all users with a high probability by the end of the rekey interval. For each rekey message, the protocol runs in two steps: a multicast step followed by a unicast step. Proactive forward error correction (FEC) multicast is used to reduce delivery latency. Our experiments show that a small FEC block size can be used to reduce encoding time at the server without increasing server bandwidth overhead. Early transition to unicast, after at most two multicast rounds, further reduces the worst-case delivery latency as well as user bandwidth requirement. The key server adaptively adjusts the proactivity factor based upon past feedback information; our experiments show that the number of NACKs after a multicast round can be effectively controlled around a target number. Throughout the protocol design, we strive to minimize processing and bandwidth requirements for both the key server and users.

Forward Error Correction (FEC) has been proposed as a technique for implementing efficient reliable multicast (RM). However, FEC incurs costs in encode/decode delay and implementation complexity. How much benefit is provided relative to these costs and how dependent is the benefit on the specific RM protocol? In this paper, we evaluate the benefits of FEC for RM, considering both proactive and reactive use with three RM recovery techniques: duplicate avoidance, limited scope multicast and subcast. Our simulationbased results indicate that FEC provides little benefit for an efficient RM protocol like OTERS and introduces extra delay for multi-point streaming data applications. 1

The use of proxies for local error recovery and congestion control is a scalable technique used to overcome a number of wellknown problems in Reliable Multicast (RM). The idea is that the multicast delivery tree is partitioned into subgroups that form a hierarchy rooted at the source, hence the term Hierarchical Reliable Multicast (HRM). For each subgroup, there is a designated node, the proxy, which is responsible for collecting the feedback from the subgroup and for locally re-transmitting the lost packets. The performance of any RM protocol is affected by the underlying multicast routing tree and its loss characteristics. Furthermore, the performance of the HRM approach, in particular, strongly depends on the appropriate partitioning of the tree and the selection of proxies. In this paper, we first model the HRM problem, then define and compute appropriate performance metrics and finally give insights on the optimal location of proxies. Keywords Performance analysis, reliable mult...