This paper presents the design, implementation and performance of a reliable multicast transport protocol called RMTP. RMTP is based on a hierarchical structure in which receivers are grouped into local regions or domains and in each domain there is a special receiver called a Designated Receiver (DR) which is responsible for sending acknowledgments periodically to the sender, for processing acknowledgements from receivers in its domain and for retransmitting lost packets to the corresponding receivers. Since lost packets are recovered by local retransmissions as opposed to retransmissions from the original sender, end-to-end latency is significantly reduced, and the overall throughput is improved as well. Also, since only the DRs send their acknowledgments to the sender, instead of all receivers sending their acknowledgments to the sender, a single acknowledgement is generated per local region, and this prevents acknowledgement implosion. Receivers in RMTP send their acknowledgments to the DRs periodically, thereby simplifying error recovery. In addition, lost packets are recovered by selective repeat retransmissions, leading to improved throughput at the cost of minimal additional buffering at the receivers. This paper also describes the implementation of RMTP and its performance on the Internet.

Optimizing TCP (Transport Layer) for mobility has been researched extensively. We present a brief summary of existing results which indicates that most schemes require intermediaries (such as base stations) to monitor the TCP traffic and actively participate in flow control in order to enhance performance. Although these methods simulate end-to-end semantics, they do not comprise true end-to-end signaling. As a result, these techniques are not applicable when the IP payload is encrypted. For instance IPSEC, which is expected to be standard under IPv6, encrypts the entire IP payload making it impossible for intermediaries to monitor TCP traffic unless those entities are part of the security association. In addition, these schemes require changes (in the TCP/IP code) at intermediate nodes making it difficult for the mobile clients to inter-operate with the existing infrastructure. In this paper we explore the "Freeze-TCP" mechanism which is a true end-to-end scheme and does not require ...

In recent years, the standards community has developed techniques for traversing NAT/firewall boxes with UDP (that is, establishing UDP flows between hosts behind NATs). Because of the asymmetric nature of TCP connection establishment, however, NAT traversal of TCP is more difficult. Researchers have recently proposed a variety of promising approaches for TCP NAT traversal. The success of these approaches, however, depend on how NAT boxes respond to various sequences of TCP (and ICMP) packets. This paper presents the first broad study of NAT behavior for a comprehensive set of TCP NAT traversal techniques over a wide range of commercial NAT products. We developed a publicly available software test suite that measures the NAT's responses both to a variety of isolated probes and to complete TCP connection establishments. We test sixteen NAT products in the lab, and 93 home NATs in the wild. Using these results, as well as market data for NAT products, we estimate the likelihood of successful NAT traversal for home networks. The insights gained from this paper can be used to guide both design of TCP NAT traversal protocols and the standardization of NAT/firewall behavior, including the IPv4-IPv6 translating NATs critical for IPv6 transition.

As the Internet has grown, so has the challenge of accurate measurement and modeling of its topology. Commonly used but coarse methods of measuring topology, e.g., BGP tables, suffer from several limitations. To pursue more accurate empirically-based topology modeling. CAIDA began its Macroscopic Topology Project in 1998, The project focus is actively measuring topology and round trip time (RTT) information across a wide cross-section of the commodity Internet. In this paper we describe CAIDA's topology measurement architecture and our analysis and visualization tools. We describe differences between IP and AS (BGPbased) granularities of topology modeling, including advantages and limitations of both, as well as how correlation between both types of data can yield more relevant insights. We introduce four new visualization metaphors for handling macroscopic topology data, as well as a tool for aggregating multiple IP addresses into the same physical router. We highlight results of our analyses, in particular relationships between RTT and topology data, and how source and destination selection and geopolitical boundaries affect those relationships.

In this paper, we demonstrate a technique called active probing used to study TCP implementations. Active probing treats a TCP implementation as a black box, and uses a set of procedures to probe the black box. By studying the way TCP responds to the probes, one can deduce several characteristics of the implementation. The technique is particularly useful if TCP source code is unavailable.

This paper reports a series of experiments to measure TCP performance when transferring data through an Asynchronous Transfer Mode (ATM) switch. The results show that TCP buffer sizes and the ATM interface maximum transmission unit have a dramatic impact on throughput. We observe a throughput anomaly in which an increase in the receiver's buffer size decreases throughput substantially. For example, when using a 16K octet send buffer and ATM Adaptation Layer 5 on a 100 megabit per second (Mb/s) ATM path, the mean throughput for a bulk transfer drops from 15.05 Mb/s to 0.322 Mb/s if the receiver's buffer size is increased from 16K octets to 24K octets. This paper analyzes the performance, explains the anomalous behavior, and describes a solution that prevent the anomaly from occurring. This work was supported in part by a fellowship from UniForum Association. 1 Introduction Asynchronous Transfer Mode (or ATM) is a connection-oriented data communication technology that switches 53-octe...

The availability of a variety of communication devices offers a choice among networks with vastly different characteristics. A mobile host is likely to encounter these different networks and no single protocol or application can be expected to perform well under all these networks. The problem of adapting to a changing network environment is further complicated by the fact that changes in network conditions are usually transparent to the applications. In order to allow automatic adaptation of applications and protocols, awareness of link conditions and network environment is required. In this paper we present a uniform mechanism based on ICMP messages for providing environmental information to all the network protocol layers. We also show how applications and protocols can adapt to changes in the environment and in particular, demonstrate dynamic fine tuning of some of the well known protocols such as UDP and TCP. Performance measurements demonstrate that our mechanism imposes very lit...

Abstract specifications for network protocols are usually based on several assumptions that make verification easier but which make strictly faithful implementations of the specified protocols impractical, expensive, or impossible. (These assumptions usually deal with timeouts, action atomicity, action fairness, message propagation, and fault atomicity.) An implementation, therefore, is an approximation of the given abstract specification. In this dissertation, we develop an execution model that can be used in translating abstract protocol specifications written in the AP notation into C programs. Although this execution model does not strictly abide by the formal semantics of the AP notation, we show that it does maintain some important safety and progress properties of the original specifications. 1 Abstract Specifications of Protocols A protocol specification has two complementary objectives. On one hand, the

Abstract not found

We introduce a novel passive network monitoring architecture which enables us to capture and integrate data from different levels of the protocol stack using a probe which can be placed at any arbitrary point in the network. We subject the data gathered to off-line analysis and show how, by modelling the dynamics of the protocols observed, we are able to extract information which would not be available using existing means. In this paper, we show how the technique can be used to passively measure Web Server response latency. Experiments are described which use artificially generated loads to validate the techniques used, and results obtained from the observation of real traffic are presented. 1.