This note proposes a modification for TCP's slow-start for use with TCP connections with large congestion windows. For TCP connections that are able to use congestion windows of thousands (or tens of thousands) of MSS-sized segments (for MSS the sender's MAXIMUM SEGMENT SIZE), the current slow-start procedure can result in increasing the congestion window by thousands of segments in a single round-trip time. Such an increase can easily result in thousands of packets being dropped in one round-trip time. This is often counterproductive for the TCP flow itself, and is also hard on the rest of the traffic sharing the congested link. This note proposes Limited Slow-Start as an optional mechanism for limiting the number of segments by which the congestion window is increased for one window of data during slow-start, in order to improve performance for TCP connections with large congestion windows.

In this paper, we present STP, a system in which communicating end hosts use untrusted mobile code to remotely upgrade each other with the transport protocols that they use to communicate. New transport protocols are written in a type-safe version of C, distributed out-of-band, and run in-kernel. Communicating peers select a transport protocol to use as part of a TCP-like connection setup handshake that is backwards-compatible with TCP and incurs minimum connection setup latency. New transports can be invoked by unmodified applications. By providing a late binding of protocols to hosts, STP removes many of the delays and constraints that are otherwise commonplace when upgrading the transport protocols deployed on the Internet. STP is simultaneously able to provide a high level of security and performance. It allows each host to protect itself from untrusted transport code and to ensure that this code does not harm other network users by sending significantly faster than a compliant TCP. It runs untrusted code with low enough overhead that new transport protocols can sustain near gigabit rates on commodity hardware. We believe that these properties, plus compatibility with existing applications and transports, complete the features that are needed to make STP useful in practice. Categories and Subject Descriptors D.4.4 [Operating Systems]: Communications Management; D.4.6 [Operating Systems]: Security and Protection; C.2.2 [Network Protocols]: Protocol architecture General Terms Design, Implementation, Deployment Keywords Transport Protocols, TCP-friendliness, Untrusted Mobile Code Permission to make digital or hard copies of all or part of this work for personal or classroom use is granted without fee provided that copies are not made or distributed for profit or ...

Internet routers require buffers to hold packets during times of congestion. The buffers need to be fast, and so ideally they should be small enough to use fast memory technologies such as SRAM or all-optical buffering. Unfortunately, a widely used rule-of-thumb says we need a bandwidth-delay product of buffering at each router so as not to lose link utilization. This can be prohibitively large. In a recent paper, Appenzeller et al. challenged this rule-of-thumb and showed that for a backbone network, the buffer size can be divided by √ N without sacrificing throughput, where N is the number of flows sharing the bottleneck. In this paper, we explore how buffers in the backbone can be significantly reduced even more, to as little as a few dozen packets, if we are willing to sacrifice a small amount of link capacity. We argue that if the TCP sources are not overly bursty, then fewer than twenty packet buffers are sufficient for high throughput. Specifically, we argue that O(log W) buffers are sufficient, where W is the window size of each flow. We support our claim with analysis and a variety of simulations. The change we need to make to TCP is minimal—each sender just needs ∗ This work was supported under DARPA/MTO DOD-N award no. W911NF-04-0001/KK4118 (LASOR PROJECT)

This paper describes our experiences in the development of the UDP-based Data Transport (UDT) protocol, an application level transport protocol used in distributed data intensive applications. The new protocol is motivated by the emergence of wide area high-speed optical networks, in which TCP is often found to fail to utilize the abundant bandwidth. UDT demonstrates good efficiency and fairness (including RTT fairness and TCP friendliness) characteristics in high performance computing applications where a small number of bulk sources share the abundant bandwidth. It combines both rate and window control and uses bandwidth estimation to determine the control parameters automatically. This paper presents the rationale behind UDT: how UDT integrates these schemes to support high performance data transfer, why these schemes are used, and what the main issues are in the design and implementation of this high performance transport protocol.

Random Early Detection (RED) is the recommended active queue management scheme for rapid deployment throughout the Internet. As a result, there have been considerable research efforts in studying the performance of RED. However, previous studies have often focused on relatively homogeneous environment. The effects of RED in a heterogeneous environment are not thoroughly understood. In this paper, we use extensive simulations to explore the interaction between RED and various types of heterogeneity, as well as the impact of such interaction on the user-perceived end-to-end performance. Our results show that overall RED improves performance at least for the types of heterogeneity we have considered.

Bandwidth-delay product (BDP) and its upper bound (BDP-UB) have been well-understood in wireline networks such as the Internet. However, they have not been carefully studied in the multi-hop wireless ad hoc network (MANET) domain. In this paper, we show that the most significant difference of computing BDP and BDP-UB in MANET is the coupling of bandwidth and delay over a wireless link, where only one packet is allowed to be transmitted over the channel at a time. Based on this observation, we prove that BDP-UB of a path in MANET is upper bounded by is the number of round-trip hops of the path. We then further obtain two tighter bounds of BDP-UB, and verify them through ns-2 simulations.

In heterogeneous networks, TCP connections that incorporate a terrestrial or satellite radio link are greatly disadvantaged with respect to entirely wired connections, because of their longer round trip times (RTTs). To cope with this problem, a new TCP proposal, the TCP Hybla, is presented and discussed in the paper. It stems from an analytical evaluation of the congestion window dynamics in the TCP standard versions (Tahoe, Reno, NewReno), which suggests the necessary modifications to remove the performance dependence on RTT. TCP Hybla performance is firstly evaluated in the case of an ideal channel, with good correlation between analytical and simulation data. Then, more realistic situations, which require the adoption of a benchmark network topology and a careful ns-2 simulation set-up, are examined. In particular, TCP Hybla performance is compared with that achievable by TCP standard in the presence of congestion and link losses, either separately or jointly considered. In all the examined cases, the superiority of TCP Hybla is evident, as it greatly reduces the severe penalization suffered by wireless, and especially satellite, TCP connections. Finally, it is worth noting that TCP Hybla does not infringe the end to end semantics of TCP and is compatible with other promising enhancements.

Internet traffic exhibits multifaceted burstiness and correlation structure over a wide span of time scales. Previous work analyzed this structure in terms of heavy-tailed session characteristics, as well as TCP timeouts and congestion avoidance, in relatively long time scales. We focus on shorter scales, typically less than 100-1000 milliseconds. Our objective is to identify the actual mechanisms that are responsible for creating bursty traffic in those scales. We show that TCP self-clocking, joint with queueing in the network, can shape the packet interarrivals of a TCP connection in a two-level ON-OFF pattern. This structure creates strong correlations and burstiness in time scales that extend up to the Round-Trip Time (RTT) of the connection. This effect is more important for bulk transfers that have a large bandwidth-delay product relative to their window size. Also, the aggregation of many flows, without rescaling their packet interarrivals, does not converge to a Poisson stream, as one might expect from classical superposition results. Instead, the burstiness in those scales can be significantly reduced by TCP pacing. In particular, we focus on the importance of the minimum pacing timer, and show that a 10-millisecond timer would be too coarse for removing short-scale traffic burstiness, while a 1-millisecond timer would be sufficient to make the traffic almost as smooth as a Poisson stream in sub-RTT scales.

Today’s Internet traffic is dominated by short Web data transfers. Such a workload is well known to interact poorly with the TCP protocol. TCP uses the slow start procedure to probe the network for bandwidth both at connection start up and upon restart after an idle period. This usually requires several roundtrips and is inefficient when the duration of a transfer is short. In this paper, we propose a new technique, which we call TCP/SPAND, to speed up short data transfers. In TCP/SPAND, network performance information is shared among many co-located hosts to estimate each connection’s fair share of the network resources. Based on such estimation and the transfer size, the TCP sender determines the optimal initial congestion window size. Instead of doing slow start, it uses a pacing scheme to smoothly send out the packets in its initial congestion window. We use extensive simulations to evaluate the performance of the resulting system. Our results show that TCP/SPAND significantly reduces latency for short transfers even in presence of multiple heavily congested bottlenecks. Meanwhile, the performance benefit does not come at the expense of degrading the performance of connections using the standard TCP. That is, TCP/SPAND is TCP friendly. 1.

In this paper, we present the design, implementation, and evaluation of a novel endpoint congestion control system that achieves near-optimal performance in all likely circumstances. Our approach, called the Probe Control Protocol (PCP), emulates network-based control by using explicit short probes to test and temporarily acquire available bandwidth. Like TCP, PCP requires no network support beyond plain FIFO queues. Our initial experiments show that PCP, unlike TCP, achieves rapid startup, small queues, and low loss rates, and that the efficiency of our approach does not compromise eventual fairness and stability. Further, PCP is compatible with sharing links with legacy TCP hosts, making it feasible to deploy. 1