TCP is widely used in commercial media streaming systems, with recent measurement studies indicating that a significant fraction of Internet streaming media is currently delivered over HTTP/TCP. These observations motivate us to develop analytic performance models to systematically investigate the performance of TCP for both live and stored media streaming. We validate the models via ns simulations and experiments conducted over the Internet. Our models provide guidelines indicating the circumstances under which TCP streaming leads to satisfactory performance, showing, for example, that TCP generally provides good streaming performance when the achievable TCP throughput is roughly twice the media bitrate, with only a few seconds of startup delay.

The issue of router buffer sizing is still open and significant. Previous work either considers open-loop traffic or only analyzes persistent TCP flows. This paper differs in two ways. First, it considers the more realistic case of non-persistent TCP flows with heavy-tailed size distribution. Second, instead of only looking at link metrics, we focus on the impact of buffer sizing on TCP performance. Specifically, our goal is to find the buffer size that maximizes the average per-flow TCP throughput. Through a combination of testbed experiments, simulation, and analysis, we reach the following conclusions. The output/input capacity ratio at a network link largely determines the required buffer size. If that ratio is larger than one, the loss rate drops exponentially with the buffer size and the optimal buffer size is close to zero. Otherwise, if the output/input capacity ratio is lower than one, the loss rate follows a power-law reduction with the buffer size and significant buffering is needed, especially with flows that are mostly in congestion-avoidance. Smaller transfers, which are mostly in slow-start, require significantly smaller buffers. We conclude by revisiting the ongoing debate on “small versus large” buffers from a new perspective.

Motivated by the wide use of TCP for streaming in practice and the increasing availability of multipath between end hosts, we study multipath live streaming via TCP in this paper. We first design a simple and practical TCP-based multipath streaming scheme, named Dynamic MPath-streaming (DMP-streaming), which dynamically distributes packets over multiple paths by implicitly inferring the available bandwidths on these paths. To allow systematic performance study, we develop an analytical model for DMP-streaming and validate the model using extensive ns simulation and Internet experiments. We explore the parameter space of this model and find that DMP-streaming generally provides satisfactory performance when the aggregate achievable TCP throughput is 1.6 times the video bitrate, with a few seconds of startup delay. Last, we comment on the benefits of using multipath versus single path for TCP-based streaming. 1.

Modeling TCP performance is an important issue that attracted research attention over the past decade. In this paper, we present an overview of models used to capture the TCP behavior. We compare several existing analytical models with respect to modeled attributes, modeling assumptions, and validation techniques. We also identify features that new TCP models should possess. Finally, we address the importance of devising common validation techniques and performance evaluation metrics for TCP models.

Abstract—We introduce a stochastic model of a bottleneck ECN/RED gateway under a large number of heterogeneous TCP flows, i.e., flows with diverse round-trip delays and session dynamics. We investigate the asymptotic behavior of the system and show that as the number of flows becomes large, the buffer dynamics and aggregate traffic simplify and can be accurately described by simple stochastic recursions independent of the number of flows, resulting in a scalable model. Based on the Central Limit analysis in the paper, we identify the sources of fluctuations in queue size and describe the relationship between the system parameters such as the marking function and variance of queue size. A closed-form approximation for the mean queue size as a function of system parameters is provided from a simple steady-state analysis. Numerical examples are provided to validate our results. Index Terms—Congestion control, modeling, stochastic systems. I.

Short-lived TCP traffic (e.g., web mice) composes the majority of the current Internet traffic. Accurate traffic modeling of a large number of short-lived TCP flows is extremely difficult due to (i) the interaction between session, transport, and network layers; and (ii) the explosion of the size of state space when the number of flows is large. Typically, ad-hoc assumptions are required for the analysis to be tractable under a certain regime.

Abstract — While earlier studies have pointed out that short-lived TCP flows (mice) may hurt long-lived TCP flows (elephants) in the long term, they provide insufficient insight for developing scenarios leading to drastic drop in throughputs of long-lived TCP flows. We have systematically developed TCP adversarial scenarios where we use short-lived TCP flows to adversely influence long-lived TCP flows. Our scenarios are interesting since, (a) they point out the increased vulnerabilities of recently proposed scheduling, AQM and routing techniques that further favor short-lived TCP flows, and (b) they are more difficult to detect when intentionally found to target long-lived TCP flows. We systematically exploit the ability of TCP flows in slow-start to rapidly capture greater proportion of bandwidth compared to long-lived TCP flows in congestion avoidance phase, to a point where they drive long-lived TCP flows into timeout. We use simulations, analysis, and experiments to systematically study the dependence of the severity of impact on long-lived TCP flows on key parameters of short-lived TCP flows – including their locations, durations, and numbers, as well as the intervals between consecutive flows. We derive characteristics of pattern of short-lived flows that exhibit extreme adverse impact on longlived TCP flows. Counter to common beliefs, we show that targeting bottleneck links does not always cause maximal performance degradation for the long-lived flows. In particular, our approach illustrates the interactions between TCP flows and multiple bottleneck links and their sensitivities to correlated losses in the absence of ‘non-TCP friendly ’ flows and paves the way for a systematic synthesis of worst-case congestion scenarios. While randomly generated sequences of short-lived TCP flows may provide some reductions (up to 10%) in the throughput of the long-lived flows, the scenarios we generate cause much greater reductions (>85%) for several TCP variants (Tahoe, Reno, New Reno, Sack), and for different packet drop policies (DropTail, RED). Keywords-TCP, short-lived flow, long-lived flow, DoS

this paper we will refer to in-profile traffic as Green, while Yellow class will be used to represent both out-of-profile and Best Effort traffic The main contribution of this paper is twofold: the proposal a Fair Marker that, by using a limited amount of memory in edge nodes, achieves the same performance enhancement of a perflow marker, but with a much simpler design; also, a simple analytical model of the completion time of short-lived flows in a DiffServ domain using our proposed Fair Marker

In this paper we propose an enhancement to the TCP protocol, called TCP Smart Framing, or TCP-SF for short, that enables the Fast Retransmit/Recovery algorithm even when the congestion window is small. TCP-SF is particularly effective for short-lived flows, as most of the current Internet traffic is. Without modifying the TCP congestion control based on the additive-increase/multiplicativedecrease paradigm, TCP-SF adopts a novel segmentation algorithm: while Classic TCP starts sending one segment, a TCP-SF source is allowed to send an initial window of 4 smaller segments, whose aggregate payload is equal to the connection 's MSS. This key idea can be implemented on top of any TCP flavor, from Tahoe to SACK, and requires modifications to the server behavior only.

this paper we extend the model in [10] by incorporating explicitly an additional layer of dynamics, namely the session layer, with TCP in the transport layer and RED gateway [2] with ECN [1] capability in the network layer. As the number of sessions increases, we show that (i) the queue size per session and the workload per session brought in during a round-trip time converge to deterministic processes; and (ii) the sessions become asymptotically independent, indicating that the RED mechanism does alleviate the synchronization problem among the connections