Roman Dunaytsev

TCP NewReno Throughput in the Presence of Correlated Losses: The Slow-but-Steady Variant

This paper presents an analytical model of steady state throughput of the Slow-but-Steady variant of TCP NewReno as a function of loss event rate, average number of segments lost per loss event, average round trip time, and retransmission timeout value. The presented model is based on the approach proposed by Padhye et al. for TCP Reno throughput modeling and extends it by capturing the effect of fast recovery algorithm of the Slow-but-Steady variant of TCP NewReno on throughput and taking into consideration slow start phase after timeout expiration. Validation by ns-2 simulations shows good agreement and accuracy of the proposed model.

Modeling TCP SACK performance over wireless channels with completely reliable ARQ/FEC

SUMMARY In this paper, we propose an analytical cross-layer model for a Transmission Control Protocol (TCP) connection running over a covariance-stationary wireless channel with a completely reliable Automatic Repeat reQuest scheme combined with Forward Error Correction (FEC) coding. Since backbone networks today are highly overprovisioned, we assume that the wireless channel is the only one bottleneck in the system which causes packets to be buffered at the wired/wireless interface and dropped as a result of buffer overflow. We develop the model in two steps. At the first step, we consider the service process of the wireless channel and derive the probability distribution of the time required to successfully transmit an IP packet over the wireless channel. This distribution is used at the next step of the modeling, where we derive expressions for the TCP long-term steady-state throughput, the mean round-trip time, and the spurious timeout probability. The developed model allows to quantify the joint effect of many implementationspecific parameters on the TCP performance over both correlated and non-correlated wireless channels. We also demonstrate that TCP spurious timeouts, reported in some empirical studies, do not occur when wireless channel conditions are covariance-stationary and their presence in those measurements should be attributed to non-stationary behavior of the wireless channel characteristics. Copyright 2011 John Wiley & Sons, Ltd.

Modeling TCP performance over wireless channels with a semi-reliable data link layer

Providing reliable data communications over wireless channels is a challenging task due to the time-varying wireless channel characteristics that often lead to bit errors. These errors propagate to the higher layers, causing loss of protocol data units. ARQ and FEC try to prevent this, providing a reliable service to the higher layers. Most analytical models that studied the effect of ARQ and FEC assumed that the ARQ scheme is fully reliable, meaning that a frame is successfully transmitted irrespective of the amount of retransmission attempts it takes. In this paper, we study the effect of a semi-reliable data link layer on performance experienced by TCP. We develop an analytical model for a TCP SACK connection running over a wireless channel with a semi-reliable ARQ, where the amount of retransmission attempts is limited by some number. The model allows to evaluate the effect of many parameters of wireless channels on TCP performance, making it suitable for performance optimization studies. These parameters include wireless channel characteristics, ARQ/FEC settings, size of protocol data units at different layers, buffer size at the IP layer, and the raw data rate of the wireless channel.

Modeling TCP SACK performance over wireless channels with semi-reliable ARQ/FEC

Providing reliable data communications over wireless channels is a challenging task because time-varying wireless channel characteristics often lead to bit errors. These errors result in loss of IP packets and, consequently, TCP segments encapsulated into these packets. Since TCP cannot distinguish packet losses due to bit corruption from those due to network congestion, any packet loss caused by wireless channel impairments leads to unnecessary execution of the TCP congestion control algorithms and, hence, sub-optimal performance. Automatic Repeat reQuest (ARQ) and Forward Error Correction (FEC) try to improve communication reliability and reduce packet losses by detecting and recovering corrupted bits. Most analytical models that studied the effect of ARQ and FEC on TCP performance assumed that the ARQ scheme is perfectly persistent (i.e., completely reliable), thus a frame is always successfully transmitted irrespective of the number of transmission attempts it takes. In this paper, we develop an analytical cross-layer model for a TCP connection running over a wireless channel with a semi-reliable ARQ scheme, where the amount of transmission attempts is limited by some number. The model allows to evaluate the joint effect of stochastic properties of the wireless channel characteristics and various implementation-specific parameters on TCP performance, which makes it suitable for performance optimization studies. The input parameters include the bit error rate, the value of the normalized autocorrelation function of bit error observations at lag 1, the strength of the FEC code, the persistency of ARQ, the size of protocol data units at different layers, the raw data rate of the wireless channel, and the bottleneck link buffer size.

Cross-layer modeling of TCP SACK performance over wireless channels with completely reliable ARQ/FEC

We propose an analytical model for a TCP SACK connection running over a wireless channel with completely reliable ARQ/FEC. We develop the model in two steps. At the first step, we consider the service process of the wireless channel and derive the probability distribution function of the time required to successfully transmit a single IP packet over the wireless channel. This distribution is used at the next step of the modeling where we derive the expression for TCP SACK steady state goodput. The developed model allows to quantify the effect of many implementation-specific parameters on TCP performance in wireless domain. We also demonstrate that TCP spurious timeouts, reported in many empirical studies, do not occur when wireless channel conditions are stationary and their presence in empirical measurements should be attributed to non-stationary behavior of wireless channel characteristics.

Modeling TCP performance over wireless channels using fixed-point approximation

Wireless channels are prone to transmission errors. Loss of IP packets due to imperfect local error concealment strategy implemented at the data-link layer is the major reason for TCP performance degradation in wireless networks. These losses are misinterpreted by TCP senders as congestion indications and subsequently lead to decrease of the sending rate. In addition to wireless losses IP packets can also be lost due to buffer overflow at the IP layer. To study performance experienced by a number of concurrent TCP sessions sharing the same wireless bottleneck we adopt the fixed-point approach. We demonstrate that throughput provided to TCP flows is complex cross-layer function depending on wireless channel statistics and configuration of the protocol stack at the wireless interface.

Refined PFTK-Model of TCP reno throughput in the presence of correlated losses

This paper presents a simple and accurate analytical model of TCP Reno throughput as a function of loss rate, average round trip time and receiver window size based on PFTK-model. The presented model refines previous work by careful examination of fast retransmit/fast recovery dynamics in the presence of correlated losses and taking into consideration slow start phase after timeout. The accuracy of the proposed model is validated against simulation results and compared with those of PFTK-model. Simulation results show that our model gives a more accurate estimation of TCP Reno throughput in the presence of correlated losses than PFTK-model.

An Analytical Comparison of the Slow-but-Steady and Impatient Variants of TCP New Reno

Current standard defines two variants of TCP NewReno: the Slow-but-Steady and Impatient. While the behavior of various TCP implementations has been extensively studied over the last years, little attention has been paid to performance analysis of different variants of TCP NewReno. In this paper, we first develop an analytical model of the Impatient variant, which being combined with the earlier proposed model of the Slow-but-Steady variant gives a comprehensive analytical model of TCP NewReno throughput. We then make an analytical comparison of the Impatient and Slow-but-Steady throughputs in the presence of correlated losses. We show that, although neither of the two variants is optimal, the Impatient variant provides the same throughput as the Slow-but-Steady one in a wide range of network conditions and significantly outperforms it in case of large windows and multiple packet drops. This could be of special interest for networks with large bandwidth and long delay.