TCP over OBS: Impact of consecutive multiple packet losses and improvements

In TCP over OBS networks, consecutive multiple packet losses are common since an optical burst usually contains a number of consecutive packets from the same TCP sender. In this paper, we first present a new theoretical method to analyze the behavior of Reno when consecutive multiple packet losses occur. Results of the analysis indicate that even a small number of consecutive multiple packet losses can force Reno to timeout. Then we propose B-Reno, a newly designed TCP implementation that can overcome Reno’s inefficiency in dealing with consecutive multiple packet losses over OBS networks and can avoid the shortcomings of New-Reno and SACK. Results of comprehensive simulations indicate that B-Reno over OBS networks can achieve a performance better than Reno and New-Reno, and that it can also achieve a performance similar to that of SACK. Moreover, B-Reno only needs some simple modifications to New-Reno at the sender’s protocol stack, and thus has less difficulty in deployment and less protocol complexity than that of SACK.

TCP Veno: TCP enhancement for transmission over wireless access networks

Wireless access networks in the form of wireless local area networks, home networks, and cellular networks are becoming an integral part of the Internet. Unlike wired networks, random packet loss due to bit errors is not negligible in wireless networks, and this causes significant performance degradation of transmission control protocol (TCP). We propose and study a novel end-to-end congestion control mechanism called TCP Veno that is simple and effective for dealing with random packet loss. A key ingredient of Veno is that it monitors the network congestion level and uses that information to decide whether packet losses are likely to be due to congestion or random bit errors. Specifically: (1) it refines the multiplicative decrease algorithm of TCP Reno-the most widely deployed TCP version in practice-by adjusting the slow-start threshold according to the perceived network congestion level rather than a fixed drop factor and (2) it refines the linear increase algorithm so that the connection can stay longer in an operating region in which the network bandwidth is fully utilized. Based on extensive network testbed experiments and live Internet measurements, we show that Veno can achieve significant throughput improvements without adversely affecting other concurrent TCP connections, including other concurrent Reno connections. In typical wireless access networks with 1% random packet loss rate, throughput improvement of up to 80% can be demonstrated. A salient feature of Veno is that it modifies only the sender-side protocol of Reno without changing the receiver-side protocol stack.     

Analytic models for the latency and steady-state throughput of TCP Tahoe, Reno, and SACK

Continuing the process of improvements made to TCP through the addition of new algorithms in Tahoe and Reno, TCP SACK aims to provide robustness to TCP in the presence of multiple losses from the same window. In this paper we present analytic models to estimate the latency and steady-state throughput of TCP Tahoe, Reno, and SACK and validate our models using both simulations and TCP traces collected from the Internet. In addition to being the first models for the latency of finite Tahoe and SACK flows, our model for the latency of TCP Reno gives a more accurate estimation of the transfer times than existing models. The improved accuracy is partly due to a more accurate modeling of the timeouts, evolution of cwnd during slow start and the delayed ACK timer. Our models also show that, under the losses introduced by the droptail queues which dominate most routers in the Internet, current implementations of SACK can fail to provide adequate protection against timeouts and a loss of roughly more than half the packets in a round will lead to timeouts. We also show that with independent losses SACK performs better than Tahoe and Reno and, as losses become correlated, Tahoe can outperform both Reno and SACK.     

TCP over optical burst-switched networks with controlled burst retransmission

For optical burst-switched (OBS) networks in which TCP is implemented at a higher layer, the loss of bursts can lead to serious degradation of TCP performance. Due to the bufferless nature of OBS, random burst losses may occur, even at low traffic loads. Consequently, these random burst losses may be mistakenly interpreted by the TCP layer as congestion in the network. The TCP sender will then trigger congestion control mechanisms, thereby reducing TCP throughput unnecessarily. In this paper, we introduce a controlled retransmission scheme in which the bursts lost due to contention in the OBS network are retransmitted at the OBS layer. The OBS retransmission scheme can reduce the burst loss probability in the OBS core network. Also, the OBS retransmission scheme can reduce the probability that the TCP layer falsely detects congestion, thereby improving the TCP throughput. We develop an analytical model for evaluating the burst loss probability in an OBS network that uses a retransmission scheme, and we also analyze TCP throughput when the OBS layer implements burst retransmission. We develop a simulation model to validate the analytical results. Simulation and analytical results show that an OBS layer with controlled burst retransmission provides up to two to three orders of magnitude improvement in TCP throughput over an OBS layer without burst retransmission. This significant improvement is primarily because the TCP layer triggers fewer time-outs when the OBS retransmission scheme is used.     

Reducing the impact of false time out on TCP performance in TCP over OBS networks

Random burst contention losses plague the performance of Optical Burst Switched networks. Such random losses occur even in low load network condition due to the analogous behavior of wavelength and routing algorithms. Since a burst may carry many packets from many TCP sources, its loss can trick the TCP sources to conclude/infer that the underlying (optical) network is congested. Accordingly, TCP reduces sending rate and switches over to either fast retransmission or slow start state. This reaction by TCP is uncalled-for in TCP over OBS networks as the optical network may not be congested during such random burst contention losses. Hence, these losses are to be addressed in order to improve the performance of TCP over OBS networks. Existing work in the literature achieves the above laid objective at the cost of violating the semantics of OBS and/or TCP. Several other works make delay inducing assumptions. In our work, we introduce a new layer, called Adaptation Layer, in between TCP and OBS layers. This layer uses burst retransmission to mitigate the effect of burst loss due to contention on TCP by leveraging the difference between round trip times of TCP and OBS. We achieve our objective with the added advantage of maintaining the semantics of the layers intact.     

Performance evaluation of high speed TCP over optical burst switching networks

Study of TCP performance over OBS networks has been an important problem of research lately and it was found that due to the congestion control mechanism of TCP and the inherent bursty losses in the Optical Burst Switching (OBS) network, the throughput of TCP connections degrade. On the other hand, High Speed TCP (HSTCP) was proposed as an alternative to the use of TCP in high bandwidth-delay product networks. HSTCP aggressively increases the congestion window used in TCP, when the available bandwidth is high and decreases the window cautiously in response to a congestion event. In this work, we make a thorough simulation study of HSTCP over OBS networks. While the earlier works in the literature used a linear chain of nodes as the network topology for the simulation, we use the popular 14-node NSFNET topology that represents an arbitrary mesh network in our study. We also study the performance of HSTCP over OBS for different bandwidths of access networks. We use two different cases for simulations where in the first HSTCP connections are routed on disjoint paths while in the second they contend for resources in the network links. These cases of simulations along with the mesh topology help us clearly distinguish between the congestion and contention losses in the OBS network and their effect on HSTCP throughput. For completeness of study, we also simulate TCP traffic over OBS networks in all these cases and compare its throughput with that of HSTCP. We observe that irrespective of the access network bandwidth and the burst loss rate in the network, HSTCP outperforms TCP in terms of the throughput and robustness against multiple burst losses up to the expected theoretical burst loss probability of 10⁻³.     

Shielding TCP from last hop wireless losses

The pervasiveness of the transport control protocol (TCP) and the proliferation of wireless local area networks (WLAN) of the 802.11 type make the topic of TCP performance over last hop wireless networks very relevant. The Snoop protocol, a link layer solution introduced several years ago to improve the performance of TCP in this scenario, has been shown to neglect its benefits to the most widely used TCP version, TCP SACK. In this paper, we introduce the TCP SACK‐Aware Snoop protocol to address this problem. Our results indicate that the TCP SACK‐Aware Snoop protocol improves the performance of TCP SACK by around 30% compared with the original Snoop protocol and by about 8% in an environment where no TCP enhancing mechanism is in place. In addition, we introduce further modifications to the proposed protocol to make its advantages available to any TCP sender. We also show that the mechanism does not introduce unfairness among TCP sources and somewhat protects TCP against UDP traffic. Our results show important throughput improvements to all TCP versions and demonstrate that the TCP SACK‐Aware Snoop protocol shields TCP from last hop wireless losses providing throughtput values very close to the maximum possible. Copyright © 2006 John Wiley & Sons, Ltd.     

Performance of TCP extensions on noisy high BDP networks

Experiments on high bandwidth delay product (BDP) networks assist in the design and understanding of future global networks. This paper describes the results of experiments with different implementations of TCP on a high speed ATM/SONET network over high delay and noisy channels. Comparisons are also made with host/traffic configurations over various smaller BDP systems, experimental comparisons of three different implementations of TCP; TCP Reno, TCP new Reno, and TCP SACK as a function of bit error rates (BER) and round-trip times (RTT) are presented     

Detecting spurious timeouts in wireless cellular networks using DS‐Agent

Large and sudden variations in packet transmission delays are often unavoidable in wireless networks. Such large delays, refer to as delay spikes (DSs), are likely to exceed several times the typical network round‐trip‐time figures, which can cause TCP spurious timeouts. The spurious timeouts lead to unnecessary retransmissions and reduction of the TCP sender's transmission rate, and degradation of TCP throughput. In this paper we propose a new scheme called DS‐Agent. The spurious timeout is detected by a DS‐Agent and thus TCP sender can response to this spurious timeout accordingly. The simulation results show the better performance of DS‐Agent scheme compared with F‐RTO and TCP Reno in the presence of DSs which is caused by mobility. Copyright © 2006 John Wiley & Sons, Ltd.     

Formal specification and verification of safety and performance ofTCP selective acknowledgment

We present a formal specification of the selective acknowledgment (SACK) mechanism that is being proposed as a new standard option for TCP. The formal specification allows one to reason about the SACK protocol; thus, we are able to formally prove that the SACK mechanism does not violate the safety properties (reliable, at most once, and in order message delivery) of the acknowledgment (ACK) mechanism that is currently used with TCP. The new mechanism is being proposed to improve the performance of TCP when multiple packets are lost from one window of data. The proposed mechanism for implementing the SACK option for TCP is sufficiently complicated that it is not obvious that it is indeed safe, so we think it is important to formally verify its safety properties. In addition to safety, we are also able to show that SACK can improve the time it takes for the sender to recover from multiple packet losses. With the additional information available at a SACK sender, the round-trip time that a cumulative ACK sender waits before retransmitting each subsequent packet lost after the very first loss can be saved. We also show that SACK can improve performance even with window sizes as small as four packets and in situations where acknowledgment packets are lost     

Improving reliable transport and handoff performance in cellular wireless networks

TCP is a reliable transport protocol tuned to perform well in traditional networks where congestion is the primary cause of packet loss. However, networks with wireless links and mobile hosts incur significant losses due to bit-errors and handoffs. This environment violates many of the assumptions made by TCP, causing degraded end-to-end performance. In this paper, we describe the additions and modifications to the standard Internet protocol stack (TCP/IP) to improve end-to-end reliable transport performance in mobile environments. The protocol changes are made to network-layer software at the base station and mobile host, and preserve the end-to-end semantics of TCP. One part of the modifications, called the snoop module, caches packets at the base station and performs local retransmissions across the wireless link to alleviate the problems caused by high bit-error rates. The second part is a routing protocol that enables low-latency handoff to occur with negligible data loss. We have implemented this new protocol stack on a wireless testbed. Our experiments show that this system is significantly more robust at dealing with unreliable wireless links than normal TCP; we have achieved throughput speedups of up to 20 times over regular TCP and handoff latencies over 10 times shorter than other mobile routing protocols.This work was supported by ARPA Contract J-FBI-93-153. This paper was in part presented at the ACM Mobile Computing and Networking Conference (Mobicom '95), Berkeley, California, 14–15 November 1995.     

A stochastic model for the behavior of multiple TCP NewReno sources over optical burst switching network

This work proposes a stochastic model to characterize the transmission control protocol (TCP) over optical burst switching (OBS) networks which helps to understand the interaction between the congestion control mechanism of TCP and the characteristic bursty losses in the OBS network. We derive the steady-state throughput of a TCP NewReno source by modeling it as a Markov chain and the OBS network as an open queueing network with rejection blocking. We model all the phases in the evolution of TCP congestion window and evaluate the number of packets sent and time spent in different states of TCP. We model the mixed assembly process, burst assembler and disassembler modules, and the core network using queueing theory and compute the burst loss probability and end-to-end delay in the network. We derive expression for the throughput of a TCP source by solving the models developed for the source and the network with a set of fixed-point equations. To evaluate the impact of a burst loss on each TCP flow accurately, we define the burst as a composition of per-flow-bursts (which is a burst of packets from a single source). Analytical and simulation results validate the model and highlight the importance of accounting for individual phases in the evolution of TCP congestion window.     

Performance evaluation of TCP extensions on ATM over high bandwidthdelay product networks

Practical experiments in a satellite network environment assist in the design and understanding of future global networks. This article describes the practical experiences gained from TCP/IP on ATM networks over a high-speed satellite link and presents performance comparison studies of such networks with the same host/traffic configurations over local area and wide area networks. These comparison studies on the LAN, WAN, and satellite environments increase our understanding of the behavior of high-bandwidth networks. NASA's Advanced Communications Technology Satellite (ACTS), with its special characteristics and high data rate satellite channels, and the ACTS ATM Internetwork (AAI) were used in these experiments to deliver broadband traffic. Network performance tests were carried out using application-level software (Netspec) on SONET OC-3 (155.52 Mb/s) satellite links. Finally, we experimentally study the performance, efficiency, fairness, and aggressiveness of TCP Reno, TCP New Reno, and TCP SACK end hosts on ATM networks over high BDP networks     

FAST TCP over optical burst switched networks: Modeling and stability analysis

FAST TCP is important for promoting data-intensive applications since it can cleverly react to both packet loss and delay for detecting network congestion. This paper provides a continuous time model and extensive stability analysis of FAST TCP congestion-control mechanism in bufferless Optical Burst Switched Networks (OBS). The paper first shows that random burst contentions are essential to stabilize the network, but cause throughput degradation in FAST TCP flows when a burst with all the packets from a single round is dropped. Second, it shows that FAST TCP is vulnerable to burst delay and fails to detect network congestion due to the little variation of round-trip time, thus unstable. Finally it shows that introducing extra delays by implementing burst retransmission stabilizes FAST TCP over OBS. The paper proves that FAST TCP is not stable over barebone OBS. However, it is locally, exponentially, and asymptotically stable over OBS with burst retransmission.     

TCP Window-Based Flow-Oriented Dynamic Assembly Algorithm for OBS Networks

In transport control protocol (TCP) over optical burst switching (OBS) networks, TCP window size and OBS parameters, including assembly period and burst dropping probability, will impact the network performance. In this paper, a parameter window data dropping probability(WDDP), is defined to analyze the impact of the assembly and the burst loss on the network performance in terms of the round trip time and the throughput. To reduce the WDDP without introducing the extra assembly delay penalty, we propose a novel TCP window based flow-oriented assembly algorithm dynamic assembly period (DAP). In the traditional OBS assembly algorithms, the packets with the same destination and class of service (CoS) are assembled into the same burst, i.e., the packets from different sources will be assembled into one burst. In that case, one burst loss will influence multiple TCP sources. In DAP, the packets from one TCP connection are assembled into bursts, which can avoid the above situation. Through comparing the two consecutive burst lengths, DAP can track the variation of TCP window dynamically and update the assembly period for the next assembly. In addition, the ingress node architecture for the flow-oriented assembly is designed. The performance of DAP is evaluated and compared with that of fixed assembly period (FAP) over a single TCP connection and multiple TCP connections. The results show that DAP performs better than FAP at almost the whole range of burst dropping probability.     

TCP拥塞控制机制浅析

TCP是当今网络中主要的传输协议,它采用了慢启动、拥塞避免、快速重传、快速恢复四种算法,能满足IP网络中数据的可靠传输。但是当出现多个数据包丢失时,由于TCP采用了累计确认机制,造成系统吞吐量下降。文章介绍了一种SACK拥塞控制机制,与传统的Tahoe、Reno对比,并通过仿真实验说明了SACK是一种最好的TCP恢复机制。     

光突发交换网络中TCP Reno的性能分析

该文在分析光突发交换(OBS)网络对TCP性能影响的基础上,研究了单个突发所包含的属于同一TCP/ IP连接的分组数对TCP Reno吞吐量性能的影响,得到了一个吞吐量与突发丢失率、单个突发所包含分组数以及往返时延(RTT)的闭合表达式;并通过仿真验证了分析的正确性;分析和仿真结果表明,在接入链路带宽较大时,突发所包含的分组数存在一个最佳值,使TCP吞吐量达到最大。     

TFRC friendliness and the case of ECN

The TFRC protocol has been proposed as a TCP‐friendly protocol to transport streaming media over the Internet. However, its deployment is still questionable because it has not been compared to other important protocols, analysed in the presence of important mechanisms, such as the explicit congestion notification (ECN), and studied under more realistic network conditions. In this paper, we address these three aspects, including other congestion control protocols not considered before in the same investigation, such as TCP Tahoe, Reno, Newreno, Vegas, Sack, GAIMD, and the Binomial algorithms, the effect of using ECN in the friendliness of the protocols, and the fairness of the protocols under static and dynamic network conditions. We found that TFRC can be safely deployed in the Internet if competing with TCP Tahoe, New Reno and SACK since fairness is achieved under all scenarios considered. We also found that ECN actually helps in achieving better fairness. However, fairness problems arise when TFRC competes with TCP Reno, GAIMD, SQRT or IIAD in static or dynamic conditions, or both. We used normalized throughput, fairness index, and convergence time as the main performance metrics for comparison. Copyright © 2004 John Wiley & Sons, Ltd.     

JTCP: jitter-based TCP for heterogeneous wireless networks

Transmission control protocol (TCP), a widely used transport protocol performs well over the traditional network which is constructed by purely wired links. As wireless access networks are growing rapidly, the wired/wireless mixed internetwork, a heterogeneous environment will get wide deployment in the next-generation ALL-IP wireless networks. TCP which detects the losses as congestion events could not suit the heterogeneous network in which the losses will be introduced by higher bit-error rates or handoffs. There exist some unsolved challenges for applying TCP over wireless links. End-to-end congestion control and fairness issues are two significant factors. To satisfy these two criteria, we propose a jitter-based scheme to adapt sending rates to the packet losses and jitter ratios. The experiment results show that our jitter-based TCP (JTCP) conducts good performance over the heterogeneous network.     

A time-dominated TCP congestion control over heterogeneous networks

The traditional transmission control protocol (TCP) suffers from performance problems such as throughput bias against flows with longer packet roundtrip time (RTT), which leads to burst traffic flows producing high packet loss, long delays, and high delay jitter. This paper proposes a TCP congestion control mechanism, TD-TCP, that the sender increases the congestion window according to time rather than receipt of acknowledgement. Since this mechanism spaces out data sent into the network, data are not sent in bursts. In addition, the proposed mechanism reduces throughput bias because all flows increase the congestion window at the same rate regardless of their packet RTT. The implementation of the mechanism affects only the protocol stack at the sender; hence, neither the receiver nor the routers need modifications. The mechanism has been implemented in the Linux platform and tested in conjunction with various TCP variants in real environments. The experimental result shows that the proposed mechanism improves transmission performance, especially in networks with congestion and/or high packet loss rates. Experiments in real commercial wireless networks have also been conducted to support the proposed mechanism's practical use. Copyright © 2008 John Wiley & Sons, Ltd.