An Overview of Packet Reordering in Transmission Control Protocol (TCP): Problems, Solutions, and Challenges

Dept. of Electr. & Electron. Eng., Hong Kong Univ.
IEEE Transactions on Parallel and Distributed Systems (Impact Factor: 2.17). 05/2007; 18(4):522-535. DOI: 10.1109/TPDS.2007.1011
Source: IEEE Xplore


Transmission control protocol (TCP) is the most popular transport layer protocol for the Internet. Due to various reasons, such as multipath routing, route fluttering, and retransmissions, packets belonging to the same flow may arrive out of order at a destination. Such packet reordering violates the design principles of some traffic control mechanisms in TCP and, thus, poses performance problems. In this paper, we provide a comprehensive and in-depth survey on recent research on packet reordering in TCP. The causes and problems for packet reordering are discussed. Various representative algorithms are examined and compared by computer simulations. The ported program codes and simulation scripts are available for download. Some open questions are discussed to stimulate further research in this area

  • Source
    • "In wireless networks, the sender side estimate for RTT and the ack-arrival time may fluctuate greatly due to various factors like transmission errors, channel contention and frequent route re-establishments [4] and it may not correctly infer to the network congestion all the time. Other factors such as TCP coarse-grained clocks, ack-clustering and ack-compression, delayed acknowledgments and route asymmetry between forward and reverse path also pose challenges to the accuracy for Bandwidth Estimation [31]. With this constraint, network estimate based on the statistics obtained for a period limited by RTT and utilizing the same for adjusting TCP's congestion control parameters during loss recovery is indecorous. "
    [Show abstract] [Hide abstract]
    ABSTRACT: This paper presents a sender side only TCP mechanism to prevent compromise for bandwidth utilization in IEEE 802.11 wireless networks. In absence of mechanism for accurate and immediate loss discrimination, the TCP sender unnecessarily reduces its Loss Window in response to the packet losses due to transmission errors. At the same time, frequent transmission losses and associated link retransmissions cause inaccuracy for available bandwidth estimate. The proposal, Adaptive TCP tackles the above issues using two refinements. First, sender estimates the degree of congestion by exploiting the statistics for estimated Round Trip Time (RTT). With this, it prevents unnecessary shrinkage of Loss Window and bandwidth estimate. Second, by concluding the uninterrupted evolution of its sending rate in recent past, the Adaptive TCP advances bandwidth estimate under favorable network conditions. This in turn, facilitates for quick growth in TCP’s sending rate after loss recovery and consequently alleviates bandwidth utilization. The authors implement the algorithm on top of TCP NewReno, evaluate and compare its performance with the wireless TCP variants deployed in current Internet. Through intensive simulations it is demonstrated that the Adaptive TCP outperforms other well-established TCP variants, and yields more than 100% of the throughput performance and more than 60% of improvement for bandwidth utilization, compared to TCP NewReno. The simulation results also demonstrated compatibility of Adaptive TCP in a shared wireless environment.
    Full-text · Article · May 2015 · International Journal of Communications, Network and System Sciences
  • Source
    • "As a result, congestion control is one of the extensively studied are as in the Internet research conducted over the last 20 years, and a number of proposals aimed at improving various aspects of the congestion-responsive data flows is very large. Several groups of these proposals have been studied in AlHanbali et al. (2005) (congestion control in adhoc networks), Lochert et al. (2007) (congestion control for mobile adhoc networks), Widmer et al. (2001) (congestion control for non-TCP protocols), (Balakrishnan et al. (1997) (congestion control for wireless networks), Leung et al. (2007) "
    [Show abstract] [Hide abstract]
    ABSTRACT: During last two decades researchers, scholars and students are continuously embracing and improving TCP congestion performance both in wired and wireless networks by focusing on four modules of congestion control algorithms i.e., slow start, congestion avoidance, fast recovery, and fast retransmit, which are considered to be the integrated models for network congestion. This paper presents the creativity to collect and classify bibliography on different flavors TCP/IP congestion control during these two decades. We have extracted some core results from the bibliography provided here which are described in the form of tables and diagrams.
    Full-text · Article · Jan 2013 · Journal of Network and Computer Applications
  • Source
    • "First of the mentioned problems, out-of-order delivery is owed to the fact that TCP data segments are carried over multiple paths with different propagation delays. TCP performance suffers from out-of-order packets in different ways [12]. The most important one is that out-of-order arrivals at the receiver cause the generation of duplicate acknowledgements (DUPACKs), what in turn unnecessarily triggers the fast retransmission/recovery algorithms [13]. "
    [Show abstract] [Hide abstract]
    ABSTRACT: The idea of using multiple paths to transport TCP traffic seems very attractive due to the potential benefits it may offer. Despite a solid theoretical background and quite substantial implementation attempts, none of the provided proposals was successful enough to fully establish its mark. Therefore, this paper presents and develops a novel approach to carry TCP traffic over multiple paths. The main features of the proposed architecture include: (1) end-to-end transparency, (2) autonomy of the operation (protocol stack transparency), and (3) interoperability with different operating systems (OSs). To demonstrate the potential of the presented architecture, experimental results are provided in one of possible application scenarios, a high capacity local mesh backbone network with bottleneck gateway links.
    Full-text · Conference Paper · Jun 2012
Show more