Reducing transport latency for short flows with multipath TCP
Reducing transport latency for short flows with multipath TCP
复制标题
DOI：
10.1016/j.jnca.2018.02.005
复制
发表时间：
2018-04
期刊：
影响因子：
--
通讯作者：
Pingping Dong;Wenjun Yang;Wen-sheng Tang;Jiawei Huang;Haodong Wang;Yi Pan;Jianxin Wang
中科院分区：
文献类型：
--
作者：
Pingping Dong;Wenjun Yang;Wen-sheng Tang;Jiawei Huang;Haodong Wang;Yi Pan;Jianxin Wang
多路径 TCP (MPTCP) 是一种新兴的传输协议，它通过将数据流拆分为跨所有可用多路径的多个子流来提供网络故障恢复能力并提高吞吐量。虽然 MPTCP 通常有利于具有大量子流的吞吐量敏感的大流，但对于延迟敏感的小流可能有害。 MPTCP 为每个子流分配一个拥塞窗口，当流仅包含几个数据包时，短流容易超时。当路径具有异构特征时，这种情况会变得更糟，因为发生数据包重新排序，并且慢速路径可以与 MPTCP 一起使用，导致端到端延迟增加和应用程序良好吞吐量降低。因此，为每个 MPTCP 连接选择合适的子流以获得良好的性能非常重要。然而，MPTCP 中的子流是在连接建立之前确定的，并且它们通常在该连接的生命周期内保持不变。为了解决这个问题，我们提出了 DMPTCP，它根据应用程序工作负载动态调整子流。具体来说，DMPTCP首先利用TCP建模的思想来估计调度路径上的延迟以及同时在其他路径上发送的数据量，然后周期性地决定用于特定应用的子流集合，其目标是减少短流的完成时间并为长流实现更高的吞吐量。我们在 Linux 服务器中实现了 DMPTCP，并在 NS3 和 Linux 测试平台中进行了广泛的实验以验证其有效性。我们的评估表明，与传统 MPTCP 相比，DMPTCP 将短流的完成时间缩短了 46.55% 以上，同时将长寿命流的 Goodput 提高了 21.3%。
Multipath TCP (MPTCP) has been an emerging transport protocol that provides network resilience to failures and improves throughput by splitting a data stream into multiple subflows across all the available multiple paths. While MPTCP is generally beneficial for throughput-sensitive large flows with large number of subflows, it may be harmful for latency-sensitive small flows. MPTCP assigns each subflow a congestion window, making short flows susceptible to timeout when a flow only contains a few packets. This condition becomes even worse when the paths have heterogeneous characteristics as packet reordering occurs and the slow paths can be used with MPTCP, causing the increased end-to-end delay and the lower application Goodput. Thus, it is important to choose the appropriate subflows for each MPTCP connection to achieve the good performance. However, the subflows in MPTCP are determined before a connection is established, and they usually remain unchanged during the lifetime of that connection. To address this issue, we propose DMPTCP, which dynamically adjusts the subflows according to application workloads. Specifically, DMPTCP first utilizes the idea of TCP modeling to estimate the latency on the path under scheduling and the data amount sent on the other paths simultaneously, and then decides the set of subflows to be used for certain application periodically with the goal of reducing completion time for short flows and achieving a higher throughput for long flows. We implement DMPTCP in a Linux server and conduct extensive experiments both in NS3 and in Linux testbed to validate its effectiveness. Our evaluation shows that DMPTCP decreases the completion time by over 46.55% compared to conventional MPTCP for short flows while increases the Goodput up to 21.3% for long-lived flows.