Accurately detecting super host that establishes connections to a large number of distinct peers is significant for mitigating web attacks and ensuring high quality of web service. Existing sketch-based approaches estimate the number of distinct connections called flow cardinality according to full IP addresses, while ignoring the fact that a malicious or victim super host often communicates with hosts within the same subnet, resulting in high false positive rates and low accuracy. Though hierarchical-structure based approaches could capture flow cardinality in subnet, they inherently suffer from high memory usage. To address these limitations, we propose SegSketch, a segmented cardinality estimation approach that employs a lightweight halved-segment hashing strategy to infer common prefix lengths of IP addresses, and estimates cardinality within subnet to enhance detection accuracy under constrained memory size. Experiments driven by real-world traces demonstrate that, SegSketch improves F1-Score by up to 8.04× compared to state-of-the-art solutions, particularly under small memory budgets.

Sketch is widely used in many traffic estimation tasks due to its good balance among accuracy, speed, and memory usage. In scenarios with priority flows, priority-aware sketch, as an emerging method, provides differentiated detection accuracy for flows of different priorities, optimizing resource allocation and improving the detection accuracy of high-priority flows. However, existing priority-aware sketches methods struggle to effectively handle the dynamic changes in flow priority distribution in realworld detection environments, leading to wasted or insufficient storage space. To address this issue, this paper proposes a new priority-aware sketch with Dynamic Memory Allocation called DMA-Sketch. It dynamically adjusts the detection framework based on flow priority distribution information and adaptively allocates appropriate memory space to each storage region. The experimental results show that DMA-Sketch improves the overall priority accuracy, high-priority accuracy and throughput by up to 1.33×, 16.39× and 1.88×, respectively, under the scenarios with changing flow priority distribution over the state-of-the-art schemes.

Remote Direct Memory Access (RDMA) over Converged Ethernet (RoCE) cooperating with Priority Flow Control (PFC) has been widely deployed in production datacenters to enable low latency, lossless transmission. At the same time, modern datacenters typically offer parallel transmission paths between any pair of end-hosts, underscoring the importance of load balancing. However, the well-studied load balancing mechanisms designed for lossy datacenter networks (DCNs) are ill-suited for such lossless environments. Through extensive experiments, we are among the first to comprehensively inspect the interactions between PFC and load balancing, and uncover that existing fine-grained rerouting schemes can be counterproductive to spread the congested flows among more paths, further aggravating PFC’s head-of-line (HoL) blocking. Motivated by this, we present FLB, a Fine-grained Load Balancing scheme for lossless DCNs. At its core, FLB employs threshold-free rerouting to effectively balance traffic load and improve link utilization during normal conditions and leverages timely congested flow isolation to eliminate HoL blocking on non-congested flows when congestion occurs. To handle complex multi-bottleneck scenarios, we further introduce FLB*, which incorporates an enhanced congestion-point-aware isolation mechanism using Congestion Point Identifiers (CPI) to eliminate HoL blocking among different congested flows.We have fully implemented a FLB prototype, and our evaluation results show that FLB reduces PFC PAUSE rate by up to 96% and avoids HoL blocking, translating to up to 45% improvement in goodput over CONGA+DCQCN and 40%, 36%, 29% and 18% reduction in average flow completion time (FCT) over LetFlow+Swift, MP-RDMA, Proteus+DCQCN and LetFlow+PCN, respectively.

Sketch has gained wide deployment and application for approximate flow estimation, because of its ability to maintain good accuracy and high throughput with limited memory resources. However, most existing sketch approaches ignore the distinctions between flow priorities, though the high-priority flows are relatively scarce but hold significant information. Therefore, a class of priority-aware sketches has appeared recently to provide differentiated measurement accuracy for flows with different priorities. Unfortunately, it is challenging for these priority-aware sketches to strike a good balance between accuracy and throughput. To address this issue, we propose a priority-adaptive architecture PA-Sketch, which utilizes priority-aware hash to dynamically allocate appropriate numbers of hash functions for different flows according to their priorities. Moreover, to further achieve good accuracy in extremely small memory space, we introduce the priority-aware sampling into PA-Sketch. The test results show that PA-Sketch significantly improves accuracy while minimizing the hash overhead. Compared to the state-of-the-art priority-aware sketches, PA-Sketch reduces the ARE of high-priority flows by 86% and improves the F1 score by 1.83 times, meanwhile maintaining slight accuracy loss for low-priority flows.

In recent years, end-to-end congestion control algorithms or flow pausing mechanisms are proposed to achieve high throughput and low latency in datacenter networks. However, prior end-to-end congestion control works without complex signals fail to achieve fast convergence to a stable equilibrium state and effectively handle the transient congestion, while existing flow pausing mechanisms are decoupled from congestion control, which leads to long convergence time after transient states and incomplete queue elimination in equilibrium states. To address these issues, we present FAR, a rate control protocol that combines the advantages of flow pausing and congestion control. At its heart, FAR couples the bandwidth-estimation-based congestion control and the end-to-end flow pausing mechanisms. After flow pausing, FAR quickly explore the available bandwidth with a binary-search probe to achieve high throughput and low latency. Meanwhile, FAR employs a probe staggering mechanism to address the queue oscillation issue in high-concurrency scenarios. We implement the prototype of FAR using DPDK. Extensive evaluation results demonstrate that our protocol achieves accurate bandwidth estimation and reduces the tail flow completion time (FCT) by up to 67% compared with the state-of-the-art designs.

With the growing ubiquity of video streaming, ensuring a fair and high quality of experience (QoE) for users has emerged as a shared concern among video content providers. State-of-the-art video delivery systems achieve QoE fairness through bottleneck bandwidth allocation across multiple video streams, all based on the assumption of a unified congestion control (CC) protocol. However, the widespread use of heterogeneous CC protocols on the Internet not only disrupts QoE fairness among video streams but also poses challenges in achieving fast convergence under dynamic bandwidth. To address these issues, we propose a QoE-Fairnessawarebandwidthallocationmechanism called Fabam, which establishes a unified QoE control plane across heterogeneous CC protocols. Fabam constructs independent virtual targets based on the real-time QoE of each video stream to achieve QoE fairness, and offers rapid convergence for the underlying CC protocols to improve efficiency. In addition, we propose a Deep Neural Network (DNN)-based multi-step mapping model aimed at balancing the performance and overhead of Fabam, thereby enhancing its deployment potential in practical applications. We implement Fabam on QUIC and integrate it with Dash.js. The evaluation results demonstrate the significant superiority of Fabam over the state-of-the-art approaches, including an enhancement of 44.01% in QoE fairness and an improvement of 36.39% in QoE efficiency. Meanwhile, Fabam-DNN maintains satisfactory QoE fairness while supporting multiple users at a low cost.

Synchronous stochastic gradient descent (SSGD) has been widely used in distributed deep learning. However, since the local gradients need to be shared among workers at every iteration, SSGD performance is significantly influenced by network bottlenecks caused by either heterogeneous environment or bandwidth contention. To solve this problem, asynchronous parallel (ASP) strategy allows each worker to update parameters independently without synchronization, while suffering from accuracy loss and convergence inefficiency. In this paper, we propose a novel similarity-based synchronization scheme called SIM, which mitigates the impact of network bottlenecks and ensures convergence efficiency. Specifically, SIM reduces the number of aggregation workers based on the gradient similarity between global and local gradients, therefore shrinking the waiting time for the stragglers. We provide a theoretical analysis of convergence efficiency and conduct large-scale testbed experiments on CIFAR-10 and SQUAD dataset. The experimental results show that SIM reduces the convergence time of four classical deep learning models by up to 40%.

Time-Sensitive Networking (TSN) is the most promising network infrastructure for various time-critical applications in Industry 4.0. However, industry applications generate a mix of time-triggered (TT) and event-triggered (ET) flows. Scheduling such mix-flows is a key challenge for TSN. Though current TSN scheduling mechanisms commonly provide deterministic transmission for TT flows with stringent latency requirements, they cannot flexibly accommodate ET flows, which are usually generated by emergency events. In this paper, we propose Elastic Backoff (EBO), a systematic solution for scheduling mix-flows in an elastic way. Our key insight is that the network resources should be reasonably allocated for ET flows while minimally impacting TT flows. To this end, we incorporate the elasticity into the TSN scheduling to make resource reservations for ET flows without hurting TT flows. We conduct extensive experiments on both testbeds and simulations. The evaluation results show that, compared with the state-of-the-art designs, EBO improves the schedulability of ET flows by up to 7.5×, while still ensuring the deterministic transmission of TT flows.

Quantiles over data stream have been recognized to be an essential feature in network traffic. To provide accurate estimation results, current quantile estimation approaches are preconfigured according to traffic distributions such as log-normal and Pareto distributions. In most practical applications, however, the traffic distributions are not known a priori or are highly dynamic, disturbing estimation results. In this paper, we propose SplitSketch, a sketch-based mechanism that aims to accurately estimate quantiles over data stream without any prior knowledge of traffic distributions. SplitSketch adjusts its estimation granularity according to the changing process of the traffic distribution. The granularities with denser distribution will be recorded with the finer granularity to provide more accurate estimation results. Experimental results demonstrate that, compared to existing approaches, SplitSketch reduces the absolute error in quantile estimation by 59.1% for heavy-tailed distributions and 79.2% for general distributions.

Sketch-based measurement approaches have recently become a promising solution for detecting periodic flows. However, current sketch approaches struggle to achieve accurate detection of periodic flows due to their short-sighted record of the flow arrival information. Recording the long-term information of flow arrivals could mitigate this issue, while the large memory consumption will hurt the detection accuracy. Consequently, achieving a satisfactory trade-off between memory efficiency and detection accuracy remains a tough challenge. To address this issue, we propose Aion for periodic flow detection. Specifically, Aion uses the Sidon sequence to compress historical flow arrival information in multiple time windows into very small size. Based on the periodicity information from successive time windows, Aion updates the estimated frequencies of periodic flows and promptly evicts non-periodic ones to enable accurate detection of frequent periodic flows. We implement Aion on a P4-based testbed and demonstrate that it achieves superior resource efficiency compared to state-of-the-art approaches. Trace-driven evaluations show that Aion improves F1-Score by up to 9.88×, particularly under small memory conditions.

As the foundational components of large-scale applications, distributed key-value stores must respond to user requests quickly. However, a user request typically comprises multiple key-value access operations, which are processed in parallel across different servers, and the response time is determined by the slowest operation. To reduce the mean response time of requests, existing approaches schedule the sequence of key-value access operations across different servers so that all operations of a request complete at approximately the same time. Nevertheless, all of these approaches operate in a distributive manner, and their theoretical performance boundaries are unknown. To address these issues, we designed SDN-KVS (Key-Value Scheduler based on Software Defined Network), which migrates the waiting queue of key-value access operations from overloaded servers to the SDN controller. In this way, SDN-KVS centrally schedules the requests from different clients to overloaded servers without extra latency overhead. The scheduling result is proven to be$(1+2 \eta)$-approximation, i.e., the mean response time of requests is smaller than ($1+2 \eta$) times of the optimal value, where$\eta$is the parameter to make a trade-off between mean and tail response time. Simulation results confirm the excellent performance of the SDN-KVS algorithm. Specifically, SDN-KVS outperforms existing algorithms up to 37.6% and 79.8% in terms of mean and tail response time, respectively.

Distributed deep learning has been widely deployed in data centers to provide various services such as image classification and speech recognition. To reduce the training time, Top-k compression has become one of the most popular solutions used to shrink the data volume of gradients. Nevertheless, we observe that existing Top-k compression solutions are inefficient when used for large-scale distributed training due to gradient build-up, missing of Top-k gradients, and high compression overhead at the end hosts. To address these problems, we propose SwitchTop-k, which improves the accuracy of selecting Top-k values while ensuring a high compression rate and zero compression overhead. Specifically, SwitchTop-k offloads the Top-k compression from the end hosts to the programmable switches, thus alleviating the gradient build-up and compression overhead. Meanwhile, we propose a sketch-based solution to achieve high accuracy in selecting global Top-k gradients. We also co-design switch logic and end host logic to improve communication efficiency of uncompressed traffic. Finally, we implement SwitchTop-k on Intel Tofino switches and integrate it with Pytorch. The test results show that SwitchTop-k reduces iteration time by up to 91% compared with existing compression algorithms.

The rise of multiaccess edge computing (MEC) speeds up mobile user services and resolves service delays caused by long-distance transmission to cloud servers. However, in task-intensive scenarios, edge server processing limitations lead to buffer congestion, increasing latency and reducing Quality of Service (QoS). Furthermore, the challenges of edge server task processing are increased by the varying deadline requirements of different tasks, the time variability of task arrivals, and the real-time fluctuations of the network. In this work, we propose an adaptive slicing-based task admission scheduling strategy (ASTA) to address these issues. ASTA consists of an adaptive time slice adjustment algorithm (ASTA-I) and a task admission scheduling algorithm (ASTA-II). ASTA-I dynamically adjusts time slices based on real-time network conditions and task flow. ASTA-II first adjusts task priorities dynamically by considering factors, such as data volume, deadlines, network conditions, and buffer locations. After that, ASTA-II formulates different scheduling strategies based on changes in task priorities. These strategies are formulated to improve the throughput efficiency of edge servers and enhance the average response speed of tasks. Simulation results show that compared with the existing O2A and OTDS in different scenarios, the proposed ASTA can reduce the average number of waiting requests in the edge server buffer by 19.53%–57.73% and 20.42%–50.26%, and accelerates the average response speed of tasks by about 39.76% and 32.41%.

With the rapid growth scale of dataset and model, the training of deep neural networks (DNN) tends to be deployed in a distributed manner. In the large-scale distributed training, the bottlenecks have gradually moved from computational resources to communication process. Recent researches adopt in-network aggregation (INA) that offloads the gradient aggregation process to programmable switches, thereby reducing network traffic amount and transmission latency. Unfortunately, due to the bandwidth competition in shared training clusters, the straggler will slow down the training efficiency of INA. To address this issue, we propose an Asynchronous Control based Aggregation Transport Protocol (AC-ATP), which makes full use uncongested links to transmit gradients and the switch memory to cache gradients from the fast workers to accelerate the gradient aggregation. Meanwhile, AC-ATP performs congestion control according to the transmission progress of worker and the remaining completion time of the job. The evaluation results of real testbed and large-scale simulations show that AC-ATP reduces the aggregate time by up to 68% and speeds up training in real-world benchmark models.

To meet the stringent requirements of industrial applications, modern Ethernet datacenter networks widely deployed with remote direct memory access (RDMA) technology and priority-based flow control (PFC) scheme aim at providing low latency and high throughput transmission performance. However, the existing end-to-end congestion control cannot handle the transient congestion timely due to the round-trip-time (RTT) level control loop, inevitably resulting in PFC triggering. In this article, we propose a Sub-RTT congestion control mechanism called SRCC to alleviate bursty congestion timely. Specifically, SRCC identifies the congested flows accurately, notifies congestion directly from the hotspot to the corresponding source at the sub-RTT control loop and adjusts the sending rate to avoid PFC's head-of-line blocking. Compared to the state-of-the-art end-to-end transmission protocols, the evaluation results show that SRCC effectively reduces the average flow completion time (FCT) by up to 61%, 52%, 40%, and 24% over datacenter quantized congestion notification (DCQCN), Swift, high precision congestion control (HPCC), and photonic congestion notification (PCN), respectively.

To meet the stringent demanding low latency and high throughput of cloud datacenter applications, recent receiver-driven transport protocols transmit only one packet once receiving each credit packet from the receiver to achieve ultra-low queueing delay. However, the round-trip time variation and the highly dynamic background traffic significantly deteriorate the performance of receiver-driven transport protocols, resulting in under-utilized bandwidth. This paper designs a simple yet effective solution called RPO, which retains the advantages of receiver-driven transmission while efficiently utilizing the available bandwidth. Specifically, RPO rationally uses low-priority opportunistic packets to ensure high network utilization without increasing the queueing delay of high-priority normal packets. Furthermore, to tackle the queueing buildup due to line-rate transmission in the first RTT, we design a selective dropping mechanism called SDM to help the majority of small flows complete within only one RTT by prioritizing the first-RTT bursty packets over the packets triggered by grants. We implement RPO in Linux hosts with DPDK. The experimental results show that RPO significantly improves the network utilization by up to 35% over the state-of-the-art schemes, without introducing additional queueing delay. Moreover, RPO integrated with SDM reduces the AFCT of small flows by up to 45% compared with RPO integrated with Aeolus.

Nowadays, Congestion Control (CC) algorithms are expected to satisfy the diverse demands of applications running over diverse networks. To achieve this goal, the combinations, which are expected to inherit both the advantages of classic CC in terms of convergence, overhead, and explainability, and the advantages of learning-based CC on adapting to diverse networks and demands, become a hot topic. In this paper, we reveal the existing combination works are eithergiving a fishorteaching to fish. Based on the insight of their essential issues, we develop the Concentrator method ofteaching to fish. According to this method, we propose Seagull as a step further. Specifically, Seagull captures the network characteristics and application demands in a coarse-grained manner via an online learning module. Moreover, the online learning module guides the customization of the rate adjustment rules of the classic CC module for fine-grained system evolution. Replacing the assumption on networks by the captured characteristics, the classic CC module of Seagull can fulfill the specified application demands. Real-world experimental results show Seagull respectively outperforms Orca, PCC-Vivace, and CUBIC by$49.3\%,\ 30.4\%$, and 24.9% in terms of throughput over the Internet, and improves the video quality of experience (QoE) by$12.9\sim 33.5\%$compared to CUBIC over cellular links.

Large-scale machine learning typically adopts distributed machine learning (DML) techniques to accelerate model training. Due to the large communication overhead, unfortunately, the phase of gradient aggregation has become the performance bottleneck for data-parallel DML. To reduce traffic volume, several in-network aggregation (INA) transmission protocols are proposed to offload gradient aggregation function into the programmable switches. However, since existing INA transmission protocols use synchronous congestion control mechanism to drive each round of gradient aggregation, the straggling workers lead to long iteration time and significant performance degradation. Besides, we reveal that existing INA solutions cannot provide the fairness performance among multiple jobs with varying number of workers. To solve the above problem, we propose PA-ATP, a progress-aware INA transmission protocol, which adopts the progress-aware asynchronous congestion control. PA-ATP adjusts the sending rate in accordance with the transmission progress, allowing the straggling flow to grab more bandwidth than the leading flow and control the asynchronous degree of straggling job. Moreover, to ensure the fair throughput among multiple jobs, we dynamically adjust the aggregator allocation for each job by tuning the number of hash operations. We use a P4 programmable switch and a kernel-bypass protocol stack to implement PA-ATP. The results of testbed and large-scale NS3 simulations show that PA-ATP reduces training time by up to 62% compared to the state-of-the-art INA transmission protocols.

For the widely deployed on-chip shared buffer, efficient buffer management is the key to absorbing bursts and avoiding packet loss during transient congestion. However, as the buffer-per-port-per-Gbps in production data centers decreases, it becomes more challenging to provide efficient buffer management to meet the requirements of heterogeneous traffic. We observe that typical shared buffer management policies have two steps: first, they identify short flows arriving at ports and then allocate more buffer room for these ports. Unfortunately, the lack of isolation between long and short flows leads to increased queue buildup and even packet loss of short flows. To address this limitation, we propose D2T, which uses different queue length thresholds for long and short flows. Specifically, we first design a compact data structure to distinguish between long and short flows. Then when two kinds of flows coexist at the same port, the threshold of long flows will decrease to absorb the bursty short flows. What’s more, we introduce D2T${}^{*}$which combines D2T with advanced DRL techniques to move toward mastering buffer management for further improving performance across various scenarios. We implement D2T at a P4-programmable switch and large-scale simulations. The results demonstrate that D2T reduces both average and tail flow completion times (FCT) of short flows by up to 29% and 62% compared with the state-of-the-art policies, respectively.

Each output port in a shared memory switch can compete for shared memory pool resources. The allocation strategy of the shared buffer directly affects the ability of each output port to absorb network traffic. Due to the unpredictability of traditional network traffic, existing switch buffer management strategies take a passive approach, allocating buffers to each port only after traffic arrives. This passive response has the problem of untimely buffer allocation and cannot effectively absorb burst traffic. Distributed deep learning follows a specific training pattern, and network traffic exhibits obvious periodic characteristics during transmission. Thus, we propose a Proactive Dynamic Threshold (PDT) strategy, which realizes the pre-adjustment of switch port threshold by detecting the traffic characteristics of distributed training.

Data augmentation plays a crucial role in AIOps for enhancing the performance of classification models in scenarios with limited supervision. However, current methods used for generating pseudo-anomaly samples may fail in AIOps: existing data augmentation methods suffer from poor sample quality due to class imbalance, high dimensionality, and high diversity. Inspired by the conditional DDPM, we address the problem by generating realistic anomaly samples between normal and abnormal ones. Unfortunately, due to the lack of pre-trained encoders and the difficulty of determining conditional information, it is hard to directly use conditional DDPM. In this work, we present C-Aug which combines sample mixing and conditional diffusion to overcome the above issues. C-Aug respectively achieves F1-Scores of 0.76, 0.98, and 0.90 on three public datasets, which significantly outperforms the other five baselines.

Nowadays the data center switches employ the on-chip shared buffer to absorb bursts and avoid packet loss during transient congestion. However, as the buffer-per-port-per-Gbps in production data centers decreases, it becomes more challenging to provide efficient buffer management to meet the requirements of heterogeneous traffic. We observe that typical shared buffer management policies have two steps: first, they identify short flows arriving at ports and then allocate more buffer room for these ports. Unfortunately, the lack of isolation between long and short flows leads to increased queue buildup and even packet loss of short flows. To address this limitation, we propose D2T, which uses different queue length thresholds for long and short flows. Specifically, we first design a compact data structure to distinguish between long and short flows. Then when two kinds of flows coexist at the same port, the threshold of long flows will decrease to absorb the bursty short flows. We implement D2T at a P4- programmable switch and large-scale simulations. The results demonstrate that D2T reduces both average and tail flow completion times (FCT) of short flows by up to 29% and 62% compared with the state-of-the-art policies, respectively.

Multicast serves as an important approach for one-to-many communication in data center networks. To reduce overhead and improve scalability, bloom filters are employed in current multicast approaches to store forwarding ports of switches. However, the well-known false positive issue of bloom filter incurs wrong forwarding behaviors and redundant traffic in multicast tree, degrading transmission efficiency and increasing the risk of data leakage. Inspired by the fact that, given the same false positive ratio, the switch in the upper layers of multicast tree generates more redundant traffic, we propose RSBF, a fine-grained and resource-aware multicast approach using skewed bloom filters. Specifically, RSBF maintains multiple bloom filters corresponding to different layers of multicast tree, and allocates more ample space to the bloom filter of the upper layer switches, thereby reducing the overall redundant traffic. The test results of large-scale simulation demonstrate that RSBF reduces both redundant traffic and header overhead by up to 64% and 49% compared with the state-of-the-art approaches, respectively.

In modern data centers, many flow scheduling schemes are proposed to accelerate data transfer and improve user experience. However, these schemes assume ideally the prior knowledge of the flow size information, which, unfortunately, is hard to obtain without modifying data center applications. The sketch-based approaches measure the flow size at switch with a compact memory structure, high throughput, and acceptable accuracy loss. However, existing sketches commonly focus on large or specific flows, while most flows in data center networks are small, resulting in missing or overestimated size information about small flows. We propose Strainer Sketch, which enables accurate and fast measurement of small flows with small memory and flexible deployment in a variety of scheduling algorithms. Specifically, Strainer Sketch uses the hierarchical structure to mitigate hash collisions between large and small flows, and the probabilistic counting algorithm to mitigate overestimation due to hash collisions between small flows. Furthermore, we propose a packet scheduling algorithm SW-PIFO, which provides the flow discrimination for a huge number of small flows by using a limited number of queues. Through the testbed experiments and simulations of typical data center applications, we show that our scheme reduces the small flow completion time (FCT) by up to 56.7 <?TeX $\%$?> Math 1 compared with flow scheduling using classic sketches.

To achieve high throughput and low latency for data center applications, there are two broad lines of work: end-to-end congestion control algorithms and flow pausing mechanisms. It is challenging for end-to-end congestion control algorithms without complex signals to achieve fast convergence to a stable equilibrium state while effectively handling the transient congestion. Additionally, flow pausing mechanisms are decoupled from congestion control, which leads to long convergence time after transient state and incomplete queue elimination in equilibrium state. We propose a transport protocol that combines the advantages of flow pausing and congestion control, called FAR. The key idea is coupling the bandwidth-estimation based congestion control and the end-to-end flow pausing mechanisms. FAR quickly explores the available bandwidth with binary-search based packet train probe to achieve high throughput and low latency. Extensive evaluation results demonstrate that our protocol achieves accurate bandwidth estimation and reduces the tail flow completion time (FCT) by up to 67 <?TeX $\%$?> Math 1 compared with the state-of-the-art designs.

Distributed deep learning has been widely employed to train deep neural network over large-scale dataset. However, the commonly used parameter server architecture suffers from long synchronization time in data-parallel training. Although the existing solutions are proposed to reduce synchronization overhead by breaking the synchronization barriers or limiting the staleness bound, they inevitably experience low convergence efficiency and long synchronization waiting. To address these problems, we propose Gsyn to reduce both synchronization overhead and staleness. Specifically, Gsyn divides workers into multiple groups. The workers in the same group coordinate with each other using the bulk synchronous parallel scheme to achieve high convergence efficiency, and each group communicates with parameter server asynchronously to reduce the synchronization waiting time, consequently increasing the convergence efficiency. Furthermore, we theoretically analyze the optimal number of groups to achieve a good tradeoff between staleness and synchronization waiting. The evaluation test in the realistic cluster with multiple training tasks demonstrates that Gsyn is beneficial and accelerates distributed training by up to 27% over the state-of-the-art solutions.

Finding persistent items provides indispensable information for data stream tasks. However, accurately identifying persistent items becomes very challenging with the increasing volume of data streams in memory-constrained environments. Existing solutions for finding persistent items often rely solely on estimating the persistence of items to make replacement decisions, requiring sufficiently large memory for acceptable performance. However, persistent items are frequently erroneously replaced in scenarios with numerous non-persistent items, leading to suboptimal accuracy. To address this issue, we reveal that periodically arriving items provide another useful feature for finding persistent items. We further propose P2Sketch that selectively replaces non-persistent items and preserves persistent items based on multi-dimensional statistics of estimated persistence and periodicity of items. Specifically, P2Sketch leverages the characteristics of persistence and periodic arrival to replace stored items selectively. When multiple candidates map to the same bucket, we replace items with longer periodic intervals and smaller estimated persistence to ensure more persistent items are protected. Experimental results show that P2Sketch significantly improves the F1 score by 1.15x and reduces the ARE by 2.66x under the condition of 50KB of memory compared with the state-of-the-art solutions.

Finding flows that are persistent but do not occur frequently is very important to intrusion detection and network management. Current solutions detect persistent and infrequent (PI) flows according to the accumulative statistics from all time windows. However, PI flows may generate a huge amount of traffic in very short time period. The bursty traffic significantly disrupts the accumulative statistics, leading to misjudgments of PI flows. To address this issue, we propose a novel design called PIB sketch, which finds PI flows based on the global information from all time windows and local information within each independent time window. By filtering out the bursty traffic, PIB sketch avoids the negative effect on measurement accuracy. We conduct large-scale trace-driven test to evaluate PIB sketch. The test results show that PIB effectively improves the F1 score and ARE compared with the state-of-the-art solutions.

The strategy for 40/100Gbps Energy Efficient Ethernet (EEE) determines when to enter and leave the power-saving modes. Accordingly, it directly decides both the energy savings and the incurred latency of frames in the EEE. The performance of the EEE strategy is greatly influenced by the network traffic, and thus existing EEE strategies need either proper parameter configuration under certain traffic loads or parameter adaptation mechanisms based on the traffic prediction under the assumptions of certain distribution. Consequently, these EEE strategies hardly keep consistent high performance under variable traffic in reality. To address this issue, we bring the reinforcement learning method into the design of the EEE strategy and propose the reinforcement learning based periodic strategy (RLPS) in this paper. Specifically, RLPS transmits existing frames at first and then stays in the selected power-saving mode for the rest time in each cycle. Moreover, RLPS learns the time length of each cycle online to reflect the impacts of traffic, instead of directly outputting power-saving mode transition decisions. In this way, the power consumption in each cycle is optimal with the help of learned information, and the overhead of online learning is reduced. Simulations driven by both synthetic traffic and real traces confirm that RLPS outperforms existing strategies, i.e., can achieve consistent high performance regardless of the traffic loads and distributions.

Increasingly large-scale models and rich data sets make communication overhead a key bottleneck for distributed Deep Neural Network (DNN) training, constantly attracting the attention of academia and industry. Despite continuous efforts, prior solutions such as pipelining computation/communication and in-network gradient compression/scheduling do not focus on how to accelerate DNN training through load balancing in datacenter networks (DCNs). However, the existing load balancing mechanisms are unaware of tensor integrity and priority for gradient parameter synchronization during the DNN training iterations, resulting in severe tensor tail latency and slow model convergence speed. In this paper, we present a Tensor-aware Load Balancing (TaLB) scheme to accelerate DNN training. Specifically, TaLB identifies the different priority tensors and makes (re)routing decisions based on the tensor-level granularity to cut the high-priority tensors tail delay. The testbed implementation and large-scale NS-3 simulation results show that TaLB effectively accelerates DNN training speed. For example, TaLB significantly reduces the average flow completion time (FCT) by up to 55%, and accelerates the model training speed up to 2.37× on VGG19, ResNet50 and AlexNet models.

Nowadays, Congestion Control (CC) algorithms based on In-Network Telemetry (INT) are popular in datacenter networks, because INT is supported by many commercial devices and provides comprehensive information about network congestion. In this work, we analyze the recent INT-based CC algorithm PowerTCP and reveal its fairness and large delay issues under the conditions of a large number of flows. Inspired by the analytical results, we propose Trident, which enhances PowerTCP by accelerating the speed of converging to fairness and maintaining a low queuing delay with a small equilibrium point. Simulations based on the open source codes confirm that Trident outperforms existing INT-based CC algorithms HPCC, PowerTCP and Poseidon by 14.3%, 14.8%, and 63.5% in terms of FCT of short flows. Moreover, Trident outperforms existing CC algorithms such as Timely, DCQCN and DCTCP, benefiting from the advantages inherited from PowerTCP.

The collective communication has become the bottleneck of large-scale distributed deep learning due to the huge volume of gradients aggregated during the training process. Despite much recent progress in reducing traffic volume by compressing the stochastic gradients inside each training worker, how to share the inter-worker data redundancy to alleviate communication overhead has remained elusive. In this paper, we reveal that most gradients have a great similarity with close value among training workers. From this hypothesis, we propose a Similarity-guided Gradient Compression framework named SGC which skips aggregating the similar gradients among each worker which utilizes local one rather than average value to save communication expenses. Each worker utilizes local SGC firstly quantifies the similarity of gradients among workers, and then elaborately adjusts the aggregation frequency of similar gradients without hurting DNN model accuracy. Meanwhile, we theoretically analyze the convergency accuracy of SGC. The comprehensive evaluation demonstrates that SGC outperforms the state-of-the-art schemes by up to 47% in convergence time.

With the growing ubiquity of video streaming, ensuring a fair and high quality of experience (QoE) for users has emerged as a shared concern among video content providers. State-of-the-art video delivery systems achieve QoE fairness through bottleneck bandwidth allocation across multiple video streaming, all based on the assumption of a unified congestion control (CC) protocol. However, the widespread use of heterogeneous CC protocols on the Internet not only disrupts QoE fairness among video streaming but also poses challenges in achieving fast convergence under dynamic bandwidth. To address these issues, we propose a QoE-Fairness aware bandwidth allocation mechanism called Fabam, which establishes a unified QoE control plane across heterogeneous CC protocols. Fabam constructs independent virtual targets based on the real-time QoE of each video streaming to achieve QoE fairness, and offers rapid convergence for the underlying CC protocols to improve efficiency. We implement Fabam on QUIC and integrate it with Dash.js. The evaluation results demonstrate the significant superiority of Fabam over the state-of-the-art approaches, including an enhancement of 24.48% in QoE fairness and an improvement of 16.63% in QoE efficiency.

Multi-path Transmission Control Protocol (MPTCP) has shown great potential in improving network bandwidth and robustness by utilizing multiple subflows in data center networks (DCNs). However, the delay and loss heterogeneities of multiple paths potentially cause packet reordering, resulting in the ACK blocking and increased latency. Recent coding-based solutions use forward error correction (FEC) to mitigate path heterogeneity with redundant encoded packets. However, current FEC-based solutions work at the subflow level, that is, each subflow independently generates redundant encoded packets. This intra-subflow coding, however, does not leverage the path diversity, easily suffering from long tail latency. In this article, we propose a new MPTCP based on cyclic matrix coding, called as CM-MPTCP, which encodes packets inter subflow to leverage the path diversity. Specifically, to let good paths help bad ones, the good paths deliver more redundant packets encoded based on cyclic matrix, which gives more coding opportunities to packets on bad paths, thus achieving high packet decoding ratio at the receiver side. The results of large-scale NS2 simulations show that CM-MPTCP effectively mitigates the ACK blocking and reduces the average flow completion time (AFCT) by about$45\% \sim 70\%$under a wide variety of network conditions compared with the state-of-the-art coding-based MPTCPs.

Remote Direct Memory Access (RDMA) achieves ultra-low latency, high throughput and low CPU overhead in data center by implementing the transport logic in hardware network interface card (NIC). However, RDMA faces new challenges in the heterogeneous multipath environment as it is very sensitive to packet reordering. When some packets are blocked in slow paths, the other packets delivered through fast paths have to be buffered at the receiver's NIC, consuming the limited on-chip memory resources. In this paper, we propose a new RDMA-based multipath transmission scheme with advanced fast retransmission called as AFR-MPRDMA. Specifically, once detecting congestion at the slow path, the sender will retransmit the blocked packets on other fast paths to speed up the transmission of blocked packets. Moreover, the receiver dynamically adjusts the buffer size for the out-of-order packets to avoid either unnecessary retransmission or long latency. The results of large-scale tests show that AFR-MPRDMA effectively mitigates packets blocking issue and reduces average flow completion time (AFCT) by up to 61% compared with the state-of-the-art RDMA-based schemes.

Mobile video streaming dominates today's network traffic, and adaptive bitrate (ABR) algorithms have been routinely adopted for transmitting media content across dynamic mobile networks. State-of-the-art ABR algorithms mainly alter video bitrate without considering audio bitrate as they consider the impact on the video negligible due to their small size. However, to bring users an immersive experience, recent content providers have applied high-quality audio with large sizes, like stereophonic sound. Therefore, improper audio bitrate selection will adversely affect video bitrate selection, leading to undesirable audio/video combinations (the highest video quality with the lowest audio quality, and vice versa) and frequent playback interruptions. To address these inefficiencies, we propose a Self-Play reinforcement learning-based Audio-aware ABR algorithm named SPA to learn strategies for audio and video bitrate selections. By learning from explicit goals, SPA can match the actual requirements and attain good performance. By conducting trace-driven and testbed-based experiments, we observe SPA's considerable superiority compared to existing approaches, including reducing the undesirable combinations by up to 34.17× and achieving zero stall time across 88.57% of traces. We also invite 35 volunteers to join a subjective test, and the result shows that 33/35 people consider SPA provides them with a satisfactory viewing experience.

Adaptive Bitrate (ABR) algorithms have become increasingly important for delivering high-quality video content over fluctuating networks. Considering the complexity of video scenes, video chunks can be separated into two categories: those with intricate scenes and those with simple scenes. In practice, it has been observed that improving the quality of intricate chunks yields more substantial improvements in Quality of Experience (QoE) compared with focusing solely on simple chunks. However, the current ABR schemes either treat all chunks equally or rely on fixed linear-based reward functions, which limits their ability to meet real-world requirements. To tackle these limitations, this paper introduces a novel ABR approach called CAST (Complex-scene Aware bitrate algorithm via Self-play reinforcemenT learning), which considers the scene complexity and formulates the bitrate adaptation task as an explicit objective. Leveraging the power of parallel computing with multiple agents, CAST trains a neural network to achieve superior video playback quality for intricate scenes while minimizing playback freezing time. Moreover, we also introduce a new variant of our proposed approach called CAST-DU, to address the critical issue of efficiently managing users' limited cellular data budgets while ensuring a satisfactory viewing experience. Furthermore, we present CAST-Live, tailored for live streaming scenarios with constrained playback buffers and considerations for energy costs. Extensive trace-driven evaluations and subjective tests demonstrate that CAST, CAST-DU, and CAST-Live outperform existing off-the-shelf schemes, delivering a superior video streaming experience over fluctuating networks while efficiently utilizing data resources. Moreover, CAST-Live demonstrates effectiveness even under limited buffer size constraints while incurring minimal energy costs.

Adaptive BitRate (ABR) algorithms have become increasingly prevalent in modern streaming platforms, offering users significant improvements in the Quality of Experience (QoE). With streaming providers like YouTube and Netflix shifting to high-fidelity audio formats such as stereophonic sound and Dolby Atoms, ensuring proper audio and video adaptation has become a critical aspect of modern streaming platforms. Additionally, Variable Bitrate (VBR) encoding has gained great popularity in encoding audio and video content, given its higher quality-to-bits ratio. However, the considerable variability in network bandwidth, in combination with VBR features such as significantly fluctuating audio/video chunk sizes and diverse content complexity, makes existing ABR schemes formidable to make optimal bitrate selection due to their overlook of audio adaptation or oblivious to VBR features. In this paper, we introduce a new ABR approach forVBR-basedAudio-aware videoStrEaming named VASE, which harnesses deep reinforcement learning (DRL) and exploits parallel computing with multiple agents to swiftly and adeptly manage fluctuations in video/audio chunk sizes, network bandwidth, and varying content complexity, all while operating without any assumptions. Besides, two variants are proposed to mitigate the download energy cost and handle audio and video content in finer granularity. Extensive trace-driven, testbed, and subjective evaluations show that our scheme surpasses existing advanced adaptation schemes regarding the overall QoE, effectively demonstrating its superiority.

In tile-based 360° video streaming, the users employ the tile rate allocation algorithm to select appropriate bitrate to maximize the quality of experience (QoE). The preferences and viewports, however, can vary significantly across the different users. Since the users independently choose their bitrate according to their own preferences and viewports, it is hard to ensure QoE fairness for users under the constraint of available bandwidth. In this article, we propose a QoE-fairness aware bitrate allocation algorithm for multi-users (QBAM) to reduce difference of user QoE. According to the trajectory of the user viewpoint and user preferences for video quality, rebuffer time and quality switching, we leverage multi-agent reinforcement learning to train the bitrate allocation strategy. The experimental results show, compared with the current tile rate allocation algorithm, QBAM effectively improves the QoE fairness.

In recent years, motivated by new datacenter applications and the well-known shortcomings of TCP in data center, many receiver-driven transport protocols have been proposed to provide ultra-low latency and zero packet loss by using the proactive congestion control. However, in the scenario of mixed short and long flows, the short flows with ON/OFF pattern generate micro-burst traffic, which significantly deteriorates the performance of existing receiver-driven transport protocols. Firstly, when the short flows turn into ON mode, the long flows cannot immediately concede bandwidth to the short ones, resulting in queue buildup and even packet loss. Secondly, when the short flows change from ON to OFF mode, the released bandwidth cannot be fully utilized by the long flows, leading to serious bandwidth waste. To address these issues, we propose a new receiver-driven transport protocol, called SAR, which predicts the micro burst generated by short flows and adjusts the sending rate of long flows accordingly. With the aid of micro-burst prediction mechanism, SAR mitigates the bandwidth competition due to the arrival of short flows, and alleviates the bandwidth waste when the short flows leave. The testbed and NS2 simulation experiments demonstrate that SAR reduces the average flow completion time (AFCT) by up to 66% compared to typical receiver-driven transport protocols.

Many existing load balancing mechanisms work effectively in lossy datacenter networks (DCNs), but they suffer from serious packet reordering in lossless Ethernet DCNs deployed with the hop-by-hop Priority-based Flow Control (PFC). The key reason is that the prior solutions are not able to perceive PFC triggering correctly and in a timely manner when making load balancing decisions. Once the forwarding path pauses transmission due to PFC triggering, the packets allocated on it are blocked, inevitably leading to out-of-order packets and retransmission. In this paper, we present an Reordering-robust Load Balancing (RLB) scheme with PFC prediction in lossless DCNs. At its heart, RLB leverages the derivative of ingress queue length to predict PFC triggering and proactively notifies the upstream switches to choose an appropriate rerouting path or perform packet recirculation to avoid reordering. Furthermore, under switch failure scenarios, RLB adjusts the recirculation threshold adaptively to mitigate the risk of packets over-recirculation. We have implemented RLB in the hardware programmable switch. As a building block for existing load balancing mechanisms, we have integrated RLB into Presto, LetFlow, Hermes and DRILL. The evaluation results show that the RLB-enhanced solutions deliver significant performance by avoiding packet reordering. For example, it reduces the$99^{th}$percentile flow completion time (FCT) by up to 72%, 67%, 58% and 54% over DRILL, Presto, LetFlow and Hermes, respectively.

Various datacenter network (DCN) load balancing schemes have been proposed in the past decade. Unfortunately, most of these solutions designed for lossy DCNs do not work well for Priority Flow Control (PFC) enabled lossless DCNs, primarily due to the reason that the individual congestion signals used in these solutions, e.g., link load, queue length, Round Trip Time (RTT) and Explicit Congestion Notification (ECN), may not be able to correctly or timely reflect the hop-by-hop PFC pausing. This paper first reveals the above problems via extensive experiments, and then based on the insights learned, we present Proteus, a PFC-aware load balancing scheme that is resilient to PFC pausing by exploring a combination of multi-level congestion signals. At its heart, Proteus leverages RTT-level signals (i.e., RTT and link utilization) to detect path status for initial routing decision, and exploits sub-RTT level signal (i.e., cumulative sojourn time) to reflect instantaneous PFC pausing and make timely rerouting choices based on the idea of better-late-than-never. We have implemented Proteus in the hardware programmable switch. Our testbed experiments as well as large-scale simulations show that Proteus can effectively handle PFC pausing under realistic workloads and achieve up to 35%, 31%, 28%, 22% and 46%, 42%, 34%, 29% better average FCT and$99^{th}$percentile FCT than CONGA, DRILL, Hermes and MP-RDMA, respectively.

Delay-based congestion control has drawn a lot of attention in both academics and industry recently. Specifically, the Copa algorithm proposed in NSDI can achieve consistent high performance under various network environments and has already been deployed on Facebook. In this paper, we theoretically analyze Copa and reveal its large queuing delay and poor fairness issue under certain conditions. The root cause is that Copa fails to achieve its expected behaviors, i.e., clear the bottleneck buffer occupancy periodically. Moreover, we also reveal that the pathological competitive mode of Copa fails to guarantee friendliness. To address these issues, we propose Copa+, which enhances Copa with a parameter adaptation mechanism and an optimized competitive mode. Designed based on our theoretical analysis, Copa+ can adaptively clear the bottleneck buffer occupancy and become friendly to Cubic in the competitive mode. As a result, Copa+ inherits the advantages of Copa but achieves lower queuing delay and better fairness under different environments, as confirmed by real-world experiments and simulations. Specifically, Copa+ has the highest average throughput over different Internet links among different cloud nodes, compared to Cubic, BBR, PCC Vivace, Remy, and Indigo. Meanwhile, Copa+ has an 8.1% increase in throughput and similar low queuing delay compared to Copa. Moreover, Copa+ achieves 14.6% lower queuing delay and 2.4% higher throughput compared to Sprout over emulated cellular links.

Deep Neural Network (DNN) is a critical component of a wide range of applications. However, with the rapid growth of the training dataset and model size, communication becomes the bottleneck, resulting in low utilization of computing resources. To accelerate communication, recent works propose to aggregate gradients from multiple workers in the programmable switch to reduce the volume of exchanged data. Unfortunately, since using synchronization transmission to aggregate data, current in-network aggregation designs suffer from the straggler problem, which often occurs in shared clusters due to resource contention. To address this issue, we propose a straggler-aware aggregation transport protocol (SA-ATP), which enables the leading worker to leverage the spare computing and storage resources to help the straggling worker. We implement SA-ATP atop clusters using P4-programmable switches. The evaluation results show that SA-ATP reduces the iteration time by up to 57% and accelerates training by up to$1.8\times $in real-world benchmark models.

To achieve low latency and high link utilization, ECN-based transport protocols (i.e., DCTCP) are widely deployed in data center networks (DCN). In multi-tenant environment, however, the newly introduced ECN-enabled TCP greatly impairs the performance of applications with out-dated and misconfigured TCP stacks. The reason is that the ECN-enabled switch fails to treat the mixed TCP traffic fairly, resulting in the distinguished performance gap between the ECN-enabled and ECN-disabled TCPs. This paper proposes DDT (Dual Dynamic Thresholds), an active queue management algorithm (AQM) to achieve the flow-level fairness for coexisting heterogeneous TCP traffic. DDT monitors the switch queue in real time, and dynamically tunes the distance between ECN-marking and packet-dropping thresholds to mitigate the aggressiveness difference between the ECN-enabled and ECN-disabled TCPs. The results of real implementations and large-scaled simulations show that DDT elegantly fills the aggressiveness gap of heterogeneous TCP traffic without disturbing their own control loops, while only introducing acceptable deployment overhead at switch.

Multicast is widely used to support a huge range of applications with one-to-many or many-to-many communication patterns. However, multicast systems do not scale due to considerable state and communication overheads. Some stateful multicast approaches require maintaining the state of each multicast session at switches, thus incurring large memory overhead. Some stateless ones utilize Bloom filter (BF) to encode multicast tree into the packet header to minimize communication overhead, but potentially suffer from the substantial false positive due to the probabilistic nature of Bloom filter. In this paper, we propose a stateless multicast scheme MEB, which uses Bloom filter to achieve large-scale multicast communication with low error, small overhead and high scalability. Specifically, to control the rate of false positive, MEB elaborately selects the hash functions for Bloom filters when constructing the packet header at the sender side, and makes forwarding decision according to packet header at the switch with negligible overhead. We compare MEB against the state-of-the-art multicast system in large-scale simulations. The test results show that MEB reduces the traffic overhead by up to 70% with small error rate.

Distributed Machine Learning (DML) techniques are widely used to accelerate the training of large-scale machine learning models. However, during training iterations, gradients need to be frequently aggregated across multiple workers, resulting in communication bottleneck. To reduce the communication overhead of DML, several In-Network Aggregation (INA) protocols are proposed to reduce the volume of aggregation traffic by offloading aggregation functions into switches, thus alleviating network bottlenecks. Nevertheless, these protocols couple the congestion control of in-switch aggregator resources and link bandwidth resources, together with the straggler-oblivious manner in aggregator allocation, leading to low aggregation efficiency.

To reduce the communication overhead in distributed sketch system, it is desirable to compress sketches before uploading. However, current sketch compression approaches hardly achieve high speed of compression procedure and low error of compressed sketches at the same time. In this paper, we take a clean slate approach to design a sketch compression scheme called as Fast-Mapping that achieves both fast compression speed and high accuracy. Based on the prior statistics knowledge of bucket data distribution, Fast-Mapping compresses the similar buckets to obtain high accuracy. We also theoretically derive the compression error bound of Fast-Mapping. The experimental results show that, Fast-Mapping achieves higher speed and lower error than the-state-of-art works.

Various datacenter network (DCN) load balancing schemes have been proposed in the past decade. Unfortunately, most of these solutions designed for lossy DCNs do not work well for Priority Flow Control (PFC) enabled lossless DCNs, primarily due to the reason that the individual congestion signals used in these solutions, e.g., link load, queue length, Round Trip Time (RTT) and Explicit Congestion Notification (ECN), may not be able to correctly or timely reflect the hop-by-hop PFC pausing. This paper first reveals the above problems via extensive experiments, and then based on the insights learned, we present Proteus, a PFC-aware load balancing scheme that is resilient to PFC pausing by exploring a combination of multi-level congestion signals. At its heart, Proteus leverages RTT-Ievel signals (i.e., RTT and link utilization) to detect path status for initial routing decision, and exploits sub-RTT level signal (i.e., cumulative sojourn time) to reflect instantaneous PFC pausing and make timely rerouting choices based on the idea of better-late-than-never. We have implemented Proteus in the hardware programmable switch. Our testbed experiments as well as large-scale simulations show that Proteus can effectively handle PFC pausing under realistic workloads and achieve up to 35 %, 31 %, 28%, 22% and 46 %, 42 %, 34 %, 29 % better average FCT and 99thpercentile FCT than CONGA, DRILL, Hermes and MP-RDMA, respectively.

Due to the ability to maintain good accuracy and high throughput with limited memory resources, sketch has gained wide deployment and application for approximate flow estimation. However, most existing sketch approaches ignore the distinctions between flow priorities, though the high-priority flows are relatively scarce but hold significant information. Therefore, a class of priority-aware sketches has appeared recently to provide differentiated measurement accuracy for flows with different priorities. Unfortunately, it is challenging for these priority-aware sketches to strike a good balance between accuracy and throughput. To address this issue, we propose a priority-adaptive architecture PA-Sketch, which utilizes priority-aware hash to dynamically allocate appropriate numbers of hash functions for different flows according to their priorities. For the scenarios we experimented, we observed that PA-Sketch significantly improves accuracy while minimizing the hash overhead. Compared to the state-of-the-art priority-aware sketches, PA-Sketch achieves around 4.83x higher accuracy and 1.83x higher F1 score for high-priority flows on average, meanwhile maintaining slight accuracy loss for low-priority flows.

Large-scale machine learning typically adopts distributed machine learning (DML) techniques to accelerate model training. Due to the large communication overhead, unfortu-nately, the phase of gradient aggregation has become the performance bottleneck for DML. To reduce traffic volume, several in-network aggregation (INA) transmission protocols are proposed to offload gradient aggregation function into the programmable switches. However, since existing INA transmission protocols use synchronous congestion control mechanism to drive each round of gradient aggregation, the straggling workers lead to long iteration time and significant performance degradation. To solve the above problem, we propose PA-ATP, a progress-aware INA transmission protocol, which adopts the progress-aware asynchronous congestion control. PA-ATP adjusts the sending rate in accordance with the transmission progress, allowing the straggling flow to grab more bandwidth than the leading flow and control the asynchronous degree of straggling job. We use a P4 programmable switch and a kernel-bypass protocol stack to implement PA-ATP. The results of testbed and large-scale NS3 simulations show that PA-ATP reduces training time by up to 62% compared to the state-of-the-art INA transmission protocols.

Many existing load balancing mechanisms work effectively in lossy datacenter networks (DCNs), but they suffer from serious packet reordering in lossless Ethernet DCNs deployed with the hop-by-hop Priority-based Flow Control (PFC). The key reason is that the prior solutions are not able to correctly and timely perceive PFC triggering when making load balancing decisions. Once the forwarding path pauses transmission due to PFC triggering, the packets allocated on it are blocked, inevitably leading to out-of-order packets and retransmission. In this paper, we present a Reordering-robust Load Balancing (RLB) scheme with PFC prediction in lossless DCNs. At its heart, RLB leverages the derivative of ingress queue length to predict PFC triggering and proactively notifies the upstream switches to choose an appropriate rerouting path or perform packet recirculation to avoid reordering. As a building block for existing load balancing mechanisms, we have integrated RLB into Presto, LetFlow, Hermes and DRILL. The test results show that the RLB-enhanced solutions deliver significant performance by avoiding packet reordering. For example, it reduces the 99th percentile flow completion time (FCT) by up to 58%, 67%, 72% and 54% over Presto, LetFlow, Hermes and DRILL, respectively.

Abstract Datacenter applications generate diverse flows, including deadline flows and non‐deadline flows. The deadline flows require to complete within strict deadline, while non‐deadline flows seek a shorter flow completion time. The state‐of‐the‐art deadline‐aware methods either transmit deadline flows with best‐effort at high priority, resulting in the starvation of non‐deadline flows, or blindly restrict the sending rates of deadline flows, leading to a high deadline missing ratio. To meet the different requirements of mix‐flows, a novel traffic‐aware rate control (TRC) method is proposed. TRC dynamically adjusts the sending rates of deadline flows according to their deadlines and the predicted future traffic patterns. If the intense competition is predicted among deadline flows, TRC will adopt a more aggressive manner to transmit the current deadline flows to avoid bandwidth contention in the future, reducing the deadline missing ratio. Otherwise, TRC will conservatively transmit deadline flows and complete these flows near their respective deadlines, relinquishing the excess bandwidth to non‐deadline flows. Meanwhile, TRC schedules non‐deadline flows in accordance with their sizes, minimizing the average FCT. The performance of TRC in large‐scale scenarios is evaluated through NS2 simulations. The test results show that TRC reduces the deadline missing ratio of deadline flows and the FCT of non‐deadline flows by up to 69.5% and 78.7% compared to the state‐of‐the‐art deadline‐aware schemes, respectively.

Since most flows are short-lived in data center networks, fast convergence becomes very important to help the short flows effectively utilize high bandwidth. Though current explicit feedback-based transport control protocols (TCPs) provide fast convergence via fine-grained congestion information from customized switches, they unavoidably incur large traffic overhead for widely existing small packets in data center applications, resulting in suboptimal network efficiency. To solve this issue, we propose a datacenter TCP based onDifferentialExplicitCongestionNotification, called DECN, to achieve fast convergence without any traffic overhead. Specifically, DECN feeds rate difference between the target and current rate back to the source by using multiple consecutive packets. Besides, we propose an enhanced version DECN* which obtains the optimal number of consecutive packets according to the packet loss rate. The experimental results of NS2 simulation and testbed implementation show that DECN and its enhanced version DECN* achieve comparable fast convergence as XCP without incurring any extra feedback overhead. Compared with the state-of-the-art explicit feedback-based TCPs, they reduce the flow completion time by up to 34% in typical data center applications.

In recent years, receiver-driven transport protocols have been proposed to use proactive congestion control to meet the stringent latency requirements of large-scale applications in data center. However, the receiver-driven proposals face the challenges brought by network dynamic. First, when the bursty flows start, the aggressive and blind line-rate transmission in the first RTT easily leads to persistent queue backlog. Second, when some flows finish transmissions, the remaining ones cannot increase their sending rates to seize the available bandwidth. To address these problems, this article presents a new receiver-driven congestion control design, called REN, which uses the under- and over-utilization notifications from switch to handle the dynamic traffic. With the aid of explicit feedback, REN alleviates the traffic burstiness due to aggressive start, mitigates the conservativeness in utilizing available bandwidth, and still retains the receiver-driven feature to achieve ultra-low latency. We implement the prototype of REN using DPDK. The experimental results of real testbed and large-scale NS2 simulation show that REN effectively reduces the average flow completion time (AFCT) by up to 68% over the state-of-the-art receiver-driven transmission schemes.

The three-dimensional deployment of Unmanned Aerial Vehicles (UAVs) has attracted extensive attention, especially for the Internet of Vehicles (IoV) in an emergency or to help the overloaded edge servers in traffic peaks. However, most existing works assume a two-dimensional road to simplify the design and modeling, while ignoring the interchange bridges scenario. In this scenario, UAVs deployment will face new challenges: the line-of-sight (LoS) transmission between the vehicles and UAVs is weakened due to the occlusion of the bridge body and vehicle movement. Meanwhile, energy consumption and the quantity of UAVs also need to be considered. In this paper, we propose an energy-aware 3D-deployment of UAVs, named 3D-UAV, to guarantee a high uplink rate with a minimized number of UAVs in IoV with Highway Interchange. First, considering the channel gain over bridges, 3D-UAV divides vehicles into several clusters. In each time slot, the number of clusters is iteratively optimized. Based on the clustering result, the flight altitude of the UAV is optimized in a stochastic gradient ascent (SGA) way aiming at maximizing the average uplink rate of transmission. Numerical results show that the proposed 3D-UAV can cover all vehicles on the highway interchange with the number of UAVs close to the theoretical lower bound. Meanwhile, it outperforms SOA, DRL, and HOLD methods in terms of the uplink rate and energy.

Most commercial players adopt adaptive bitrate (ABR) algorithms to dynamically decide each chunk's bitrate based on the perceived network bandwidth and buffer occupancy. However, current ABR algorithms are agnostic of audio bitrate selection since they deem it has negligible influence on video bitrate selection due to small size of audio chunks. Nevertheless, with the development of audio technologies, the bitrate of audio content increases dramatically in recent years. Thus, inappropriate audio selection can significantly affect video selection and deteriorate the viewing experience. To tackle these inefficiencies, we propose a deepReinforcement learning-based ABR algorithm that takesAudio andVideo quality into account (RAV) to circumvent a series of suboptimal performances, like low playback quality, frequent playback interruptions, poor playback smoothness, and undesirable combinations of video and audio chunks. Furthermore, RAV trains a neural network model that automatically outputs the bitrates for future audio and video chunks without relying on any presumptions about the environment, achieving good robustness to a broad spectrum of conditions. By conducting trace-driven and real-world experiments, we demonstrate that RAV significantly ameliorates the average overall viewing quality by 37.96%-118.20% over the state-of-the-art ABR algorithms. In addition, we also conduct subjective experiments by inviting 32 volunteers, and 27/32 users strongly agree that RAV provides them a better viewing experience than existing ABR solutions.

Video traffic has experienced an exponential increase in current years due to the growing ubiquity of mobile equipment and the constant network improvement. Most commercial players employ adaptive bitrate (ABR) algorithms to dynamically choose bitrate for each chunk based on perceived network capacity and buffer occupancy. Unluckily, even though improving the quality of chunks with dynamic scenes can achieve more QoE gain than static scenes, current ABR algorithms usually strive to maximize the average bitrate instead of perceptual quality, leading to the QoE degradation. To overcome this obstacle, we introduce a dynamic-chunk quality-aware adaptive bitrate algorithm through apprenticeship learning called DAVS (Dynamic-chunk qualityAwareVideoStreaming), where higher quality is selected for the dynamic chunks without reducing the quality of static chunks extravagantly. Furthermore, we take the user’s viewing preference into account to make DAVS adapt to the QoE diversity. The experimental results demonstrate that DAVS ameliorates the quality of dynamic chunks and significantly enhances the QoE compared with several representative ABR algorithms.

Existing reactive or proactive congestion control protocols are hard to simultaneously achieve ultra-low latency and high link utilization across all workloads ranging from delay-sensitive flows to bandwidth-hungry ones in datacenter networks. We present an Anti-ECN (Explicit Congestion Notification) Marking Receiver-driven Transport protocol called AMRT, which achieves both near-zero queueing delay and full link utilization by reasonably increasing sending rate in the case of under-utilization. Specifically, switches mark the ECN bit of data packets once detecting spare bandwidth. When receiving the anti-ECN marked packet, the receiver generates the corresponding marked grant to trigger more data packets. The testbed and simulation experiments show that AMRT effectively reduces the average flow completion time (AFCT) by up to 42% and improves the link utilization by up to 38% over the state-of-the-art receiver-driven transmission schemes.

Mobile Edge Computing (MEC) has attracted attention for its short-range and low-latency computing services for the Internet of Vehicles (IoV). However, in the Vehicle-RSU-MEC environment, incomplete task migration is a problem when Vehicle Users(VUs) are moving at high speeds around the Road Side Units (RSUs). Additionally, during peak traffic periods or road congestion, completed tasks compete for MEC server computing resources, leading to load imbalance. Therefore, making reasonable migration and offloading decisions is an important challenge. To address this challenge, the paper proposes a Cooperative Offloading strategy to jointly optimize offloading Decisions and Allocation of computing resources (CODA) step by step. First, to solve the problem of incomplete task migration, CODA proposed a Greedy-Based Task Completion Migration (GBTCM) algorithm. The algorithm calculates the required RSU set for each task to achieve complete migration, and greedily searches for the optimal migration target in the corresponding set to reduce task transmission latency. Second, after completing the task migration, CODA proposed a Distance-Based Computing Resource Allocation (DBCRA) algorithm to achieve load balancing for MEC servers. The algorithm prioritizes distance and finds MEC servers with sufficient computing resources to achieve better load balancing performance. Experimental results have shown that CODA is a low-complexity algorithm applicable to IoV, which can make reasonable and rapid decisions for task migration and offloading. Compared to three other benchmarks, CODA exhibits higher effectiveness and superiority.

Collaborative cloud-edge-end computing is a promising solution to support computation-intensive and latency-sensitive tasks by utilizing rich computing resources of cloud datacenters and low access delay of mobile edge computing (MEC) servers. Compared with traditional cloud computing and MEC, the cloud-edge environment has a stronger heterogeneity of servers and networks, resulting in significant differences between servers in the computation speed and access delay. However, few studies on cloud-edge-end task offloading focused on the characteristic of 5G heterogeneous networks in the cloud-edge environment. In this paper, we study the task offloading problem for collaborative cloud-edge-end computing in MEC-enabled small cell networks with low-cost distributed wireless backhaul. We aim to minimize the energy consumption of all user devices (UDs) via jointly optimizing the offloading decision, UDs’ transmission power, and the allocation of spectrum and computation resources. To solve the non-convex problem, we decouple the original problem into three subproblems, and design an efficient method with solving these three subproblems iteratively to obtain a high-quality solution. The simulation results indicate that our proposed method can lead to significant reduction in the energy consumption of all UDs compared with other conventional methods.

Identifying heavy flows is essential for network management. However, it is challenging to detect heavy flow quickly and accurately under the highly dynamic traffic and rapid growth of network capacity. Existing heavy flow detection schemes can make a trade-off in efficiency, accuracy and speed. However, these schemes still require memory large enough to obtain acceptable performance. To address this issue, we propose ChainSketch, which has the advantages of good memory efficiency, high accuracy and fast detection. Specifically, ChainSketch uses the selective replacement strategy to mitigate the over-estimation issue. Meanwhile, ChainSketch utilizes the hash chain and compact structure to improve memory efficiency. We implement the ChainSketch on OVS platform, P4-based testbed and large-scale simulations to process heavy hitter and heavy changer detection. The results of trace-driven tests show that, ChainSketch greatly improves the F1-score by up to$3.43\times $compared with the state-of-the-art solutions especially for small memory.

Datacenter networks provide large bisection bandwidth by load balancing traffic over rich parallel paths in multi-rooted tree topologies. Nevertheless, production datacenters operate under various path diversities caused by traffic dynamics, hardware failures and heterogeneous switching equipment. Therefore, the load balancing schemes in data center should be resilient to network asymmetry. Prior fine-grained schemes such as RPS and Presto are prone to experience packet reordering problem under asymmetric topology since they split flows into small units which are spread across all parallel paths. The coarse-grained solutions such as ECMP and LetFlow effectively avoid packet reordering, but easily leading to under-utilization of multiple paths. To solve these problems, we propose a load balancing mechanism called AG, which adaptively adjusts switching granularity according to the asymmetric degree of multiple paths. AG increases switching granularity to alleviate packet reordering under large degrees of topology asymmetry, while reducing switching granularity to obtain high link utilization under small degrees of topology asymmetry. Moreover, we design a switch-based scheme which measures the difference of one-way delay of multiple paths to obtain accurate state of topology asymmetry with low overhead. AG is a practical switch-based solution without modification at end hosts. The experimental results of NS2 simulations and real implementation show that AG reduces the average and$99^{th}$flow completion time by up to 54% and 65% compared with the state-of-the-art load balancing schemes, respectively.

Nowadays, the distributed key-value stores have become the basic building block for large-scale cloud applications. In large-scale distributed key-value stores, many key-value access operations, which will be processed in parallel on different servers, are usually generated for a single end-user request. Accordingly, the completion time of an end-user request is determined by the last completed key-value access operation. Scheduling the order of serving key-value access operations can effectively reduce the completion times of end requests, thereby improving the user experience. However, existing scheduling algorithms hardly achieve consistent low latency due to the following challenges: the large overhead of cooperating clients and servers, the time-varying load and performance of servers, the traffic distribution can be either heavy-tailed or light-tailed and both the mean and the tail completion time are expected to be low. In this paper, we formalize the problem of scheduling key-value access operations and show it is NP-hard. Furthermore, we heuristically design the distributed adaptive scheduler (DAS), which distributively combines the largest remaining processing time last and the shortest remaining process time first algorithms. Theoretical analysis shows that DAS is adaptive to the time-varying traffic and server performance and can achieve consistent low mean and tail latency regardless of traffic distributions. Extensive simulations show that DAS reduces the mean request completion time by$17 \! \sim \! 50\%$with heavy-tailed traffic and$2 \! \sim 26 \! \%$with light-tailed traffic, while keeping the smallest tail completion time, compared to the default first come first served algorithm. Moreover, DAS outperforms the existing Rein-SBF algorithm under various scenarios.

In distributed key-value stores, multiple replica servers are always available for each key-value access operation when the eventual consistency model is employed. Accordingly, the completion times of the key-value access operations generated by an end-user request at different servers may be of great difference, especially when the replica servers are heterogeneous and have time-varying performance. Accordingly, the replica selection algorithm is crucial to cut the response time of end-user requests. The main challenge of making replica selection for each light-weighted key-value access operation is to timely know the status of replica serves. Recently, the adaptive replica selection algorithm C3 suggests guiding the replica selection with the piggybacked information of replica server in the returned “value”. Although C3 has good performance, the poor timeliness of feedback information makes a large performance gap between C3 and the ideal replica selection algorithm. To narrow this gap, the Accelerated Information Dissemination (AID) mechanism is proposed in this paper. Specifically, AID removes the bottleneck of information dissemination at the “slow” servers by letting both “client” and “replica server” store the records about the status of replica servers and both “key” and “value” piggyback multiple records. AID is implemented in Cassandra and evaluated by experiments and large scale simulations. The results show AID can significantly improve the timeliness of feedback information, especially when the number of nodes is large. Accordingly, AID helps C3 to greatly reduce the latency.

In Priority Flow Control (PFC) enabled datacenter networks (DCNs), PFC is inevitably triggered due to bursty traffic even with end-to-end congestion control. Load balancing as a complementary mechanism to transport protocols can make rerouting decisions in time to alleviate PFC’s head-of-line (HoL) blocking problem. However, prior solutions designed for lossy DCNs do not work well in PFC-enabled networks, because the unreliable rerouting signals such as separate local queue length, round-trip time (RTT), explicit congestion notification (ECN), and link load cannot timely and correctly reflect PFC pausing.

Adaptive bitrate (ABR) algorithms are routinely adopted for transmitting media contents across dynamic networks. State-of-the-art ABR algorithms only adapt to video bitrate without considering audio bitrate adaption as they consider the im-pact on the video to be negligible due to the small size of the audio. However, to bring users an immersive experience, more and more content providers have applied high-quality audio with large sizes, like stereophonic and surround (Dolby Atmos). Therefore, improper audio bitrate selection will ad-versely affect video bitrate selection, leading to undesirable audio/video combinations (the highest video quality with the lowest audio quality, vice versa) and frequent playback inter-ruptions. To address these inefficiencies, we propose a Self-Play reinforcement learning-based Audio-aware ABR algorithm named SPA to learn strategies for audio and video bi-trate selections. Experimental results demonstrate SPA's con-siderable superiority as compared with existing approaches.

In the parameter-server-based distributed deep learning system, the workers simultaneously communicate with the parameter server to refine model parameters, easily resulting in severe network contention. To solve this problem, Asynchronous Parallel (ASP) strategy enables each worker to update the parameter independently without synchronization. However, due to the inconsistency of parameters among workers, ASP experiences accuracy loss and slow convergence. In this paper, we propose Hybrid Synchronous Parallelism (HSP), which mitigates the communication contention without excessive degradation of convergence speed. Specifically, the parameter server sequentially pulls gradients from workers to eliminate network congestion and synchronizes all up-to-date parameters after each iteration. Meanwhile, HSP cautiously lets idle workers to compute with out-of-date weights to maximize the utilizations of computing resources. We provide theoretical analysis of convergence efficiency and implement HSP on popular deep learning (DL) framework. The test results show that HSP improves the convergence speedup of three classical deep learning models by up to 67%.

Heavy flow detection in enormous network traffic is a critical task for network measurement. Due to the limited memory size and high link capacity, accurate detection of heavy flows becomes challenging in large-scale networks. Almost all existing approaches of detecting heavy flows use single-dimension statistics of flow size to make flow-replacement decisions. However, under the mass number of small flows, the heavy flows are prone to be frequently and mistakenly replaced, resulting in unsatisfactory accuracy. To solve this problem, we reveal that the number of uninterrupted arrival packets is a useful metric in identifying flow types. We further propose UA-Sketch that expels small flows and protects heavy ones according to the multiple-dimension statistics including both estimated flow size and number of uninterrupted arrival packets. The test results of trace-driven simulations and OVS experiments show that, even under small memory, UA-Sketch achieves higher accuracy than the existing works, with the F1 Score by up to 2.1 ×.

Copa is a delay-based congestion control algorithm proposed in NSDI recently. It can achieve consistent high performance under various network environments and has already been deployed in Facebook. In this paper, we theoretically analyze Copa and reveal its large queuing delay and poor fairness issue under certain conditions. The root cause is that Copa fails to clear the bottleneck buffer occupancy periodically as expected. Accordingly, Copa may get a wrong base RTT estimation and enter its competitive mode by mistake, leading to large delay and unfairness. To address these issues, we propose Copa+, which enhances Copa with a parameter adaptation mechanism and an optimized competitive mode entrance criterion. Designed based on our theoretical analysis, Copa+ can adaptively clear the bottleneck buffer occupancy for correct estimation of base RTT. Consequently, Copa+ inherits the advantages of Copa but achieves lower queuing delay and better fairness under different environments, as confirmed by the real-world experiments and simulations. Specifically, Copa+ has the highest throughput similar to Copa but 11.9% lower queuing delay over different Internet links among different cloud nodes, and achieves 39.4% lower queuing delay and 8.9% higher throughput compared to Sprout over emulated cellular links.

Modern data centers host diverse applications, which generate a mix of short flows with stringent latency requirement and long flows requiring large sustained throughput. To solve the problem of resource competition between the mixed flows, we propose an adaptive traffic isolation scheme APS. Based on the packet spraying scheme in the multipath transmission, APS dynamically separates long flows from short ones on different paths to provide the low latency for the short flows. Meanwhile, to resolve the out-of-order problem, APS limits the long flows to a few paths with Equal Cost Multi Path (ECMP). Experimental results of NS2 simulation and testbed implementation show that, APS reduces the average completion time for short flows by up to 60 percent and increases the throughputs for long flows by about 1.68x over the state-of-the-art multipath transmission schemes.

The video streaming system employs adaptive bitrate (ABR) algorithms to optimize a user’s quality of experience. However, it is hard for ABR algorithms to choose the right bitrate consistently under highly dynamic bandwidth fluctuations in wild Internet. In this article, we propose a building block on the client side named Opportunistic Chunk Replacement Mechanism (OCRM) to help existing ABR algorithms make full use of the available bandwidth to improve the network utilization and viewing experience of users. Specifically, the servers take advantages of the spare bandwidth to opportunistically transmit high-quality chunks (called opportunistic chunks ) with low priority to the client, without incurring any extra delay. Then, the client player replaces the low-quality chunks with the opportunistic ones that have high quality. We compare OCRM with state-of-the-art ABR algorithms by using trace-driven experiments spanning a wide variety of quality of experience metrics and network conditions. The test results show that OCRM effectively achieves high network utilization and improves the user’s viewing experience by up to 35%.

Modern data centers often host multiple applications with diverse network demands. To provide fair bandwidth allocation to several thousand traversing flows, Approximate Fair Queueing (AFQ) utilizes multiple priority queues in switch to approximate ideal fair queueing. However, due to limited number of queues in programmable switches, AFQ easily experiences high packet loss and low link utilization. In this paper, we propose Elastic Fair Queueing (EFQ), which leverages limited priority queues to flexibly achieve both high network utilization and fair bandwidth allocation. EFQ dynamically assigns the free buffer space in priority queues for each packet to obtain high utilization without sacrificing flow-level fairness. The results of simulation experiments and real implementations show that EFQ reduces the average flow completion time by up to 82% over the state-of-the-art fair bandwidth allocation mechanisms.

Explicit Congestion Notification (ECN) is widely utilized in modern data centers to achieve low latency and high throughput for various applications. In recent years, however, even with the sustainable growth of link bandwidth in data centers, the switch buffer size does not increase remarkably. Consequently, the standard per-port ECN scheme suffers from excessive packet loss. Though the shared-buffer ECN scheme alleviates the packet loss, we observe that it leads to severe unfairness, which we term as the Port Starvation problem. When flows destined for some ports have aggressively occupied the shared buffer, the later-arrival flows destined for other ports will be ECN-marked unfairly and obtain significantly lower throughput. To address the port starvation problem, we design a buffer-aware fair ECN-marking (BFEM) scheme for shallow-buffered switch. BFEM leverages the shared buffer to reduce packet loss and meanwhile punishes aggressive flows by ECN marking. We evaluate BFEM with both 40Gbps P4 testbed implementation and large-scale NS2 simulation. The test results show that, by improving fairness between egress ports, BFEM increases total link utilization and reduces the average flow completion time by up to 40% compared with the state-of-the-art per-port and shared-buffer ECN marking schemes.

To improve the quality of experience for worldwide users, an increasing number of service providers deploy their services on geographically dispersed data centers, which are connected by wide area network (WAN). In the cross-datacenter networks, however, the intra- and inter-datacenter parts have different characteristics, including switch buffer depth, round-trip time and bandwidth. Besides, most of intra-DC flows belong to interactive services that require low delay while inter-DC flows typically need to achieve high throughput. Unfortunately, existing sender-based and receiver-driven transport protocols do not consider the network heterogeneity between inter- and intra- DC networks so that they fail to simultaneously achieve low latency for intra-DC flows and high throughput for inter-DC flows. This paper proposes a general hybrid congestion control mechanism called GTCP to address this problem. When the inter-DC flow detects congestion inside data center, it switches to the receiver-driven mode to avoid the impact on intra-DC flows. Otherwise, it switches back to the sender-based mode to proactively explore the available bandwidth. Besides, the intra-DC flow leverages the pausing mechanism to eliminate the queue build-up. Through a series of testbed experiments and large-scale NS2 simulations, we demonstrate that GTCP reduces flow completion time by up to 79.3% compared with existing protocols.

Modern datacenter applications bring fundamental challenges to transport protocols as they simultaneously require low latency and high throughput. Recent receiver-driven trans-port protocols transmit only one data packet once receiving each grant or credit packet from the receiver to achieve ultra-low queueing delay and zero packet loss. However, the round-trip time variation and the highly dynamic background traffic significantly deteriorate the performance of receiver-driven transport protocols, resulting in under-utilized bandwidth. This paper designs a simple yet effective solution called RPO that retains the advantages of receiver-driven transmission while efficiently utilizing the available bandwidth. Specifically, RPO rationally uses low-priority opportunistic packets to ensure high network utilization without increasing the queueing delay of high-priority normal packets. In addition, since RPO only uses Explicit Congestion Notification (ECN) marking function and priority queues, RPO is ready to deploy on switches. We implement RPO in Linux hosts with DPDK. Our small-scale testbed experiments and large-scale simulations show that RPO significantly improves the network utilization by up to 35% under high workload over the state-of-the-art receiver-driven transmission schemes, without introducing additional queueing delay.

For massive access to the Internet of Things, edge computing servers are installed on cellular ground base stations (GBSs) with fixed geographical locations, which easily suffer from traffic overload of the end user (EU) with high density and mobility. To provide reliable and flexible offloading service, unmanned aerial vehicles (UAV) are explored to assist edge computing, which relieves the computation offloading pressure of both EUs and GBS. However, most existing UAV researches focus on trajectory design to reduce offloading delay, which ignoring the variability of user distribution and the energy limitation of UAV. This article proposes a novel UAV-assisted edge computing framework, named as HOTSPOT, which locates the UAV in 3-D space according to the time-varying hot spot of user distribution and provides the corresponding edge computing offloading assistance. By formulating the UAV positioning problem into a maximum clique problem, a light-weighted deterministic algorithm is proposed based on stochastic gradient descent to search the optimal location of UAV. With the elaborate UAV position, HOTSPOT further gives an opportunistic offloading balanced scheme to reach low latency. Simulation results show that when the GBS load is 75%, HOTSPOT reduces the average offloading delay by 33%. When the GBS load reaches 90%, the average delay reduction is up to 80%.

Mobile-edge computing (MEC) is expected to provide reliable and low-latency computation offloading for massive Internet of Things (IoT) with the next generation networks, such as the sixth-generation (6G) network. However, the successful implementation of 6G depends on network densification, which brings new offloading challenges for edge computing, one of which is how to make offloading decisions facing densified servers considering both channel interference and queuing, which is an NP-hard problem. This article proposes a distributed-two-stage offloading (DTSO) strategy to give tradeoff solutions. In the first stage, by introducing the queuing theory and considering channel interference, a combinatorial optimization problem is formulated to calculate the offloading probability of each station. In the second stage, the original problem is converted to a nonlinear optimization problem, which is solved by a designed sequential quadratic programming (SQP) algorithm. To make an adjustable tradeoff between the latency and energy requirement among heterogeneous applications, an elasticity parameter is specially designed in DTSO. Simulation results show that compared to the latest works, DTSO can effectively reduce latency and energy consumption and achieve a balance between them based on application preferences.

Data center networks typically adopt multi-rooted tree topologies to provide high bisection bandwidth. Various fine-grained load balancing schemes have been proposed to split flows across multiple paths. However, data center networks suffer from many uncertainties such as highly dynamic traffic. These uncertainties easily make network become asymmetric, resulting in significant packet reordering. Unfortunately, existing solutions passively deal with packet reordering based on a threshold and hardly adapt to asymmetric networks because of lacking the explicit reordering feedback. These solutions either fail to quickly respond to packet loss or cause unnecessary fast retransmission, which reduces link utilization and increases flow completion time. In this paper, we propose a fine-grained load balancing scheme RMC to eliminate the impact of packet reordering and handle uncertainties in asymmetric networks. To avoid unnecessary fast retransmission, the switch proactively identifies reordered packet according to local queue length and global path latency. Furthermore, we employ a coding technique with redundancy optimization to reduce long-tailed flow completion time under network asymmetry. Through a series of large-scale NS2 simulations and testbed experiments, we demonstrate that RMC effectively avoids unnecessary fast retransmission under different network scenarios and reduces flow completion time by up to 72% compared with state-of-the-art schemes.

The state-of-the-art datacenter load balancing designs commonly optimize bisection bandwidth with homogeneous switching granularity. Their performances surprisingly degrade under mixed traffic containing both short and long flows. Specifically, the short flows suffer from long-tailed delay, while the throughputs of long flows also degrade dramatically due to low link utilization and packet reordering. To solve these problems, we design a traffic-aware load balancing (TLB) scheme to adaptively adjust the switching granularity of long flows according to the load strength of short ones. Under the heavy load of short flows, the long flows use large switching granularity to help short ones obtain more opportunities in choosing short queues to complete quickly. On the contrary, the long flows reroute flexibly with small switching granularity to achieve high throughput. Furthermore, under extremely bursty scenario, we utilize the packet slicing scheme for long flows to release bandwidth for short ones. The experimental results of NS2 simulation and testbed implementation show that TLB significantly reduces the average flow completion time of short flows by 16%-67% over the state-of-the-art load balancers and achieves the high throughput for long flows. Moreover, for extreme bursty case, at the acceptable throughput degradation of long flows, TLB with packet slicing reduces the deadline missing ratio of bursty short flows by up to 80%.

Modern data center networks are usually constructed in multi-rooted tree topologies, which require the highly efficient multi-path load balancing to achieve high link utilization. Recent packet-level load balancer obtains high throughput by spraying packets to all paths, but it easily leads to the packet reordering under network asymmetry. The flow-level or flowlet-level load balancer avoids the packet reordering, while reducing the link utilization due to their inflexibility. To solve these problems, we design a Queueing Delay Aware Packet Spraying (QDAPS), that effectively mitigates the packet reordering for packet-level load balancer. QDAPS selects paths for packets according to the queueing delay of output buffer, and lets the packet arriving earlier be forwarded before the later packets to avoid packet reordering. Moreover, we adopt the “power-of- n-choices” paradigm on QDAPS to alleviate the impact of herd behavior under multiple forwarding engines. We compare QDAPS with ECMP, LetFlow and RPS through NS2 simulation and Mininet implementation. The test results show that QDAPS reduces flow completion time (FCT) by ~30%-50% over the state-of-the-art load balancing mechanism.

In data center networks, many network-intensive applications leverage large fan-in and many-to-one communication to achieve high performance. However, the special traffic patterns, such as micro-burst and high concurrency, easily cause TCP Incast problem and seriously degrade the application performance. To address the TCP Incast problem, we first reveal theoretically and empirically that alleviating packet burstiness is much more effective in reducing the Incast probability than controlling the congestion window. Inspired by the findings and insights from our experimental observations, we further propose a general supporting scheme Adaptive Pacing (AP), which dynamically adjusts burstiness according to the flow concurrency without any change on switch. Additionally, a sender-based approach is proposed to estimate the flow concurrency. Another feature of AP is its broad applicability. We integrate AP transparently into different TCP protocols (i.e., DCTCP, L2DCT and D2TCP). Through a series of large-scale NS2 simulations and testbed experiments, we show that AP significantly reduces the Incast probability across different TCP protocols and the network goodput can be increased consistently by on average 7× under severe congestion.

To deliver video in a high quality across various network conditions, adaptive bitrate (ABR) algorithms dynamically select bitrate for each chunk according to perceived network rate and buffer occupancy. Unfortunately, though ameliorating the quality of chunks with dynamic scenes can obtain more QoE gain than the ones with static scenes, current ABR algorithms generally aim to maximize the average bitrate rather than perceptual quality, resulting in the QoE degradation. To address this issue, we propose a dynamic-chunk quality aware adaptive bitrate scheme via apprenticeship learning named DAVS, in which higher quality is chosen for the dynamic chunks without decreasing the quality of static chunks excessively. The experimental results show that DAVS enhances the quality of dynamic chunks and greatly improves the overall QoE compared with the state-of-the-art ABR algorithms.

Since most flows are short-lived in data center networks, fast convergence becomes very important to help the short flows effectively utilize high bandwidth. Though current feedback-based transport control protocols (TCPs) provide fast convergence via fine-grained explicit congestion information from customized switches, they unavoidably incur large traffic overhead for widely existing small packets in data center applications, resulting in suboptimal network efficiency. To solve this issue, we propose a datacenter TCP based on differential feedbacks, called DECN, to achieve fast convergence without any traffic overhead. Specifically, DECN feeds rate difference between the target and current rate back to the source by using multiple consecutive packets. The experimental results of NS2 simulation and testbed implementation show that DECN achieves comparable fast convergence as XCP without incurring any extra feedback overhead. Compared with the state-of-the-art feedback-based TCPs, DECN reduces the flow completion time by up to 34.1% in typical data center applications.

To deliver high quality video across different network conditions, the video chunks are explicitly fetched by client or proactively pushed by server in Dynamic Adaptive Streaming over HTTP (DASH) system. Unfortunately, on the one hand, the client fetch mechanism suffers from bandwidth wastage due to its stop-and-wait fashion when the network delay becomes large. On the other hand, the server push mechanism performs poorly because of its inflexibility in bitrate switching under fluctuating bandwidth. To address these inefficiencies, we propose a pipeline-based chunk scheduling scheme called PCS to auto-turn the sending time of each chunk. For a given ABR algorithm, PCS dynamically pre-schedules the chunk delivery according to the real-time network conditions. Using the pipelined-based chunk delivery, PCS flexibly adjusts the bitrate of each chunk and meanwhile avoids the unnecessary waiting time in the stop-and-wait transmission. The experimental results of testbed implementations show that PCS greatly improves the average bitrate of the state-of-the-art ABR algorithms by up to 26%, and reduces the rebuffer rate by up to 31%.

Modern data centers often host multiple applications with diverse network demands. To provide fair bandwidth allocation to several thousand traversing flows, Approximate Fair Queueing (AFQ) utilizes multiple priority queues in switch to approximate ideal fair queueing. However, due to limited number of queues in commodity switches, AFQ easily experiences high packet loss and low link utilization. In this paper, we propose Elastic Fair Queueing (EFQ), which leverages limited priority queues to flexibly achieve both high network utilization and fair bandwidth allocation. EFQ dynamically assigns the free buffer space in priority queues for each packet to obtain high utilization without sacrificing flow-level fairness. The results of simulation experiments and real implementations show that EFQ reduces the average flow completion time by up to 82% over the state-of-the-art fair bandwidth allocation mechanisms.

Cloud applications generate a variety of workloads ranging from delay-sensitive flows to bandwidth-hungry ones in data centers. Existing reactive or proactive congestion control protocols are hard to simultaneously achieve ultra-low latency and high link utilization across all workloads in data center networks. We present a new receiver-driven transport scheme using anti-ECN (Explicit Congestion Notification) marking to achieve both near-zero queueing delay and full link utilization by reasonably increasing sending rate in the case of under-utilization. Specifically, switches mark the ECN bit of data packets once detecting spare bandwidth. When receiving the anti-ECN marked packet, the receiver generates the corresponding marked grant to trigger more data packets. The experimental results of small-scale testbed implementation and large-scale NS2 simulation show that AMRT effectively reduces the average flow completion time (AFCT) by up to 40.8% and improves the link utilization by up to 36.8% under high workload over the state-of-the-art receiver-driven transmission schemes.

In data center networks, a large number of concurrent TCP connections suffer the TCP Incast throughput collapse due to packet drops in shallow-buffered Ethernet switches. In this work, we first reveal theoretically and empirically that controlling the IP packet size is much more effective in avoiding Incast than cutting congestion window under severe congestion. We further design a general supporting scheme Packet Slicing, which adjusts the IP packet on widely used commodity switches. The design uses standard ICMP signaling, which makes no modification on TCP protocols and can be transparently utilized by various TCP protocols. To alleviate the impact of micro-burst caused by high flow concurrency, we utilize the TCP Pacing scheme to disperse packets over the round trip time, helping Packet Slicing to support more concurrent TCP flows. We integrate Packet Slicing with three state-of-the-art data center TCP protocols on NS2 simulation and a physical testbed. The experimental results show that Packet Slicing broadly improves the goodput of different data center TCP protocols by average 26x, while having almost no effect on the I/O performance of switches and end hosts.

The sustainable growth of bandwidth has been an inevitable tendency in current Data Center Networks (DCN). However, the dramatic expansion of link capacity offers a remarkable challenge to the transport layer protocols of DCN, i.e., how to converge fast and enable data flow to utilize the high bandwidth effectively. Meanwhile, the new protocol should be compatible to the traditional TCP because the applications with old TCP versions are still widely deployed. Therefore, it is important to achieve a trade-off between the aggressiveness and TCP-friendliness in protocol design. In this article, we first empirically investigate why the existing typical data center TCP variants naturally fail to guarantee both fast convergence and TCP friendliness. Then, we design a new transport protocol for DCN, namely Fast and Friendly Converging (FFC), which makes independent decisions and self-adjustment through retrieving the two-dimensional congestion notification from both RTT and ECN. We further present a mathematic model to analyze its competing behavior and converging process. The results from simulation experiments and real implementation show that FFC can achieve fast convergence, thus benefiting the flow completion time. Moreover, when coexisting with the traditional TCP, FFC also presents a moderate behavior, while introducing trivial deployment overhead only at the end-hosts.

To achieve better network performance, the cloud service providers are widely deploying the ECN-based transport protocols (i.e., DCTCP) in their data center networks (DCN). In multi-tenant environment, however, the newly introduced ECN-enabled TCP greatly impairs the performance of applications with out-dated and miscon figured TCP stacks. The reason is that the ECN-enabled datacenter switch fails to treat the mixed TCP traffic fairly, causing the distinguished performance gap between the ECN-enabled and ECN-disabled TCPs. This paper proposes DDT (Dual Dynamic Thresholds), an active queue management algorithm (AQM) that aims to achieve the flow-level fairness when the heterogeneous TCP traffic coexists. DDT monitors the switch queue in real time, and dynamically tunes the distance between ECN-marking and packet-dropping thresholds to mitigate the competitiveness difference between the ECN-enabled and ECN-disabled TCP. Our preliminary real implementations and testing results show that DDT elegantly fills the competitiveness gap of heterogeneous TCP traffic without disturbing their own control loops, while only introducing acceptable deployment overhead at the switch.

In the data center network, a packet-level load balancer such as random packet spraying (RPS) achieves high throughput by spraying data packets to all transmission paths, which easily suffers from the packet out-of-order problem under network asymmetry. While state-of-the-art network coding schemes can mitigate the issue, too many encoded redundant packets introduced by the network coding will cause extra traffic overhead, larger queueing delay and even TCP time out. In this paper, we propose OPportunistic Encoded Redundant (OPER), a middle-layer design upon existing coding schemes to mitigate the curse of redundant packets. Specifically, OPER uses opportunistic redundant packets which are replaceable by the data packets in the switches under heavy congestion. OPER is implemented as a shim layer between TCP and IP layers at end-hosts and a loadable plugin at switches, leaving existing TCP/IP protocols unmodified. The testbed and NS2 experiments show that, OPER reduces the average flow completion time by up to 71% compared with the state-of-the-art multipath coding schemes.

Modern data center networks provide multiple paths between host pairs to guarantee high aggregated network bandwidth and transmission reliability. However, data center networks suffer from various uncertainties such as highly dynamic traffic and heterogeneous devices. These uncertainties easily lead to network asymmetry and cause significant packet reordering. Unfortunately, due to lacking the explicit reordering feedback, existing sender-based and receiver-based solutions hardly adapt to asymmetric data center networks and cause long-tailed flow completion time as well as throughput loss. In this paper, we propose a per-packet transmission scheme RMC to eliminate the impact of packet reordering and handle uncertainties in asymmetric networks. To avoid unnecessary fast retransmission, the switch proactively identifies packet reordering according to local queue length and global path latency. Furthermore, we employ a coding technique to reduce long-tailed flow completion time under network asymmetry. Through a series of large-scale NS2 simulations and testbed experiments, we demonstrate that RMC reduces flow completion time by up to 72% compared with existing protocols.

Modern data center topologies often take the form of a multi-rooted tree with rich parallel paths to provide high bandwidth. However, various path diversities caused by traffic dynamics, link failures and heterogeneous switching equipments widely exist in production datacenter network. Therefore, the multi-path load balancer in data center should be robust to these diversities. Although prior fine-grained schemes such as RPS and Presto make full use of available paths, they are prone to experience packet reordering problem under asymmetric topology. The coarse-grained solutions such as ECMP and LetFlow effectively avoid packet reordering, but easily lead to under-utilization of multiple paths. To cope with these inefficiencies, we propose a load balancing mechanism called AG, which adaptively adjusts switching granularity according to the asymmetric degree of multiple paths. AG increases switching granularity to alleviate packet reordering under large degrees of topology asymmetry, while reducing switching granularity to obtain high link utilization under small degrees of topology asymmetry. AG is deployed on the switches with negligible overhead, while making no modification on end-hosts. We evaluate AG through both Mininet testbed and large-scale NS2 simulations. The experimental results show that AG reduces the average and 99thflow completion time by up to 51% and 56% over the state-of-the-art load balancing schemes, respectively.

Modern datacenter topologies typically are multi-rooted trees consisting of multiple paths between any given pair of hosts. Recent load balancing designs focus on making full use of available parallel paths to provide high bisection bandwidth. However, they are agnostic to the mixed traffic generated by diverse applications in data centers and respectively use the same granularity in rerouting flows regardless of the flow type. Therefore, the short flows suffer the long-tailed queueing delay and reordering problems, while the throughputs of long flows are also degraded dramatically due to low link utilization and packet reordering under the non-adaptive granularity. To solve these problems, we design a traffic-aware load balancing (TLB) scheme to adopt different rerouting granularities for two kinds of flows. Specifically, TLB adaptively adjusts the switching granularity of long flows according to the load strength of short ones. Under the heavy load of short flows, the long flows use large switching granularity to help short ones obtain more opportunities in choosing short queues to complete quickly. When the load strength of short flows is low, the long flows switch paths more flexibly with small switching granularity to achieve high throughput. TLB is deployed at the switch, without any modifications on the end-hosts. The experimental results of NS2 simulations and Mininet implementation show that TLB significantly reduces the average flow completion time (AFCT) of short flows by ~15%-40% over the state-of-the-art load balancing schemes and achieves the high throughput for long flows.

Modern data-center applications generate a diverse mix of short and long flows with different performance requirements and weaknesses. The short flows are typically delay-sensitive but to suffer the head-of-line blocking and out-of-order problems. Recent solutions prioritize the short flows to meet their latency requirements, while damaging the throughput-sensitive long flows. To solve these problems, we design a Coding-based Adaptive Packet Spraying (CAPS) that effectively mitigates the negative impact of short and long flows on each other. To exploit the availability of multiple paths and avoid the head-of-line blocking, CAPS spreads the packets of short flows to all paths, while the long flows are limited to a few paths with Equal Cost Multi Path (ECMP). Meanwhile, to resolve the out-of-order problem with low overhead, CAPS encodes the short flows using forward error correction (FEC) technology and adjusts the coding redundancy according to the blocking probability. Moreover, since the coding efficiency decreases when the coding unit is too small or large, we demonstrate how to obtain the optimal size of coding unit. The coding layer is deployed between the TCP and IP layers, without any modifications on the existing TCP/IP protocols. The test results of NS2 simulation and small-scale testbed experiments show that CAPS significantly reduces the average flow completion time of short flows by ~30%-70% over the state-of-the-art multipath transmission schemes and achieves the high throughput for long flows with negligible traffic overhead.

In modern data centers, many flow-based and task-based schemes have been proposed to speed up the data transmission in order to provide fast, reliable services for millions of users. However, the existing flow-based schemes treat all flows in isolation, contributing less to or even hurting user experience due to the stalled flows. Other prevalent task-based approaches, such as centralized and decentralized scheduling, are sophisticated or unable to share task information. In this work, we first reveal that the relinquishing bandwidth of leading flows to the stalled ones effectively reduces the task completion time. We further present the design and implementation of a general supporting scheme that shares the flow-tardiness information through a receiver-driven coordination. Our scheme can be flexible and widely integrated with the state-of-the-art TCP protocols designed for data centers in either single stage or multiple stage scenario, while making no modification on switches. Through the testbed experiments and simulations of typical data center applications, we show that in single stage scenario, our scheme reduces the task completion time by 70% and 50% compared with the flow-based protocols (e.g., DCTCP, L2DCT) and task-based scheduling (e.g., Baraat), respectively. Moreover, our scheme also outperforms other approaches by 18%~25% in prevalent topologies of the data center. For multiple stage scenario, our scheme also has up to 50% improvement compared to other schemes.

The dramatic expansion of link capacity in current data center network causes remarkable challenges to the design of new transport layer protocol, that is, how to converge as fast as possible to help data flow effectively utilize the high bandwidth. Meanwhile, the new protocol should be friendly to the traditional TCP because the non-cooperating applications with old TCP versions are widely existing. Therefore, it is important to achieve a trade-off between the aggressiveness and TCP-friendliness in protocol design. In this paper, we first empirically study why the existing typical data center TCP variants naturally fail to guarantee both fast convergence and TCP friendliness. Then, we design FFC, a transport protocol that makes independent decisions and self-adjustment through retrieving the two-dimensional congestion notification from the RTT and ECN. The results of simulation experiments and real implementations show that the fast convergence of FFC leads to the lower flow completion time compared with DX and DCTCP. Meanwhile, when coexisting with the traditional TCP, FFC also presents a moderate competitiveness, while introducing trivial deployment overhead only at the end hosts.

Modern data center networks are usually constructed in multi-rooted tree topologies, which require the highly efficient multi-path load balancing to achieve high link utilization. Recent packet-level load balancer obtains high throughput by spraying packets to all paths, but it easily leads to the packet reordering under network asymmetry. The flow-level or flowlet-level load balancer avoids the packet reordering, while reducing the link utilization due to their inflexibility. To solve these problems, we design a Queueing Delay Aware Packet Spraying (QDAPS), that effectively mitigates the packet reordering for packet-level load balancer. QDAPS selects paths for packets according to the queueing delay of output buffer, and lets the packet arriving earlier be forwarded before the later packets to avoid packet reordering. We compare QDAPS with ECMP, LetFlow and RPS through NS2 simulation and Mininet implementation. The test results show that QDAPS reduces flow completion time (FCT) by ~30%-50% over the state-of-the-art load balancing mechanism.

Modern data-center applications generate a diverse mix of short and long flows with different performance requirements and weaknesses. The short flows are typically delay-sensitive but to suffer the head-of-line blocking and out-of-order problems. Recent solutions prioritize the short flows to meet their latency requirements, while damaging the throughput-sensitive long flows. To solve these problems, we design a Coding-based Adaptive Packet Spraying (CAPS) that effectively mitigates the negative impact of short and long flows on each other. To exploit the availability of multiple paths and avoid the head-of-line blocking, CAPS spreads the packets of short flows to all paths, while the long flows are limited to a few paths with Equal Cost Multi Path (ECMP). Meanwhile, to resolve the out-of-order problem with low overhead, CAPS encodes the short flows using forward error correction (FEC) technology and adjusts the coding redundancy according to the blocking probability. The coding layer is deployed between the TCP and IP layers, without any modifications on the existing TCP/IP protocols. The experimental results of NS2 simulation and Mininet implementation show that CAPS significantly reduces the average flow completion time of short flows by ~30% -70% over the state-of-the-art multipath transmission schemes and achieves the high throughput for long flows with negligible traffic overhead.

In modern data centers, many flow-based and task-based schemes have been proposed to speed up the data transmission in order to provide fast, reliable services for millions of users. However, existing flow-based schemes treat all flows in isolation, contributing less to or even hurting user experience due to the stalled flows. Other prevalent task-based approaches, such as centralized and decentralized scheduling, are sophisticated or unable to share task information. In this work, we first reveal that relinquishing bandwidth of leading flows to the stalled ones effectively reduces the task completion time. We further present the design and implementation of a general supporting scheme that shares the flow-tardiness information through a receiver-driven coordination. Our scheme can be flexibly and widely integrated with the state-of-the-art TCP protocols designed for data centers, while making no modification on switches. Through the testbed experiments and simulations of typical data center applications, we show that our scheme reduces the task completion time by 70% and 50% compared with the flow-based protocols (e.g. DCTCP, L2DCT) and task-based scheduling (e.g. Baraat), respectively. Moreover, our scheme also outperforms other approaches by 18% to 25% in prevalent topologies of data center.

In data center networks, many network-intensive applications leverage large fan-in and many-to-one communication to achieve high performance. However, the special traffic patterns, such as micro-burst and high concurrency, easily cause TCP Incast problem and seriously degrade the application performance. To address the TCP Incast problem, we first reveal theoretically and empirically that alleviating packet burstiness is much more effective in reducing the Incast probability than controlling the congestion window. Inspired by the findings and insights from our experimental observations, we further propose a general supporting scheme Adaptive Pacing (AP), which dynamically adjusts burstiness according to the flow concurrency without any change on switch. Another feature of AP is its broad applicability. We integrate AP transparently into different TCP protocols (i.e., DCTCP, L2DCT and D2TCP). Through a series of large-scale NS2 simulations, we show that AP significantly reduces the Incast probability across different TCP protocols and the network goodput can be increased consistently by on average 7x under severe congestion.

Modern data centers host diverse hyper text transfer protocol (HTTP)-based services, which employ persistent transmission control protocol (TCP) connections to send HTTP requests and responses. However, the ON/OFF pattern of HTTP traffic disturbs the increase of TCP congestion window, potentially triggering packet loss at the beginning of ON period. Furthermore, the transmission performance becomes worse due to severe congestion in the concurrent transfer of HTTP response. In this paper, we provide the first extensive study to investigate the root cause of performance degradation of highly concurrent HTTP connections in data center network. We further present the design and implementation of TCP-TRIM, which employs probe packets to smooth the aggressive increase of congestion window in persistent TCP connection and leverages congestion detection and control at end-host to limit the growth of switch queue length under highly concurrent TCP connections. The experimental results of at-scale simulations and real implementations demonstrate that TCP-TRIM reduces the completion time of HTTP response by up to 80%, while introducing little deployment overhead only at the end hosts.

Modern data centers host diverse HTTP-based services, which employ persistent TCP connections to send HTTP requests and responses. However, the ON/OFF pattern of HTTP traffic disturbs the increase of TCP congestion window, potentially triggering packet loss at the beginning of ON period. Furthermore, the transmission performance becomes worse due to severe congestion in the concurrent transfer of HTTP response. In this work, we first reveal that the TCP's aggressive behavior in increasing congestion window causes TCP timeouts and throughput collapse. We further present the design and implementation of TCP-TRIM, which employs probe packets to smooth the aggressive increase of congestion window in persistent TCP connection, and leverages congestion detection and control at end-host to limit the growth of switch queue length under highly concurrent TCP connections. The experimental results of at-scale simulations and real implementations show that TCPTRIM reduces the completion time of HTTP response by up to 80%, while introducing little deployment overhead only at the end hosts.

In data center networks, many network-intensive applications typically suffer TCP incast throughput collapse when bursty concurrent TCP flows share a single bottleneck link. To address the TCP incast problem, we first reveal theoretically and empirically that controlling the number of concurrent flows is much more effective in reducing the incast probability than controlling the congestion window. We further propose a novel cross-layer design called Adaptive Request Schedule (ARS), which dynamically adjusts the number of concurrent TCP flows by batching application requests according to the congestion state acquired from transport layer. ARS is deployed only at the aggregator-side, while making no modification on hundreds or thousands of workers. Broad applicability is another advantage of ARS. We integrated ARS transparently (i.e., without modification) with DCTCP and TCP NewReno on NS2 simulation and a physical testbed, respectively. The experimental results show that ARS significantly reduces the incast probability across different TCP protocols and that the network goodput can be increased consistently by on average 6x under severe congestion.

Multipath transmission control protocol (MPTCP) allows a TCP connection to operate across multiple paths simultaneously and becomes highly attractive to support the emerging mobile devices with various radio interfaces and to improve resource utilization as well as connection robustness. The existing multipath congestion control algorithms, however, are mainly loss-based and prefer the paths with lower drop rates, leading to severe performance degradation in wireless communication systems, where random packet losses occur frequently. To address this challenge and improve the performance of MPTCP in wireless networks, this paper proposes a new mVeno algorithm, which makes full use of the congestion information of all the subflows belonging to a TCP connection in order to adaptively adjust the transmission rate of each subflow. Specifically, mVeno modifies the additive increase phase of Veno so as to effectively couple all subflows by dynamically varying the congestion window increment based on the receiving ACKs. The weighted parameter of each subflow for tuning the congestion window is determined by distinguishing packet losses caused by random error of wireless links or by network congestion. We implement mVeno in a Linux server and conduct extensive experiments both in test bed and in real WAN to validate its effectiveness. The performance results demonstrate that compared with the existing schemes, mVeno increases the throughput significantly, achieves load balancing, and can keep the fairness with regular TCP.

Modern data center nowadays leverages highly concurrent TCP connections between thousands of computer servers to achieve high system performance and service reliability. However, recent works have shown that, in the many-to-one and barrier-synchronized communication pattern, a large number of concurrent TCP connections suffer the TCP Incast problem due to packet drops in shallow-buffered Ethernet switches. This problem unavoidably leads to severe under-utilization of link capacity. In this work, we first reveal theoretically and empirically that controlling the IP packet size reduces the Incast probability much more effectively than controlling the congestion windows in the presence of severe congestion. We further present the design and implementation of Packet Slicing, a general supporting scheme that adjusts the packet size through a standard ICMP signaling method. Our method can be deployed on commodity switches with small firmware updates, while making no modification on end hosts. Another highlight of our work is Packet Slicing's broad applicability and effectiveness. We integrate Packet Slicing transparently (i.e., without modification) with three state-of-the-art TCP protocols designed for data centers on NS2 simulation and a physical testbed, respectively. The experimental results show that Packet Slicing remarkably improves network goodput across different TCP protocols by average 26× under severe congestion, while introducing little I/O performance impact on both switches and end hosts.

Summary Message replication is often used to improve the delivery ratio in delay‐tolerant networks because of the short‐lived wireless connectivity environment. However, packet replication may easily incur large resource consumption and finally result in network congestion. This paper proposes a probabilistic packet acceptance and drop algorithm (PAD), which adaptively controls congestion for delay‐tolerant networks. In PAD algorithm, the queue length and the input/output rate are combined to detect congestion. Based on the congestion state, each node determines the probability of accepting or dropping packets to obtain a good trade‐off between high delivery ratio and low overhead. Furthermore, based on the birth–death model, we construct the continuous‐time Markov chain to analyze the delivery ratio of a packet. Theory analysis and simulation results show that PAD increases the delivery ratio by more than 130% with least overhead. Meanwhile, it also achieves the shortest average end‐to‐end delay when the buffer of a node is severely limited. Copyright © 2015 John Wiley & Sons, Ltd.

Providing deadline-sensitive services is a challenge in data centers. Because of the conservativeness in additive increase congestion avoidance, current transmission control protocols are inefficient in utilizing the super high bandwidth of data centers. This may cause many deadline-sensitive flows to miss their deadlines before achieving their available bandwidths. We propose an Adaptive-Acceleration Data Center TCP, A2DTCP, which takes into account both network congestion and latency requirement of application service. By using congestion avoidance with an adaptive increase rate that varies between additive and multiplicative, A2DTCP accelerates bandwidth detection thus achieving high bandwidth utilization efficiency. At-scale simulations and real testbed implementations show that A2DTCP significantly reduces the missed deadline ratio compared to D2TCP and DCTCP. In addition, A2DTCP can co-exist with conventional TCP as well without requiring more changes in switch hardware than D2TCP and DCTCP.

Due to the intermittent connectivity of content-centric opportunistic networks, data dissemination is often performed by replication which easily leads to network congestion. Many existing message forwarding or dropping strategies provide congestion avoidance based on only the node-related metrics or the message-related metrics, i.e. the meeting history between nodes, the number of subscribers, message hop count etc. However the temporal validity of user interest is ignored in most cases. In fact, the interests of users may have different durations. A message is unnecessary to be forwarded when the nodes have lost interests in it. Therefore, we present the first attempt to apply the validity duration of user interest in message dropping decision and present a novel temporal validity based buffer management scheme. In this scheme, a multi-attribute utility is designed to decide the dropping priority of messages from the standpoints of node, message and interest. Simulation results prove the superiority of our scheme with respect to delivery ratio, dissemination speed and overhead.

End-to-end reliability in Delay Tolerant Networks (DTNs) is challenging and complicated because of high delay and absence of stable end-to-end path in the intermittently-connected mobile environment. Some existing acknowledgement mechanisms use active forwarding to provide the end-to-end reliability, while incurring excessive retransmissions or replications. The other passive mechanisms aim to reduce the storage overhead but may suffer a large delay. To improve the storage-delay tradeoff, we propose a Congestion Level based end-to-end ACKnowledgement (CL-ACK) mechanism, which adaptively adjusts the spread manner of ACK packets according to the ratio of drops over replications. Simulation results show that CL-ACK effectively controls resources consumption, reduces end-to-end delay, and achieves high message delivery rate.

The fair allocation of the resources among different nodes is one of the critical problems in wireless mesh networks. Existing solutions mainly focus on rate-limitation policies or distributed fair MAC schemes at the potential expense of total network utilization. This paper investigates a special starvation problem among TCP flows that are different hops away from the BS, as well as the recently proposed solution, the "Minimum Content Window" policy based on IEEE 802.11e. It is found that the aggregate throughput degrades sharply because the effect of this policy on the TCP congestion mechanism has been overlooked. This paper proposes a priority-based congestion control by using "Cross-Layer Explicit Congestion Notification". Analysis and simulation results demonstrate that our scheme can improve the fairness of TCP flows while the aggregate throughput is at least 20% higher than the "Minimum Content Window" policy.

In wireless LAN connected with Internet, the anomaly of wireless channel and the greedy nature of closed loop control in TCP lead to the congestion problem and significant unfair bandwidth distribution between TCP flows. In this paper, an ECN-based access point congestion control algorithm called APCC (AP congestion control) is proposed. The main properties of APCC are: (1) Using both wireless channel load and buffer queue length as congestion indicators, APC Censures low packet loss rate, low queue delay, and high goodput; (2) APCC guarantees the up/down TCP fairness by marking the ECN bit in TCP DATA and ACK packets; (3) Taking into account the wireless channel rate of each TCP flow, APCC sets different ECN marking probability for each flow to achieve the time fairness and high network efficiency.

Geographic routing is widely used in wireless sensor networks. The problem that most of the geographic routing algorithms, which adopt greedy algorithm as their basic routing strategies, have to face is the ldquolocal minimal phenomenardquo. In this paper, we propose a neighbors investment geographic routing algorithm (NIGRA), which is based on the geographic information of 2-hop neighbors. NIGRA adopts NIR (neighbors investment routing) algorithm as basic routing strategy, which makes nodes be aware of the existence of voids, so that the packet can bypass the dead-end nodes ahead of time to reduce the occurrence rate of local minimal phenomena. Furthermore, PATN, a planarization algorithm based on 2-hop neighbors, is also introduced. When NIGRA resorts to perimeter routing as recovery strategy, PATN ensures the success of perimeter routing through the planarization without extra overhead, and guarantees the delivery in UDG networks. Extensive simulations further show that NIR algorithm can significantly decrease the local minimal phenomena and NIGRA has better routing performance than GPSR.

In order to increase node lifetime and system throughput, the topology of wireless Ad Hoc networks can be controlled by changing the transmission power at each node. In this paper, we propose an energy-e±cient distributed topology- control algorithm, Varying Cone on Gabriel Graph (VCGG). By selectin

In multi-rate 802.11 wireless LAN, when the wireless node with lower data rate dominates shared channel usage time, the throughput experienced by other nodes transmitting at higher data rates will be drastically reduced. While most fair AQM (active queue management) algorithms are based on throughput fairness, we prove that temporal fair AQM algorithm can achieve more total throughput than throughput fair AQM algorithm in multi-rate 802.11 WLAN. The authors propose a temporal fair AQM-TFRED (temporal fair RED) on AP (access point) in multi-rate 802.11 WLAN. Taking into account the transmission rate of each flow, TFRED sets different drop probability for each flow going through the AP to guarantee equal channel usage time for each wireless node. Analysis and simulation results show that, compared with throughput fair AQM algorithms, TFRED achieves temporal fairness, per-flow throughput protection, and a 40% increase in total throughput.