This paper evaluates QUIC deployment and performance through active measurements of 1,572 Chinese and 1,953 non-Chinese websites. Results show clear regional and categorical differences: HTTP/3 support is 35.0% for non-Chinese websites but only 5.7% for Chinese websites. Performance gains also depend on network conditions, reducing transfer time by 17.06% in Chinese long-tail networks while slightly regressing by 0.79% in optimized low-latency CDN edge scenarios. Overall, QUIC’s deployment and benefits are highly scenario-dependent.

Modern clusters rely on centralized control planes, but decisions can be slow and fragile under control-plane degradations. We present Reflex, a bi-modal recovery design for clusters under control-plane degradation. Reflex adds a reflex-arc-like path that makes rapid takeover decisions from preprocessed local priorities while suppressing contention via lightweight coordination. After services are runnable, it performs steady-state reconstruction. Our evaluation shows bounded latency and robust conflict suppression under control-plane degradation and bursty failures.

Multi-agent workflows often generate repeated action requests when specialized agents interact with shared runtime objects. When requests target the same object under explicit compatibility conditions, a single execution result can serve multiple agents. This paper presents ActShare, a pre-execution coordination layer for reusable action requests. Before invoking a tool or a model, each agent emits a schema-constrained structured action request. ActShare compares the request against an active action table and a recent result cache. A request is attached only to a compatible owner action; otherwise, it is executed independently. When the owner action completes, ActShare writes the result once and dispatches it to all attached requests. We prototype ActShare in a stateful multi-agent debugging workflow and evaluate effect on repeated action reduction, token and tool-call savings.

Modern networks increasingly suffer from gray failures that are too fine-grained and short-lived for global re-optimization, while existing local mechanisms lack the context and control boundaries needed for effective bounded response. We propose CLEX, a contract-bounded local execution framework in which the controller defines per-device action boundaries and the device-side execution agent combines local semantics, runtime signals, and short-lived memory to select bounded local policy states that are realized through the local execution substrate. Implementation shows that CLEX enables agile local responses while reducing unnecessary action oscillation.

Due to the limited permissions for upgrading dual-side (i.e., server-side and client-side) loss tolerance schemes from the perspective of CDN vendors in a multi-supplier market, modern large-scale live streaming services are still using the automatic-repeat-request (ARQ) based paradigm for loss recovery, which only requires server-side modifications. In this paper, we first conduct a large-scale measurement study with up to 50 million live streams. We find that loss showsdynamicsand live streaming contains frequenton-off mode switchingin the wild. We further find that the recovery latency, enlarged by the ubiquitous retransmission loss, is a critical factor affecting live streaming’s client-side QoE (e.g., video freezing). We then propose an enhanced recovery mechanism called AutoRec, which can transform the disadvantages of on-off mode switching into an advantage for reducing loss recovery latency without any modifications on the client side. AutoRec allows users to customize overhead tolerance and recovery latency tolerance and adaptively adjusts strategies as the network environment changes to ensure that recovery latency meets user demands whenever possible while keeping overhead under control. We implement AutoRec upon QUIC and evaluate it via testbed and real-world commercial services deployments. The experimental results demonstrate the practicability and profitability of AutoRec.

Geo-distributed consensus protocols underpin modern databases across multiple data centers, yet their performance critically depends on effective leader management under wide-area network (WAN) dynamics. Legacy approaches fall short as they either ignore application-level factors or incur high switching overheads. This paper presents GeoLM, a lightweight, performance-oriented leader management framework. GeoLM formalizes a utility model that jointly captures network delays, node popularity, and read-write ratios, and integrates three mechanisms: (i) Target Selection to identify promising leaders, (ii) Switching Damping to bound oscillations and ensure stable leadership, and (iii) Leader Handover with log pre-synchronization to minimize downtime. GeoLM is implemented in ETCD-Raft and seamlessly extends existing protocols without altering log replication. Comprehensive evaluation-including trace-driven simulations with AWS latency traces and real-world deployments across multiple regions-demonstrates that GeoLM reduces average latency by up to 30.77%, tail latency by up to 49.75%, and improves throughput by up to 28% over baseline protocols (depending on the comparison). GeoLM further delivers consistently better performance than state-of-the-art protocols such as SwiftPaxos, CURP, and EPaxos, while maintaining strong consistency guarantees.

The widespread adoption of database middleware for supporting distributed transaction processing is prevalent in numerous applications, with heterogeneous data sources deployed across national and international boundaries. However, transaction processing performance significantly drops due to the high network latency between the middleware and data sources and the long lock contention span, where transactions may be blocked while waiting for the locks held by concurrent transactions. In this paper, we propose GeoTP, a latency-aware geo-distributed transaction processing approach in database middleware. GeoTP incorporates three key techniques to enhance performance in geo-distributed scenarios. First, we propose a decentralized prepare mechanism to reduce network round-trips for distributed transactions. Second, we design a latency-aware scheduler to minimize the lock contention span by strategically delaying the lock acquisition. Third, heuristic optimizations are proposed for the scheduler to reduce the lock contention span further. We implemented GeoTP on Apache Shardingsphere, a state-of-the-art middleware, and extended it into Apache ScalarDB. Experimental results on YCSB and TPC-C demonstrate that GeoTP achieves up to 17.7x performance improvement.

In large-scale data stream analytics, accurate identification of Persistent and Infrequent (PI) flows is of great significance for monitoring and protecting against network attacks such as Advanced Persistent Threats (APT). However, existing research focuses mainly on detecting frequent flows or persistent flows, with insufficient studies on the characterization and detection methods for PI flows. Based on the analysis of sufficient APT flows, we propose a method that combines global and local features to effectively characterize PI flows. Further, we propose a novel sketch algorithm called WisePIFinder, which aims to detect PI flows more accurately and efficiently in realtime. The key idea is to continuously filter out non-PI flows while detecting flow persistence, to achieve accurate statistics on PI flows. Experimental results show that WisePIFinder improves the F1 Score by at least 20 % and insertion throughput by at least 60 % compared to the state-of-the-art solution for detecting PI flows. All related codes have been open-sourced on GitHub.

The paradigm of Intelligent DataPlane (IDP) embeds deep learning (DL) models on the network dataplane to enable intelligent traffic analysis at line-speed. However, the current use of the match-action table (MAT) abstraction on the dataplane is misaligned with DL inference, leading to several key limitations, including accuracy degradation, limited scale, and lack of generality. This paper proposes Pegasus to address these limitations. Pegasus translates DL operations into three dataplane-oriented primitives to achieve generality: Partition, Map, and SumReduce. Specifically, Partition "divides" high-dimensional features into multiple low-dimensional vectors, making them more suitable for the dataplane; Map "conquers" computations on the low-dimensional vectors in parallel with the technique of Fuzzy Matching, while SumReduce "combines" the computation results. Additionally, Pegasus employs Primitive Fusion to merge computations, improving scalability. Finally, Pegasus adopts full-precision weights with fixed-point activations to improve accuracy. Our implementation on a P4 switch demonstrates that Pegasus can effectively support various types of DL models, including Multi-Layer Perceptron (MLP), Recurrent Neural Network (RNN), Convolutional Neural Network (CNN), and AutoEncoder models on the dataplane. Meanwhile, Pegasus outperforms state-of-the-art approaches with an average accuracy improvement of up to 22.8%, along with up to 248× larger model size and 212× larger input scale.

Traffic data contains deep domain-specific knowledge, making labeling challenging, and the lack of labeled data adversely impacts the accuracy of learning-based traffic analysis. The pre-training technology is widely adopted in the fields of vision and natural language to address the problem of limited labeled data. However, the exploration in the domain of traffic analysis remains insufficient. This paper proposes an efficient pre-training model, TrafficFormer, for traffic data. In the pre-training stage, TrafficFormer introduces a fine-grained multi-classification task to enhance the representation capabilities of traffic data; in the fine-tuning stage, TrafficFormer proposes a traffic data augmentation method utilizing the random initialization feature of fields, which helps the traffic model focus on key information. We evaluate TrafficFormer using both traffic classification tasks and protocol understanding tasks. The experimental results show that TrafficFormer achieves superior performance on six traffic classification datasets, with improvements of up to 10% in the F1 score and demonstrates significantly superior protocol understanding capabilities compared to existing traffic pre-training models.

Abstract In virtual private network (VPN) tunnel mode, the entire original packet, including the header’s five-tuple information, is encrypted, which prevents traditional scheduling algorithms from evenly distributing packets to central processing unit (CPU) cores based on packet header information. To address the need for data security and encrypted packet scheduling, we propose a novel framework, named REFS (receive encrypted flow steering), for accelerated receive encrypted flow steering. This work creatively adopts a new method that allows encrypted packets to be distributed across CPU cores without decrypting them, overcoming limitations of traditional scheduling approaches. It efficiently distributes encrypted packets across CPU cores, enabling dynamic allocation of CPU resources. A key feature of REFS is its ability to perform this distribution without decrypting the packets, which enhances dynamic load balancing and improves system responsiveness. When integrated into the Linux kernel’s VPN functionality, REFS can potentially increase throughput by up to 50% compared to WireGuard, which is a benchmark for kernel-based VPN performance. Upon integration of REFS into userspace, network performance shows significant improvements: throughput doubles, while latency is reduced by 80%.

Packet losses significantly impact the user experience of content delivery network (CDN) services such as live streaming and data backup-and-archiving. However, our production network measurement studies show that the legacy loss recovery is far from satisfactory due to the wide-area loss characteristics (i.e., dynamics and burstiness) in the wild. In this paper, we propose a sender-side Adaptive ReTransmission scheme, ART, which minimizes the recovery time of lost packets with minimal redundancy cost. Distinguishing itself from forward-error-correction (FEC), which preemptively sends redundant data packets to prevent loss, ART functions as an automatic-repeat-request (ARQ) scheme. It applies redundancy specifically to lost packets instead of unlost packets, thereby addressing the characteristic patterns of wide-area losses in real-world scenarios. We implement ART upon QUIC protocol and evaluate it via both trace-driven emulation and real-world deployment. The results show that ART reduces up to 34% of flow completion time (FCT) for delay-sensitive transmissions, improves up to 26% of goodput for throughput-intensive transmissions, reduces 11.6% video playback rebuffering, and saves up to 90% of redundancy cost.

This paper proposes ElastiCast, a novel Bluetooth Low Energy (BLE) broadcast mode that reduces the neighbor discovery latency in offline finding networks (OFNs). ElastiCast adapts the broadcast mode of the lost devices to the scan modes of the finder devices, considering their diversity. We start with an overview of OFNs, followed by a detailed analysis of the issues and challenges of existing solutions, which motivates the design of ElastiCast. Then we provide Blender, a simulator that models the neighbor discovery behavior of different broadcasters and scanners. By adopting Blender, ElastiCast can be implemented with three components: Local Optima Estimation, Common Interest Extraction, and Interval Multiplexing, in which we capture the key features of BLE neighbor discovery and globally optimize the broadcast mode interacting with diverse scan modes. Experimental evaluation results and commercial product deployment experience demonstrate that ElastiCast is effective in achieving stable and bounded neighbor discovery latency within the power budget.

ECN marking has been integrated into datacenter switches to enable high-throughput and low-latency transport. We observe that current marking schemes are coarse-grained: they blindly mark all flows when congestion occurs, causing large flows to occupy undeserved bandwidth and preventing newly arriving small flows from finishing quickly. In this paper, we propose DiffECN, a differential marking strategy that marks only the flows that are the culprits of congestion and protects the remaining flows from being limited. We have implemented it in the Barefoot Tofino switch and performed extensive evaluations via both physical testbed and large-scale simulations. The results show that DiffECN can restrain flows responsible for congestion successfully while providing desirable network performance. For instance, compared to the legacy way of ECN marking, DiffECN achieves up to 32.5% (40.1%) lower average (99th percentile) flow completion time (FCT) for small flows while delivering similar FCT for large flows under production workloads.

Random Early Detection (RED) has been integrated into datacenter switches as a fundamental Active Queue Management (AQM) for decades. The accurate configuration of RED parameters is crucial to achieving high throughput and low latency. However, due to the highly dynamic nature of workloads in datacenter networks, maintaining consistently high performance with statically configured RED thresholds poses a challenge. Prior work applies reinforcement learning to predict proper thresholds, but their real-world deployment has been hindered by poor tail performance caused by instability. In this paper, we propose$\textsf {PRED}$, a novel system that enables automatic and stable RED parameter adjustment in response to traffic dynamics. Specifically, the system employs a Multiplicative-Increase Multiplicative-Decrease (MIMD) strategy to dynamically adapt to flow concurrency while utilizing an Additive-Increase Additive-Decrease (AIAD) mechanism to adapt to flow distribution. We perform extensive evaluations on our physical testbed and large-scale simulations. The results demonstrate that$\textsf {PRED}$can keep up with the real-time network dynamics generated by realistic workloads. For instance, compared with the static-threshold-based methods,$\textsf {PRED}$keeps 66% shorter switch queue length and obtains up to 80% lower Flow Completion Time (FCT). Compared with the state-of-the-art learning-based method,$\textsf {PRED}$reduces the tail FCT by 34%.

The first-frame delay is an essential indicator for evaluating the performance of cloud CDN vendors and affects the client-side QoE of live streaming. Instead of the traditional way of tuning the initial congestion window (cwnd) for all connections to a fixed value based on expert experience, this paper explores the using of transport signals unique to each connection (e.g., application-layer framing, historical QoS metrics) to initialize the sending parameters for each connection. Thus we propose Wira, a first-frame optimization mechanism that adjusts both initial cwnd and initial rate, which are two key parameters for decreasing the first-frame completion time (FFCT). Particularly, Wira provides cross-layer Frame Perception that parses frames and adapts the initial cwnd to the first-frame size. Meanwhile, Wira introduces the Transport Cookie to enable cloud-client collaborations, in which the historical QoS metrics from the clients can be reported and reused by rate initialization in the stateless cloud. This assures the initial rate matches the actual network conditions while avoiding non-trivial storage overhead in the cloud. We implement Wira upon QUIC and evaluate it via real-world deployments of commercial services. Results demonstrate the profitability of Wira, in which the average and 90th-percentile FFCT are reduced by 10.6% and 16.7%, respectively.

Wi-Fi Direct has been serving a progressively wide range of applications such as device-to-device file sharing, face-to-face interactive gaming, and wireless projection. However, when TCP meets Wi-Fi Direct, we find that two independent control loops exist, i.e., the transport-layer control loop and the link-layer control loop. First, these functionally redundant loops result in spectrum inefficiency. Second, the lack of effective information interaction between layers results in local optimal. To tackle these issues, this paper proposes Wi-Fi Direct TCP (WDTCP), a performant TCP that provides a full protocol design of the acknowledgment de-redundancy and explicit-capacity-based congestion control. WDTCP tightly couples the two control loops by capturing the WiFi Direct’s key feature of one-hop communication. Evaluation results demonstrate that WDTCP can maximize bandwidth utilization while keeping low latency. For instance, compared to legacy TCP, WDTCP improves throughput by up to 49.2% and reduces average and 95th latency by up to 32.4% and 50.7%, respectively.

This paper proposes the novel concept of reasoning-based Bluetooth Low-Energy (BLE) neighbor discovery, an indirect paradigm of device-to-device sensing to address challenges (e.g., interference and power limitations) where direct sensing falls short. Inspired by the classical Rule of Syllogism, reasoning-based BLE neighbor discovery abstracts the device-to-device sensing as the presence detection of a BLE signal in a certain space. It deduces the presence of the BLE signal according to the presence of the Wi-Fi signal through the historical correlation between BLE and Wi-Fi. To demonstrate the feasibility of this new neighbor discovery paradigm, we report the design and evaluation of a prototype called ReND. By leveraging the complementary strengths of Wi-Fi and BLE, ReND reduces up to 91.3% and 65.9% of the 50thand 95thpercentile BLE neighbor discovery latency, respectively. We further discuss the feasibility and incentive of ReND in the Polygon’s Mumbai Testnet public blockchain.

Due to the limited permissions for upgrading dual-side (i.e., server-side and client-side) loss tolerance schemes from the perspective of CDN vendors in a multi-supplier market, modern large-scale live streaming services are still using the automatic-repeat-request (ARQ) based paradigm for loss recovery, which only requires server-side modifications. In this paper, we first conduct a large-scale measurement study with up to 50 million live streams. We find that loss shows dynamics and live streaming contains frequent on-off mode switching in the wild. We further find that the recovery latency, enlarged by the ubiquitous retransmission loss, is a critical factor affecting live streaming's client side QoE (e.g., video freezing). We then propose an enhanced recovery mechanism called AutoRec, which can transform the disadvantages of on-off mode switching into an advantage for reducing loss recovery latency without any modifications on the client side. AutoRec also adopts an online learning-based policy to fit the dynamics of loss, balancing the tradeoff between the recovery latency and the incurred overhead. We implement AutoRec upon QUIC and evaluate it via both testbed and real-world commercial services deployments. The experimental results demonstrate the practicability and profitability of AutoRec, in which the average times and duration of client-side video freezing can be lowered by 11.4% and 5.2%, respectively.

Converged Ethernet employs Priority-based Flow Control (PFC) to provide a lossless network. However, issues caused by PFC, including victim flow, congestion spreading, and deadlock, impede its large-scale deployment in production systems. The fine-grained experimental observations on switch buffer occupancy find that the root cause of these performance problems is a mismatch of sending rates between end-to-end congestion control and hop-by-hop flow control. Resolving this mismatch requires the switch to provide an additional buffer, which is not supported by the classic dynamic threshold (DT) policy in current shared-buffer commercial switches. In this paper, we propose Selective-PFC (SPFC), a practical buffer management scheme that handles such mismatch. Specifically, SPFC incrementally modifies DT by proactively detecting port traffic and adjusting buffer allocation accordingly to trigger PFC PAUSE frames selectively. Extensive case studies demonstrate that SPFC can reduce the number of PFC PAUSEs on non-bursty ports by up to 69.0%, and reduce the average flow completion time by up to 83.5% for large victim flows.

Packet losses significantly impact the user experience of wide-area applications such as content distribution and remote procedure call (RPC) based services. However, our production network measurement studies show that the legacy loss recovery is far from satisfactory due to the wide-area loss characteristics (i.e., dynamics and burstiness) in the wild. In this paper, we propose a sender-side Adaptive ReTransmission scheme, ART, which minimizes the recovery time of lost packets with minimal redundancy cost. Distinguishing itself from forward-error-correction (FEC), which preemptively sends redundant data packets to prevent loss, ART functions as an automatic-repeat-request (ARQ) scheme. It applies redundancy specifically to lost packets instead of unlost packets, thereby addressing the characteristic patterns of wide-area losses in real-world scenarios. We implement ART upon QUIC protocol and evaluate it via both trace-driven emulation and real-world deployment. The results show that ART reduces up to 34% of flow completion time (FCT) for delay-sensitive transmissions, improves up to 28 % of goodput for throughput-intensive transmissions, and saves up to 90% of redundancy cost.

Recently, such industrial pioneers as Apple and Samsung have offered a new generation of offline finding network (OFN) that enables crowd search for missing devices without leaking private data. Specifically, OFN leverages nearby online finder devices to conduct neighbor discovery via Bluetooth Low Energy (BLE), so as to detect the presence of offline missing devices and report an encrypted location back to the owner via the Internet. The user experience in OFN is closely related to the success ratio (possibility) of finding the lost device, where the latency of the prerequisite stage, i.e., neighbor discovery, matters. However, the crowd-sourced finder devices show diversity in scan modes due to different power modes or different manufacturers, resulting in local optima of neighbor discovery performance. In this paper, we present a brand-new broadcast mode called ElastiCast to deal with the scan mode diversity issues. ElastiCast captures the key features of BLE neighbor discovery and globally optimizes the broadcast mode interacting with diverse scan modes. Experimental evaluation results and commercial product deployment experience demonstrate that ElastiCast is effective in achieving stable and bounded neighbor discovery latency within the power budget.

As the ubiquitous phenomenon occurs in applications such as live streaming and video conferencing, the on-off traffic pattern is regarded as a disadvantage for congestion control. However, we argue that it can be transformed as an advantage for accelerating loss recovery. In this paper, we report the design of TOO, a loss recovery acceleration mechanism that tames on-off patterns for loss duplicate reinjection without incurring non-trivial traffic overhead.

This paper demonstrates PolyCC, a general framework for simultaneous operation of poly-algorithmic congestion control. PolyCC gains benefits from taking advantage of the complementary among already existing congestion controllers.

TCP incast has become a practical problem for high-bandwidth, low-latency transmissions, resulting in throughput degradation of up to 90% and delays of hundreds of milliseconds, severely impacting application performance. However, in virtualized multi-tenant data centers, host-based advancements in the TCP stack are hard to deploy from the operators’ perspective. Operators only provide infrastructure in the form of virtual machines, in which only tenants can directly modify the end-host TCP stack. In this paper, we present R-AQM, a switch-powered reverse ACK active queue management (R-AQM) mechanism for enhancing ACK-clocking effects through assisting legacy TCP. Specifically, R-AQM proactively intercepts ACKs and paces the ACK-clocked in-flight data packets, preventing TCP from suffering incast collapse. We implement and evaluate R-AQM in NS-3 simulation and NetFPGA-based hardware switch. Both simulation and testbed results show that R-AQM greatly improves TCP performance under heavy incast workloads by significantly lowering packet loss rate, reducing retransmission timeouts, and supporting 16 times (i.e., 60 to 1000) more senders. Meanwhile, the forward queuing delays are also reduced by 4.6 times.

For the widely used Bluetooth Low-Energy (BLE) neighbor discovery, the parameter configuration of neighbor discovery directly decides the results of the trade-off between discovery latency and power consumption. Therefore, it requires evaluating whether any given parameter configuration meets the demands. The existing solutions, however, are far from satisfactory due to unsolved issues. In this paper, we propose Blender, a simulation framework that produces a determined and full probabilistic distribution of discovery latency for a given parameter configuration. To capture the key features in practice, Blender provides adaption to the stochastic factors such as the channel collision and the random behavior of the advertiser. Evaluation results show that, compared with the state-of-art simulators, Blender converges closer to the traces from the Android-based realistic estimations. Blender can be used to guide parameter configuration for BLE neighbor discovery systems where the trade-off between discovery latency and power consumption is of critical importance.

The emerging applications including cloud AR/VR gaming, ultra-high definition (UHD) streaming, Metaverse, etc. imply the demand for ultra-high bandwidth transmission in the wide area network (WAN). During the past four decades, a great number of high-speed TCP variants have been proposed to improved the throughput for WAN transmission. This paper conducts a measurement study on loss recovery for high-speed transmission, and exposes that receive buffer starvation is a result of capability mismatch between loss recovery and high-speed TCP advancements such as non-loss-based congestion control. To mitigate this mismatch problem, an opportunistic random redundant retransmission (OR3) algorithm, as well as its TCP implementation TCP-OR3, is proposed. OR3 accelerates loss recovery by minimizing the maximum retransmission times of each packet. Experiment results shows that TCP-OR3 alleviates receive buffer starvation and tackles the bottleneck of the legacy approaches such as parallel TCP in the case of high-speed transmission.

Recent advances in high-speed rails (HSRs), coupled with user demands for communication on the move, are propelling the need for acceptable quality of experience (QoE) in high-speed mobility environments. However, with throughput declining significantly the QoE on existing HSRs is still far from satisfactory. In order to improve QoE on HSRs, this paper seeks to answer the question regarding which is better of two options: the selection of the best cellular carrier applying single-path TCP or the conjunction of multiple carriers applying Multi-path TCP (MPTCP). To this end, we carefully design comparison experiments using the two approaches on HSRs with a peak speed of 310 km/h. Measurement study on MPTCP performance shows that generally carrier conjunction gives similar performance as carrier selection. We take an in-depth analysis of the details of the instances, and for the first time expose the phenomenon called subflow degradation. We further confirm that subflow degradation of MPTCP occurs due to its poor adaptability to frequent handoffs. We believe these insights can provide valuable guidance for the design, implementation, and deployment of transmission protocols in high-speed mobility environments.

The emerging applications including AR/VR inter-active gaming, ultra-high-definition live streaming, 4K wireless projection, Metaverse, etc. imply the demand for ultra-low latency and ultra-high bandwidth wireless transmission. The legacy kernel TCP stack is not fully satisfactory because it induces the CPU bottleneck on hosts. In this paper, we propose Wireless-RDMA (W-RDMA) that enables RDMA in wireless networks to tackle the CPU bottleneck issue on wireless hosts. The feasibility of W-RDMA is demonstrated through testbed experiments. Technical challenges and future opportunities are further discussed. We believe it is a small but crucial step for enabling RDMA for wireless transmission.

TCP incast has become a practical problem for high-bandwidth, low-latency transmissions, resulting in throughput degradation of up to 90% and delays of hundreds of milliseconds, severely impacting application performance. However, in virtualized multi-tenant data centers, host-based advancements in the TCP stack are hard to deploy from the operators perspective. Operators only provide infrastructure in the form of virtual machines, in which only tenants can directly modify the end-host TCP stack. In this paper, we present R-AQM, a switch-powered reverse ACK active queue management (R-AQM) mechanism for enhancing ACK-clocking effects through assisting legacy TCP. Specifically, R-AQM proactively intercepts ACKs and paces the ACK-clocked in-flight data packets, preventing TCP from suffering incast collapse. We implement and evaluate R-AQM in NS-3 simulation and NetFPGA-based hardware switch. Both simulation and testbed results show that R-AQM greatly improves TCP performance under heavy incast workloads by significantly lowering packet loss rate, reducing retransmission timeouts, and supporting 16 times (i.e., 60 → 1000) more senders. Meanwhile, the forward queuing delays are also reduced by 4.6 times.

Anycast content delivery networks rely on the underlying routing to schedule clients to their nearby service nodes, which however is not natively aware of server load or path latency. Requests burst from some regions may cause overload and hurt user experience. This scenario demands quickly adjusting clients to other nearby servers with available capacity. However, state-of-the-art solutions do not work well. On one hand, native routing-based scheduling is not flexible and precise enough, which may cause cascading damage and interrupt ongoing sessions. On the other hand, centralized algorithm is vulnerable and not responsive due to high complexity. We propose a practical distributed hybrid load management model to solve load burst problem. First, the hybrid mechanism leverages flexible DNS-based redirection, which can schedule at per-request granularity without interrupting ongoing sessions. Second, the distributed model is responsive by reducing computation overhead and theoretically guarantees to converge to the optimal solution. Based on the model, we further propose an cooperative and two heuristic distributed algorithms. At last, using a measurement dataset, we demonstrate their effectiveness and scalability, and illustrate how to adapt them to different scenarios.

The shared nature of the wireless medium induces contention between data transport and backward signaling, such as acknowledgment. The current way of TCP acknowledgment induces control overhead which is counter-productive for TCP performance especially in wireless local area network (WLAN) scenarios. In this paper, we present a new acknowledgment called TACK (“Tame ACK”), as well as its TCP implementation TCP-TACK. TACK seeks to minimize ACK frequency, which is exactly what is required by transport. TCP-TACK works on top of commodity WLAN, delivering high wireless transport goodput with minimal control overhead in the form of ACKs, without any hardware modification. Evaluation results show that TCP-TACK achieves significant advantages over legacy TCP in WLAN scenarios due to less contention between data packets and ACKs. Specifically, TCP-TACK reduces over 90% of ACKs and also obtains an improvement of up to 28% on goodput. A TACK-based protocol is a good replacement of the legacy TCP to compensate for scenarios where the acknowledgment overhead is non-negligible.

The shared nature of the wireless medium induces contention between data transport and backward signaling, such as acknowledgement. The current way of TCP acknowledgment induces control overhead which is counter-productive for TCP performance especially in wireless local area network (WLAN) scenarios.

The increasing data requirements of Internet applications have driven a dramatic surge in developing new programming paradigms and complex scheduling algorithms to handle data-intensive workloads. Due to the expanding volume and the variety of such flows, their raw data are often processed on Intermediate Processing Nodes (IPNs) before being sent to servers. However, the intermediate processing constraint is rarely considered in existing flow computing models. This paper aims to minimize the tardiness of data-intensive applications in the presence of intermediate processing constraint. Motivating cases show that the tardiness is affected by both IPN locations and flow dispatching strategies. Based on the observation that dispatching flows to IPNs is essentially building a matching between flows and IPNs, a novel solution is proposed based on matching theory. In the deployment phase, a tardiness-aware deferred acceptance algorithm is developed to optimize IPN locations. In the operation phase, the Power-of-D paradigm and matching theory are combined together to dispatch flows efficiently. Evaluation results show that our solution effectively minimizes the total tardiness of data-intensive applications in heterogeneous systems.

Recent advances in high speed rails (HSRs) are propelling the need for acceptable network service in high speed mobility environments. However, previous studies show that the performance of traditional single-path transmission degrades significantly during high speed mobility due to frequent handoff. Multi-path transmission with multiple carriers is a promising way to enhance the performance, because at any time, there is possibly at least one path not suffering a handoff. In this paper, for the first time, we measure multi-path TCP (MPTCP) with two cellular carriers on HSRs with a peak speed of 310km/h. We find a significant difference in handoff time between the two carriers. Moreover, we observe that MPTCP can provide much better performance than TCP in the poorer of the two paths. This indicates that MPTCP's robustness to handoff is much higher than TCP's. However, the efficiency of MPTCP is far from satisfactory. MPTCP performs worse than TCP in the better path most of the time. We find that the low efficiency can be attributed to poor adaptability to frequent handoff by MPTCP's key operations in sub-flow establishment, congestion control and scheduling. Finally, we discuss possible directions for improving MPTCP for such scenarios.

With the prevalence of broadband network and wireless mobile network accesses, distributed interactive applications (DIAs) such as online gaming have attracted a vast number of users over the Internet. The deployment of these systems, however, comes with peculiar hardware/software requirements on the user consoles. Recently, such industrial pioneers as Gaikai, Onlive, and Ciinow have offered a new generation of cloud-based DIAs (CDIAs), which shifts the necessary computing loads to cloud platforms and largely relieves the pressure on individual user's consoles. In this paper, we aim to understand the existing CDIA framework and highlight its design challenges. Our measurement reveals the inside structures as well as the operations of real CDIA systems and identifies the critical role of cloud proxies. While its design makes effective use of cloud resources to mitigate client's workloads, it may also significantly increase the interaction latency among clients if not carefully handled. Besides the extra network latency caused by the cloud proxy involvement, we find that computation-intensive tasks (e.g., game video encoding) and bandwidth-intensive tasks (e.g., streaming the game screens to clients) together create a severe bottleneck in CDIA. Our experiment indicates that when the cloud proxies are virtual machines (VMs) in the cloud, the computation-intensive and bandwidth-intensive tasks may seriously interfere with each other. We accordingly capture this feature in our model and present an interference-aware solution. This solution not only smartly allocates workloads but also dynamically assigns capacities across VMs based on their arrival/departure patterns.

Presents corrections to the paper, “Modeling, analysis, and implementation of universal acceleration platform across online video sharing sites,” (Xu, K. et al), IEEE Trans. Serv. Comput., vol. 11, no. 3, pp. 534–548, May/Jun. 2018.

User-generated video sharing service has attracted a vast number of users over the Internet. The most successful sites, such as YouTube and Youku, now enjoy millions of videos being watched every day. Yet, given limited network and server resources, the user experience of existing video sharing sites (VSSes) is still far from being satisfactory. To mitigate such a problem, peer-to-peer (P2P) based video accelerators have been widely suggested to enhance the video delivery on VSSes. In this paper, we find that the interference of multiple accelerators will lead to a severe bottleneck across the VSSes. Our model analysis shows that a universal video accelerator can naturally achieve better performance with lower deployment cost. Based on this observation, we further present the detailed design of Peer-to-Peer Video Accelerator (PPVA), a real-world system for universal and transparent P2P accelerating. Such a system has already attracted over 180 million users, with 48 million video transactions every day. We carefully examine the PPVA performance from extensive measurements. Our trace analysis indicates that it can significantly reduce server bandwidth cost and accelerate the video download speed by 80 percent.

Cloud radio access network (C-RAN) and mobile edge computing (MEC) have emerged as promising candidates for the next generation access network techniques. Unfortunately, although MEC tries to utilize the highly distributed computing resources in close proximity to user equipments equipments (UE), C-RAN suggests to centralize the baseband processing units (BBU) deployed in radio access networks. To better understand and address such a conflict, this paper closely investigates the MEC task offloading control in C-RAN environments. In particular, we focus on perspective of matching problem. Our model smartly captures the unique features in both MEC and C-RAN with respect to communication and computation efficiency constraints. We divide the cross-layer optimization into the following three stages: (1) matching between remote radio heads (RRH) and UEs, (2) matching between BBUs and UEs, and (3) matching between mobile clones (MC) and UEs. By applying the Gale-Shapley Matching Theory in the duplex matching framework, we propose a multi-stage heuristic to minimize the refusal rate for user's task offloading requests. Trace-based simulation confirms that our solution can successfully achieve near-optimal performance in such a hybrid deployment.

The increasing data requirement of Internet applications has driven a dramatic surge in developing new programming paradigms and complex scheduling algorithms to handle data-intensive workloads. Due to the expanding volume and the variety of such flows, their raw data are often processed on intermediate processing nodes before being sent to servers. The intermediate processing constraints are however not yet considered in existing task and flow computing models. In this paper, we aim to minimize the total tardiness of all flows in the presence of intermediate processing constraints. We build a model to consider Tardiness-aware Flow Scheduling with Processing constraints (TFS-P), which is unfortunately NP-Hard. Hence, we propose a heuristic Routing and Scheduling duplex MATching (RSMAT) framework based on the classic Gale-Shapley Matching Theory. We find that the problem can be well-addressed by classic Deferred Acceptance (DA) algorithm, in which the match is stable but inefficient for the model. We therefore propose the Tardiness-aware Deferred Acceptance algorithm with Dynamical Quota (TDA-DQ). This algorithm is enhanced by overcoming the inefficient stability and smartly considering the dynamical quota in the system. The evaluation compares TDA-DQ to the lower bound obtained by a modified subgradient optimization algorithm. The result indicates that TDA-DQ can achieve near-optimal performance for data-intensive applications.

The open shortest path first (OSPF) protocol has been widely applied to intra-domain routing in today's Internet. Since a router running OSPF distributes traffic uniformly over equal-cost multi-path (ECMP), the OSPF-based optimal traffic engineering (TE) problem (i.e., deriving optimal link weights for a given traffic demand) is computationally intractable for large-scale networks. Therefore, many studies resort to multi-protocol label switching (MPLS) based approaches to solve the optimal TE problem. In this paper we present a generalized routing framework to realize the optimal TE, which can be potentially implemented via OSPFor MPLS-based approaches. We start with viewing the conventional optimal TE problem in a fresh way, i.e., optimally allocating the residual capacity to every link. Then we make a generalization of network utility maximization (NUM) to close this problem, where the network operator is associated with a utility function of the residual capacity to be maximized. We demonstrate that under this framework, the optimal routes resulting from the optimal TE are also the shortest paths in terms of a set of non-negative link weights that are explicitly determined by the optimal residual capacity and the objective function. The network entropy maximization theory is employed to enable routers to exponentially, instead of uniformly, split traffic over ECMP. The shortest-path penalizing exponential flow-splitting (SPEF) is designed as a link-state protocol with hop-by-hop forwarding to implement our theoretical findings. An alternative MPLS-based implementation is also discussed here. Numerical simulation results have demonstrated the effectiveness of the proposed framework as well as SPEF.

In mobile Internet era, wireless traffic has become a rare resource and there is no effective ways for users to share their unused traffic with each other. This paper introduces a system solution requiring no sophisticated hardware. An incentive mechanism is designed and implemented in a novel system named Traffic Sharing Platform (TSP) for mobile users, which can optimize network resource configuration and achieve Pareto optimality of the society. Simulation results show the TSP is available and the incentive mechanism is effective.

With the popularity of information and content items that can be cached within ISP networks, developing high-quality and efficient content distribution approaches has become an important task in future internet architecture design. As one of the main techniques of content distribution, in-network caching mechanism has attracted attention from both academia and industry. However, the general evaluation model of in-network caching is seldom discussed. The trade-off between economic cost and the deployment of in-network caching still remains largely unclear, especially for heterogeneous applications. We take a first yet important step towards the design of a better evaluation model based on the Application Adaptation CapaciTy (2ACT) of the architecture to quantify the trade-off in this paper. Based on our evaluation model, we further clarify the deployment requirements for the in-network caching mechanism. Based on our findings, ISPs and users can make their own choice according to their application scenarios. © 2013 IEEE.