Traditional congestion management mechanisms, including Explicit Congestion Notification and Priority Flow Control, operate reactively, addressing congestion only after queue thresholds are exceeded. This fundamental limitation becomes increasingly critical as modern data center workloads generate microbursts lasting under 100 milliseconds, creating transient bottlenecks and elevated tail latency. AI-Driven Dynamic Load Balancing (AI-DLB) introduces an intelligent, predictive framework that combines real-time telemetry with machine learning inference to forecast congestion events and proactively redistribute traffic flows. The system employs a closed-loop control architecture integrating supervised learning for short-term congestion prediction and reinforcement learning for continuous policy optimization. By analyzing queue depth, ECN marks, link utilization, and RTT trends, AI-DLB enables sub-second load redistribution before congestion manifests. Simulation results on spine-leaf topologies demonstrate substantial reductions in tail latency and faster convergence compared to conventional mechanisms. The framework operates as a complementary enhancement to existing protocols, establishing a foundation for self-optimizing, intent-driven data center fabrics that bridge predictive analytics with autonomous control.

A Proposal of Congestion Avoidance Communication Method Using Explicit Congestion Notification in HTTP/3 Applications

In the rapidly evolving landscape of 5G networks, ensuring Quality of Service (QoS) and efficient congestion management are paramount to meet the high demands of modern applications. This paper presents a comprehensive study on the integration of Smart Explicit Congestion Notification (ECN) and Resource Reservation Protocol (RSVP) protocols, enhanced by advanced Artificial Intelligence (AI) techniques, to optimize QoS and manage congestion in 5G networks. We propose a novel framework that combines AI-driven algorithms with traditional congestion control mechanisms to dynamically adjust network parameters based on real-time traffic conditions and user demands. The Smart ECN protocol utilizes machine learning to predict and mitigate congestion before it impacts network performance, while the enhanced RSVP protocol leverages AI for intelligent resource allocation and QoS management. Through simulations and theoretical analysis, we demonstrate the effectiveness of our approach in improving network efficiency, reducing latency, and ensuring reliable service delivery across diverse 5G scenarios. This study underscores the potential of AI to revolutionize congestion control and QoS optimization, paving the way for more resilient and adaptive 5G networks. KEYWORDS: artificial intelligence (AI), machine learning (ML), Explicit Congestion Notification (ECN) and Resource Reservation Protocol (RSVP)

Reducing tail latency for multi-bottleneck in datacenter networks: A compound approach

Currently, network applications are experiencing explosive growth, and various types of network applications are showing a trend of varied demands for quality of network service. However, the existing Explicit Congestion Notification (ECN) marking methods have not taken into account the diversified Quality of Service (QoS) requirements of network applications. This article introduces a multi-queue ECN marking strategy targeting multiple QoS guarantees. The strategy utilizes virtual queues and dynamic weighted round-robin scheduling to achieve traffic partitioning in a programmable data plane. It constructs a multi-queue, multi-class QoS queuing model based on the QoS requirements of different traffic and network conditions. The model is solved by real-time to obtain the ECN marking thresholds and round-robin weights for different queues, in order to achieve dynamic QoS requirements of different network applications. We implemented this strategy in Mininet and BMv2, and compared it with DCQCN, P4QCN, and TCN. The experimental results indicate that this policy demonstrates good performance in terms of queue length, RTT, and throughput, while also ensuring fairness between traffics. Results of the experiment indicate that the proposed approach is superior to DCQCN and P4QCN in the field of performance fluctuation and rapid feedback, and it exhibits notable advantages over TCN, and also ensures the fairness of traffic.

QUIC is a transport protocol that adds congestion control, retransmission control, and TLS to UDP. QUIC can use the same congestion control algorithms as TCP. In previous work, we have shown that communication performance becomes unfair concerning the buffer size of the shared bottleneck link of two flows, one using CUBIC and the other BBR congestion control algorithms within QUIC. In this study, we improve the communication fairness performance at different bottleneck link buffer sizes by using Round Trip Time (RTT) and Explicit Congestion Notification (ECN) to regulate congestion control aggressiveness when CUBIC and BBR compete within QUIC through actual experiments.

With the development of multi-queue data centers, various efforts have leveraged Explicit Congestion Notification (ECN) to achieve high throughput and low latency for data communication. However, one of the deep-seated problems is that the micro-burst traffic in networks could cause the instantaneous queue length to exceed the ECN threshold, leading to significant mismarkings of ECN. Such mismarkings could further lead to severe network performance degradation. In this paper, we propose Micro-Burst aware ECN (MBECN+) to mitigate this issue. MBECN+ is essentially a novel scheme that decouples ECN threshold setting and ECN marking. First, it finds a more appropriate ECN threshold on a per-queue basis to eliminate the spurious congestion signals caused by micro burst traffic. Second, MBECN+ consists of a double-threshold ECN marking scheme that it marks the packets with ECN in a finer grain. Considering the queue backlog caused by micro-burst traffic, it marks the packets when they are dequeuing, instead of enqueuing. Furthermore, we analyze the feasibility of implementing MBECN+ on commodity layer-3 multi-queue switches. Through testbed experiments and large scale simulations, we demonstrate that MBECN+ can improve the throughput by up to <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"><tex-math notation="LaTeX">$\sim$</tex-math></inline-formula> 20% and reduce FCT (flow completion time) by up to <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"><tex-math notation="LaTeX">$\sim$</tex-math></inline-formula> 40%. The throughput under MBECN+ improves by 1.5 <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"><tex-math notation="LaTeX">$\sim 2.4\times$</tex-math></inline-formula> than DCTCP and 1.26 <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"><tex-math notation="LaTeX">$\sim 1.35\times$</tex-math></inline-formula> than ECN*.

DCCS: A dual congestion control signals based TCP for datacenter networks

Most of current Data Center Network (DCN) protocols leverage Explicit Congestion Notification (ECN) for congestion control to maintain low latency and high throughput. However, the majority of them only consider single-queue scenario in each switch port, making their performance inferior in multiple-service multiple-queue scenario. To this end, we propose DemePro, a DCN scheme for multiple-service multiple-queue scenario. In face of congestion signal, DemePro also leverages ECN for congestion notification, while decouples packet marking from enqueuing, in order to ensure fairness among multiple services. We also question the effectiveness of the congestion signal derived from the single threshold for port buffer queue length, via a set of experiments. Then, DemePro utilizes multiple thresholds for proactive congestion control, and packets are encapsulated to carry congestion extent information, in order to notify TCP senders for precise and fine-grained congestion control. Experiments show that DemePro has better performance than MQ-ECN [1] which is currently the best DCN scheme for multiple-service multiple-queue scenario, and DemePro can also well guarantee the fairness.

Data Center Networks (DCNs) suffer from synchronized bursts for segmentation and aggregation modes, leading to buffer overflows at switches and increasing network delay. To overcome this problem, some congestion control algorithms like DCTCP use Explicit Congestion Notification (ECN) to notify in-network congestion and reduce switch buffer occupancy. However, the traditional Additive Increase Multiplicative Decrease (AIMD) method causes high fluctuation of round-trip time (RTT) in DCNs. Some intelligent congestion control algorithms designed for Internet can achieve great flexibility, but are not applicable in DCNs for a lack of accurate congestion feedback. In this paper, we analyze the deficiencies of utilizing RTT as congestion signals and the applicability of learning algorithms in DCNs. Then, we propose DECC, a smart TCP congestion control algorithm for DCNs, which combines Deep Reinforcement Learning (DRL) with ECN to achieve high bandwidth utilization as well as low queuing delay. DECC fully utilizes precise in-network feedback and formulates several QoS requirements to a multi-objective function. Meanwhile, it decouples cwnd adjustment with DRL decision making to gradually learn the optimal congestion control policy in real-time. We evaluate the performance of DECC in various scenarios. Simulation results show that DECC can reduce the queue length at bottleneck switches by more than 50% compared to DCTCP, while maintaining high bandwidth utilization and reducing Flow Completion Time (FCTs) under burst traffic.

RLECN—A learning based dynamic threshold control of ECN

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.

This paper presents a new TCP protocol called TCPPuerto-Londero, which makes dynamic tuning in the congestion window (cwnd) by means of adaptive control theory aiming to keep stable and small the queue length in the bottleneck link located on the path from source to destination. This adaptive control loop has the relative delay in the forward path measured as an input and the cwnd value as an output. Thus, unlike classic TCP protocols, TCP-Puerto-Londero does not use Round Trip Time (RTT) information. Also, unlike classic Active Queue Management (AQM) strategies and Explicit Congestion Notification (ECN) based protocols, TCP-Puerto-Londero employs end-to-end queue management without the need for ECN resources. Moreover, TCP-Puerto-Londero also aims to attend to the new challenges of the Industrial Internet of Things (IIoT). Its performance has been tested in a Dumbbell network topology shared by TCP and UDP-like Networked Control Systems (NCS) flows. Therefore, TCP-Puerto-Londero performance has been compared with ECN-based protocols like DCTCP, E-DCTCP, TCP-Jersey, and ENCN. Furthermore, this paper employs an approach for modeling, analysis, simulation, and verification of the communication network and NCS employing UPPAAL simulation software tool, where all network constituents (transmitters, channels, routers, receivers, controllers, and plants) were modeled employing timed automata, simplifying a formal verification of the complete studied system. Simulations and statistical verification indicate that even though utilizing fewer resources (as it does not require the AQM/ECN information) TCP-Puerto-Londero overcomes TCP-Jersey, DCTCP, and EDCTCP with regard to throughput and fairness for TCP flows. Also, TCP-Puerto-Londero flows are capable to keep the queue length more stable and smaller than other protocols that were compared, and consequently, it reduces the impact in NCS-UDPlike flows sharing the same network, whose performance was measured employing the Integral Time Absolute Error (ITAE), which is a desirable feature for Industrial IoT.

Switches in cloud data centers support multiple service queues per port to provide differentiated network performance among different traffic classes. To isolate service queues, recent solutions leverage the power of Explicit Congestion Notification (ECN). However, this causes a fundamental dependency on ECN-based transport protocols, making it hard to use generic transport protocols. To this end, we design DynaQ, a protocol-independent multi-queue management solution that enables service queue isolation with generic transport protocols. The key idea of DynaQ is to adjust the packet dropping threshold of service queues dynamically. Specifically, DynaQ allows a service queue to occupy free buffer space but prevents the queue from hurting other active queues. Our solution requires only a few additional clock cycles to implement on hardware. To evaluate DynaQ comprehensively, we conduct a series of testbed experiments and large-scale simulations. Our evaluation results show that, compared to alternative schemes, DynaQ is the only solution that achieves work-conserving weighted fair sharing and low latency without protocol dependency.

Traffic congestions cause many environmental, economic and health issues. If we are unable to completely get rid of them, the least we shall try to do is to move them outside of residential areas. In this paper, a novel signal coordination method is proposed, which aims to mitigate traffic congestions. The proposed algorithm is based on the explicit congestion notification protocol, which is well-known from the domain of computer networking. Our method was tested under Eclipse SUMO. Results show that the proposed algorithm successfully limits the traffic density and the traffic flow to a certain level.

Google proposed a new congestion control algorithm (CCA) based on bottleneck bandwidth and round-trip propagation time (BBR), which is considered to open a new era of congestion control. BBR creates a network path model by measuring the available bottleneck bandwidth and the minimum round-trip time (RTT) to maximize delivery rate and minimize latency. The BBR v2 algorithm is a recently updated version by Google, which aims to improve some of the problems in the original BBR (BBRv1) algorithm, such as interprotocol fairness issues, RTT fairness issues, and excessive retransmissions. The BBRv2 evaluation results show that it can improve the coexistence with the loss_based algorithm and alleviate some of the shortcomings in BBRv1. However, when multiple BBRv2 flows enter the same link at different times, fair convergence cannot be achieved, and RTT fairness still exists. Based on these problems, we analyze the root cause and proposed an improved algorithm BBRv2+, which uses flow-aware explicit congestion notification (ECN) to quantify queue information and feedback on the accurate congestion degree. BBRv2+ algorithm can avoid blind window constraints and selectively mark packets so that different flows can converge to fairness. In the simulation experiment of Network Simulator 3 (NS3), the results show that the BBRv2+ algorithm can improve intraprotocol fairness and RTT fairness and ensure bandwidth utilization and interprotocol fairness.

Harmonia: Explicit Congestion Notification and Credit-Reservation Transport Converged Congestion Control in Datacenters

The basic philosophy behind RED is to prevent congestion. When the average queue length exceeds the minimum threshold, packets are randomly dropped, or the explicit congestion notification bit is marked. Since network requirements differ significantly, it is not an optimal approach to establish RED parameters with constant value. There is a new algorithm we are proposing called Critical Point on Target Queue (AQM-RED-CPTQ), provide greater congestion management over the network while also preserving the value of RED. To overcome the problem in RED without changing queue weight parameter, we have proposed few models to control the congestion by introducing range parameter with probability and control mechanism which will belong between minimum and maximum threshold. The current queue size is controlled together with average queue size. A new range variable has been introduced to improve the performance of priority queue of existing RED based algorithm which improves the overall performance of networks. For each packet, minimum and maximum threshold has been updated and dropped with probability (Pa) for a special condition. Instead of multiplicative increase and decrease the maximum probability, the scheme uses additive-increase and multiplicative-decrease. Once the AVG queue length is close to the minimum threshold value, our approach automatically sets queue parameter according to queue conditions and handles queuing delay and improve throughput. The simulated results proof that our approaches are better than RED in terms of throughput, end to end delay, packet delivery ratio and goodput.

Choose a correct marking position: ECN should be freed from tail mark

The past several decades have witnessed wide deployment of space information networks (SINs). More than 8000 satellites have been launched into Earth's orbit recently. Meanwhile, the space communications requirements have grown by a factor of 10 in the last decade. The sharply increasing traffic volume puts tremendous pressure on the network traffic control in SINs. Current end-host-based schemes have to collect a massive amount of data (e.g., explicit congestion notification) to respond to one network event (e.g., topology change, network congestion), which is unacceptable in SINs due to the large end-to-end delay. Recently, with the emergence of programmable network devices, it has become possible to perform flexible control strategies as the traffic is flowing through the network (i.e., in-network traffic control). By implementing the control directly inside the network, the in-network schemes are more effective at scale and more responsive to network dynamics. Therefore, in this article, we design an in-network traffic control powered SIN architecture to enhance the network performance. In addition, we present two use cases, in-network load balance and in-network congestion control, to demonstrate the feasibility of our architecture. Extensive simulations are performed to evaluate our proposed schemes.