UDP Flows Research Articles

Enhancing TCP Airtime Fairness through Precise Computation for Upload and Download Flows in WiFi Networks

Airtime fairness has emerged as a key approach to enhancing wireless throughput performance. However, existing research often overlooks the precise calculation of airtime, particularly in relation to TCP acknowledgments. This paper introduces a novel method, implemented on the access point side, for accurately calculating the airtime of TCP and UDP flows. Building on this, we propose a QoS-based scheduling algorithm designed to improve fairness between upload and download traffic. The effectiveness of the algorithm is validated through experiments that accurately measure both throughput and airtime for upload and download traffic.

Fixed-Time Congestion Control for a Class of Uncertain Multi-Bottleneck TCP/AWM Networks

As network technology continues to advance, network congestion has become an inevitable aspect of network communication. Considering the external interference, unmodeled uncertainty and the interaction between nodes, a multi-bottleneck TCP/AWM network model is established in this paper. A new fixed-time congestion controller was designed by combining a neural network and the backstepping technique. The neural network approximation property is used to eliminate the interference of unmodeled uncertainty and UDP flow in the system. The controller designed in this paper can ensure the stability of the TCP/AWM closed-loop system in a fixed time. Finally, the simulation results demonstrate the effectiveness of the proposed TCP/AWM controller.

A seven-dimensional state flow traffic modelling for multi-controller Software-Defined Networks considering multiple switches

Software-Defined Networking (SDN) improves flexibility in network management and programmability through decoupling the control and the data planes. In SDN, traffic analysis and modelling are important for tracking network activity and availability to detect abnormalities, such as security and operational problems. In this paper, a seven-dimensional state model is used to model SDN TCP and UDP flows, unlike other models that are used in previous work which use three or four-dimensional states. We formulate seven-dimensional states with transitions which consider flow-level arrivals to get more accurate knowledge about modelling the SDN traffic. The proposed work uses multiple controllers and multiple switches with limited capacities unlike other models in related work which may use one controller or switch with an infinite buffer. Using multiple controllers improves the security of the network, due to the backup capability. In addition, it enhances the network’s scalability since there are enough controllers to serve more switches. Also, using multiple switches can provide more accurate results to better reflect the actual case of traffic in the network. The packet average delay and loss probability are tracked to measure the performance of the developed models. Simulation results are formulated and compared to analytical ones. The results prove that the proposed modelling is accurate compared to the recent work.

Automatic Detection of DGA-Enabled Malware Using SDN and Traffic Behavioral Modeling

Enterprise networks are under enormous threats from sophisticated cyber-attacks. Various kinds of malware are installed by attackers on compromised hosts, acting as bots that typically use Domain Generation Algorithms (DGAs) to communicate with their Command and Control (C&C) servers. It is computationally expensive to inspect all network packets of every host connected to a large enterprise network in “real-time” at scale with hundreds of gigabits per second data rates. This paper combines Software Defined Networking (SDN) and machine learning to develop an accurate, cost-effective, and scalable system for detecting infected hosts communicating with external C&C servers, subsequent to the resolution of DGA query names. Our solution dynamically selects network flows for diagnosis by trained models in real-time, and relies more on the behavioral traffic profile, rather than packet content. Our first contribution highlights the prevalence and activity pattern of DGA-enabled malware across internal hosts. We draw insights into the behavioral profile of DGA-enabled malware flows when communicating with C&C servers. For our second contribution, we identify malware traffic attributes and train three specialized one-class classifier models using behavioral attributes of malware HTTP, HTTPS and UDP flows. We develop an SDN-based monitoring system to automatically mirror TCP/UDP flows pertinent to DGA queries for diagnosis by the trained models. Finally, we evaluate the efficacy of our approach by testing suspicious traffic flows (selectively recorded by SDN reactive rules), identifying infected hosts, and verifying our detection with an off-the-shelf Intrusion Detection System (IDS) software tool.

Resilience Enhancement at Edge Cloud Systems

It is becoming common practice to push interactive and location-based services from remote datacenters to resource-constrained edge domains. This trend creates new management challenges at the network edge, not least to ensure resilience. These challenges now need to be investigated and overcome. In this paper, we explore the use of open-source programmable asset orchestration at edge cloud systems to guarantee operational resilience and a satisfactory performance level despite system incidents such as faults, congestion, or cyber-attacks. We discuss the design and deployment of a new cross-level configurable solution, Resilient Edge Cloud Systems (RECS). Results from appropriate tests made on RECS highlight the positive effects of deploying novel service and resource management algorithms at both data and control planes of the programmable edge system to mitigate against disruptive events such as control channel issues, service overload, or link congestion. RECS offers the following benefits: i) the switch automatically selects the standalone operation mode after its disconnection from the upper-level controllers; ii) deployment of edge virtualized services is made, according to client requests; iii) the client requests are served by edge services and the related traffic is balanced among the alternative on-demand routing paths to the edge location where each service is available for its clients; iv) the TCP traffic quality is protected from unfair competitiveness of UDP flows; and v) a set of redundant controllers is orchestrated by a top-level multi-thread cluster manager, using a novel management protocol with low overhead.

Observer‐based finite‐time congestion control for TCP/AQM network system with time‐varying delays and UDP flows

AbstractIn this paper, the problem of observer‐based finite‐time congestion control for transmission control protocol (TCP)/active queue management (AQM) network systems with time‐varying delay and nonresponsive flows such as UDP is studied. The congestion control problem of TCP/AQM network system with time delay and external disturbances observer‐based finite‐time control is solved for the first time, and a theorem proved by the Lyapunov function of TCP/AQM network systems finite‐time bounded is given to ensure that the closed‐loop system is finite‐time bounded. Finally, the control gain and observer gain are obtained by linear matrix inequality (LMI) technique. Simulation results show the effectiveness and superiority of the proposed method.

Diffusion Model of a Non-Integer Order PIγ Controller with TCP/UDP Streams.

In this article, a way to employ the diffusion approximation to model interplay between TCP and UDP flows is presented. In order to control traffic congestion, an environment of IP routers applying AQM (Active Queue Management) algorithms has been introduced. Furthermore, the impact of the fractional controller and its parameters on the transport protocols is investigated. The controller has been elaborated in accordance with the control theory. The TCP and UDP flows are transmitted simultaneously and are mutually independent. Only the TCP is controlled by the AQM algorithm. Our diffusion model allows a single TCP or UDP flow to start or end at any time, which distinguishes it from those previously described in the literature.

Knowledge Discovery: Can It Shed New Light on Threshold Definition for Heavy-Hitter Detection?

Heavy-Hitter (HH) flows are well-known in the field of networking, mainly due to their resource consumption, which is considerably higher than the majority of flows. Their reliable detection and management are critical to optimising network performance. Nevertheless, to date, there is no generally accepted and widely used methodology for HH threshold selection. Indeed, different works use distinct thresholds without the support of a detailed or systematic study. In this paper, we provide useful insights and suggestions on how to determine more justified and valid thresholds. Based on the obtained results, we conclude that no threshold can be used universally to separate flows into HHs and non-HHs. A threshold that performs efficiently in one network may underperform in another. Threshold and HH definitions are often application-dependent, and therefore, threshold selection should include a detailed analysis of the network and its traffic. We also highlight that TCP and UDP flows should be classified with different thresholds because HHs exhibit different characteristics in such protocols. Lastly, we point out that the use of more than one threshold leads to accuracy and efficacy improvements in HHs classification.

A dynamic network traffic classifier using supervised ML for a Docker-based SDN network

With the rapid technological growth in the last decades, the number of devices and users has drastically increased. Software-defined networking (SDN) with machine learning (ML) has become an emerging solution for network scheduling, quality of service (QoS), resource allocations, and security. This paper focuses on the implementation of a network traffic classifier using a novel Docker-based SDN network. ML offers good performance to real-time traffic solutions without depending on well-known TCP or UDP port numbers, IP addresses, or encrypted payloads. In this paper, using three ML techniques, we first classify network flows with 3, 5, and 7 parameters giving up to 97.14% accuracy. Additionally, we present a new performance accelerator algorithm (PAA), which incorporates these three ML classifiers and accelerates the overall performance significantly. We then propose a dynamic network classifier (DNC) generated from PAA over a novel Docker-based SDN network. Finally, we propose a new controller algorithm for Ryu platforms, which integrates the DNC and classifies both TCP and UDP flows in real-time. Based on the evaluations, an improvement in latency performance has been demonstrated, where analysing a packet, controller processing time takes on an average of 10 µs. This study will certainly serve to further research on optimising SDN and QoS.

Анализ влияния параметров алгоритмов Machine Learning на результаты классификации трафика в режиме реального времени

The paper analyzes the impact of setting the parameters of Machine Learning algorithms on the results of traffic classification in realtime. The Random Forest and XGBoost algorithms are considered. A brief description of the work of both methods and methods for evaluating the results of classification is given. Experimental studies are conducted on a database obtained on a real network, separately for TCP and UDP flows. In order for the results of the study to be used in real time, a special feature matrix is created based on the first 15 packets of the flow. The main parameters of the Random Forest (RF) algorithm for configuration are the number of trees, the partition criterion used, the maximum number of features for constructing the partition function, the depth of the tree, and the minimum number of samples in the node and in the leaf. For XGBoost, the number of trees, the depth of the tree, the minimum number of samples in the leaf, for features, and the percentage of samples needed to build the tree are taken. Increasing the number of trees leads to an increase in accuracy to a certain value, but as shown in the article, it is important to make sure that the model is not overfitted. To combat overfitting, the remaining parameters of the trees are used. In the data set under study, by eliminating overfitting, it was possible to achieve an increase in classification accuracy for individual applications by 11-12% for Random Forest and by 12-19% for XGBoost. The results show that setting the parameters is a very important step in building a traffic classification model, because it helps to combat overfitting and significantly increases the accuracy of the algorithm's predictions. In addition, it was shown that if the parameters are properly configured, XGBoost, which is not very popular in traffic classification works, becomes a competitive algorithm and shows better results compared to the widespread Random Forest.

Distributed gateway-based load balancing in software defined network

To achieve an internet with high availability and reliability, needs two or more data paths so the process for sending data can be faster. Load balancing is often plays a significant role for this technique to properly utilized every gateway in the network. This research, implemented load balancing in software defined network architecture using floodlight controller. Evaluation is done by measuring QoS (delay, bit rate, packet rate, packet success rate) while sending various traffics through the network such as UDP Flow, VoIP, and DNS. Performance of load balancer is work well, because the results after load balancing is better than before. Which is the value of delay after load balancing is decreased about 30-55% compared to before load balancing, also the values of bit rate, packet rate dan packet success rate after load balancing is increased about 10-30% compared to before load balancing.

Design and Implementation of TCP-Friendly Meters in P4 Switches

Metering the traffic of a flow and dropping the portion of traffic that exceeds the target rate set by the meter is very important to provide quality of service (QoS) in a network. Most existing switches on the market provide the meter function. Although these meters regulate the traffic of a UDP flow very well, according to our tests, they are not TCP-friendly. For example, we found that a TCP flow passing the meter of a commercial switch can only achieve about 10% of the target rate. Based on our studies, we found that this poor performance is due to the bad interactions between the TCP congestion control and the meter function. In this paper, we design and implement a TCP-friendly meter in the packet processing pipelines of a P4 switch. Experimental results show that our meter regulates a TCP flow very well and can maintain its achieved rate within 5% of the target rate. Compared with the meters in several hardware switches, our meter improves the achieved rate of a TCP flow by almost 85% of the target rate.

Minstrel PIE: Curtailing queue delay in unresponsive traffic environments

Active Queue Management (AQM) algorithms aim to maintain a proper trade-off between queue delay and bottleneck link utilization. However, it is often noticed that this trade-off is not achieved convincingly when unresponsive UDP flows coexist with responsive TCP flows. This paper proposes an extension to Proportional Integral controller Enhanced (PIE) algorithm called Minstrel PIE, which adapts the reference queue delay to improve the trade-off between queue delay and link utilization when unresponsive flows share the same bottleneck queue as responsive flows. Extensive evaluations through simulations and real time experiments demonstrate that Minstrel PIE improves the performance of PIE in the presence of unresponsive flows, and delivers similar performance otherwise. Moreover, the Minstrel PIE algorithm does not introduce new knobs to improve the performance of PIE and hence, can be easily deployed without any additional complexity.

Active Queue Management Algorithm for TCP Networks with Integral Backstepping and Minimax

A novel active queue management (AQM) approach is considered for a class of TCP network systems in this paper. A sufficient condition is given and the corresponding control is obtained based on integral back-stepping technique (IB) and minimax method. The presented results not only are used to deal with the disturbances produced by UDP flows, but also can shorten the convergent time of the signals. Simulation examples are carried out to verify the effectiveness and superiority of the proposed algorithm.

Performance modeling and analysis of TCP and UDP flows over software defined networks

Software Defined Networking (SDN) decouples the control plane from the data plane, thereby facilitating network virtualization, dynamic programmability and flexibility in network management. Previous studies on SDN modeling focus only on packet-level arrivals. However, if flow-level arrivals are not also considered, the model cannot properly reflect the true probability of packets being sent to the controller. Accordingly, the present study proposes two analytical models for predicting the performance of TCP and UDP flows over SDN, respectively, given the assumption of both flow-level arrivals and packet-level arrivals. In constructing the models, the switch and controller are considered jointly and four-dimensional states are used to evaluate the steady-state probabilities of the states. Analytical formulae are derived for the average packet delay and packet loss probability of the TCP and UDP flows. Simulation results very well match with the analytical ones, thereby validating our analytical models. The results show that TCP significantly outperforms UDP over SDN architectures. In particular, TCP reduces the packet delay by 12 ∽ 50% and the packet loss probability by 25 ∽ 100%.

DeepComNet: Performance evaluation of network topologies using graph-based deep learning

Modeling the performance of network protocols and communication networks generally relies on expert knowledge and understanding of the different elements of a network, their configuration, and the overall architecture and topology. Machine learning is often proposed as a tool to help modeling such complex systems. One drawback of this method is that high-level features are generally used – which require full understanding of the network protocols to be chosen, correctly engineered, and measured – and the approaches are generally limited to a given network topology. In this article, we present DeepComNet, an approach to address the challenges of working with machine learning by using lower-level features, namely only a description of the network architecture. Our main contribution is a method for applying deep learning on network topologies via the use of Graph Gated Neural Networks, a specialized recurrent neural network for graphs. Our approach enables us to make performance predictions based only on a graph-based representation of network topologies. To evaluate the potential of DeepComNet, we apply our approach to the tasks of predicting the throughput of TCP flows and the end-to-end latencies of UDP flows. In both cases, the same base model is used. Numerical results show that our approach is able to learn and predict performance properties of TCP and UDP flows with a median absolute relative error smaller than 1%, outperforming related methods from the literature by one order of magnitude.

A Control-Theoretic Approach to In-Network Congestion Management

WANs are often over-provisioned to accommodate worst-case operating conditions, with many links typically running at only around 30% capacity. In this paper, we show that in-network congestion management can play an important role in increasing network utilization. To mitigate the effects of in-network congestion caused by rapid variations in traffic demand, we propose using high-frequency traffic control (HFTraC) algorithms that exchange real-time flow rate and buffer occupancy information between routers to dynamically coordinate their link-service rates. We show that the design of such dynamic link-service rate policies can be cast as a distributed optimal control problem that allows us to systematically explore an enlarged design space of in-network congestion management algorithms. This also provides a means of quantitatively comparing different controller architectures: we show, perhaps surprisingly, that centralized control is not always better. We implement and evaluate HFTraC in the face of rapidly varying UDP and TCP flows and in combination with AQM algorithms. Using a custom experimental testbed, a Mininet emulator, and a production WAN, we show that HFTraC leads to up to 66% decreases in packet loss rates at high link utilizations as compared to FIFO policies.

Minimax based congestion control for TCP network systems with UDP flows

The congestion control problem for TCP network systems with UDP flows is considered. A nonlinear TCP network model with strict-feedback structure is established. The unknown UDP flow is considered as the disturbance to the system, and the maximum UDP flow is calculated by using the minimax approach. And then, a congestion control algorithm is proposed by using backstepping approach. Further, a state-feedback congestion controller is presented to make the TCP networks asymptotically stable. The simulation results show the superiority and feasibility of the proposed method.

Fixed-point analysis of a network of routers with persistent TCP/UDP flows and class-based weighted fair queuing

Fixed-point models have already been successfully used to analytically study networks consisting of persistent TCP flows only, or mixed TCP/UDP flows with a single queue per link and differentiated buffer management for these two types of flows. In the current study, we propose a nested fixed-point analytical method to obtain the throughput of persistent TCP and UDP flows in a network of routers supporting class-based weighted fair queuing allowing the use of separate queues for each class. In particular, we study the case of two classes where one of the classes uses drop-tail queue management and is intended for only UDP traffic. The other class targeting TCP, but also allowing UDP traffic for the purpose of generality, is assumed to employ active queue management. The effectiveness of the proposed analytical method is validated in terms of accuracy using ns-3 simulations and the required computational effort.

Pktgen: Measuring performance on high speed networks

Pktgen is a tool for high-speed packet generation and testing. It runs in the Linux kernel, and is designed to accommodate a wide range of network performance tests. Pktgen consists of a packet generator, a receiver, and a protocol that defines the format for test packets. This paper describes the design of pktgen, and discusses its usage as a capable performance testing tool. The design is focused on performance, and in order to generate packets at high packet rate, pktgen takes advantage of multicore systems and multi-queue features on modern network cards. To support realistic heterogeneous traffic patterns, pktgen can generate multiple simultaneous UDP flows. Pktgen supports measurements of throughput, latency, packet inter-arrival time, and packet delay variation with a wide-range of fine-grained user specified rates.We evaluate and compare pktgen to other tools, and show that it is more than an order of magnitude faster than current Linux applications such as Iperf and Netperf, and has the same upper limit on performance as special-purpose high-speed tools, such as DPDK and Netmap, when it comes to throughput measurements with user-specified rate.