High Network Load Research Articles

APPROACH TO DETECTION OF ANOMALOUS NETWORK TRAFFIC USING LOF AND HBOS ALGORITHMS

The article is devoted to the problem of detecting anomalies in modern computer networks, which is one of the main threats to cyber security. With the development of Internet technologies, the number of devices and the volume of network traffic are constantly increasing, which leads to an increase in the risk of various cyber threats, such as DDoS attacks, zero-day attacks, and exploitation of protocol vulnerabilities. Abnormal network traffic can result from malicious activity and technical malfunctions, such as configuration errors or hardware failures. Specialised algorithms and methods of analysing large volumes of data are used to detect such threats. The paper considers the main methods of detecting anomalies in network traffic, including classical approaches and modern deep and machine learning methods. Special attention is paid to the efficiency of using methods based on convolutional neural networks, long-term memory and their combinations to detect anomalies. An analysis of the disadvantages and advantages of various approaches to detecting anomalous traffic, such as high computational requirements and the complexity of setting up models, is performed. Still, their effectiveness in analysing large volumes of data is noted. One of the main methods used for anomaly analysis is the local outlier algorithm, which compares the density of objects with their neighbours, allowing for the detection of anomalies in regional segments of the data. Another method is histogram-based outlier estimation, which is faster and more efficient using one-dimensional histograms for each variable. The work also explores the application of unsupervised machine learning methods, which allows for analysing network traffic in real time without the need for prior labelling of data. The article also considers the prospects of further testing the proposed methods in real networks. The combined use of LOF and HBOS balances anomaly detection accuracy and data processing speed, essential to ensure continuous system operation in high-load networks. The implementation of similar solutions in actual conditions requires further research, particularly regarding optimising the use of computing resources and adapting methods to the specific conditions of the network environment. Thus, the paper presents a thorough analysis of modern approaches to detecting anomalies in network traffic and substantiates the feasibility of their application in actual conditions to increase the effectiveness of cyber security.

Smart-KG: Partition-Based Linked Data Fragments for querying knowledge graphs

RDF and SPARQL provide a uniform way to publish and query billions of triples in open knowledge graphs (KGs) on the Web. Yet, provisioning of a fast, reliable, and responsive live querying solution for open KGs is still hardly possible through SPARQL endpoints alone: while such endpoints provide a remarkable performance for single queries, they typically can not cope with highly concurrent query workloads by multiple clients. To mitigate this, the Linked Data Fragments (LDF) framework sparked the design of different alternative low-cost interfaces such as Triple Pattern Fragments (TPF), that partially offload the query processing workload to the client side. On the downside, such interfaces still come with the expense of unnecessarily high network load due to the necessary transfer of intermediate results to the client, leading to query performance degradation compared with endpoints. To address this problem, in the present work, we investigate alternative interfaces, refining and extending the original TPF idea, which also aims at reducing server-resource consumption, by shipping query-relevant partitions of KGs from the server to the client. To this end, first, we align formal definitions and notations of the original LDF framework to uniformly present existing LDF implements and such “partition-based” LDF approaches. These novel LDF interfaces retrieve, instead of the exact triples matching a particular query pattern, a subset of pre-materialized, compressed, partitions of the original graph, containing all answers to a query pattern, to be further evaluated on the client side. As a concrete representative of partition-based LDF, we present smart-KG+, extending and refining our prior work (In WWW ’20: The Web Conference 2020 (2020) 984–994 ACM / IW3C2) in several respects. Our proposed approach is a step forward towards a better-balanced share of the query processing load between clients and servers by shipping graph partitions driven by the structure of RDF graphs to group entities described with the same sets of properties and classes, resulting in significant data transfer reduction. Our experiments demonstrate that the smart-KG+ significantly outperforms existing Web SPARQL interfaces on both pre-existing benchmarks for highly concurrent query execution as well as an accustomed query workload inspired by query logs of existing SPARQL endpoints.

Energy-efficient BS sleeping, user association, and resource allocation in full-duplex ultra-dense networks

Full-duplex ultra-dense network (FD-UDN) is a promising technology in 5G mobile networks for handling the increase in network capacity. However, interference management and energy efficiency are two of the most important challenges that must be addressed. With the small base stations (SBSs) densification, employing sleeping strategy coupled with resource management becomes an effective approach to manage interference and power consumption in FD-UDN. To the best of our knowledge, this aspect has not been addressed in previous work. To this end, we develop a framework to optimize the BS sleeping and resource management with the aim of maximizing energy efficiency and maintaining the quality of service (QoS) requirements of users. The problem is formulated as a non-convex mixed-integer non-linear programming problem, which is difficult to handle. Employing the Dinkelbach method, the objective function of the problem is converted to an equivalent parametric subtractive form. Then, the problem is decoupled into two sub-problems: user association and resource allocation, as well as BSs on/off switching. The former is solved using the iterative reweighted lq-norm minimization (IRM) method, and the latter is solved using the Lagrangian dual method and the constrained concave-convex procedure (CCCP). The simulation results demonstrate that the proposed method is more effective than traditional ones in simultaneously improving the EE, reducing power consumption, and keeping fewer SBSs active, especially in high network loads.

Low consumption automatic discovery protocol for DDS-based large-scale distributed parallel computing

DDS (Data Distribution Service) is an efficient communication specification for distributed parallel computing. However, as the scale of computation expands, high network load and memory consumption consistently limit its performance. This paper proposes a low consumption automatic discovery protocol to improve DDS in large-scale distributed parallel computing. Firstly, an improved Bloom Filter called TBF (Threshold Bloom Filter) is presented to compress the data topic. Then it is combined with the SDP(Simple Discovery Protocol) to reduce the consumption of the automatic discovery process in DDS. On this basis, data publication and subscription between the distributed computing nodes are matched using binarization threshold θ and decision threshold T , which can be obtained through iterative optimization algorithms. Experiment results show that the SDPTBF can guarantee higher transmission accuracy while reducing network load and memory consumption, and therefore improve the performance of DDS-based large-scale distributed parallel computing.

ReTSN: Resilient and Efficient Time-Sensitive Network for Automotive In-Vehicle Communication

Time-sensitive networking (TSN) is being widely investigated to provide Ethernet capabilities for in-vehicle backbone communication. However, the gate control list (GCL), as a simple mechanism for achieving timing determinism for safety-critical traffic (ST) frames with hard deadlines, is too rigid to handle the intrinsic timing uncertainty of automated driving systems (ADSs). Due to the complexity and the unpredictable operating environment, there can be delayed ST frames that can disrupt the solidly fixed timing behavior. In this work, we present one novel approach to effectively use bandwidth resources to deal with delayed ST frames, which cannot be handled by a traditional fixed GCL, and the discarding of them should be reduced to improve the system’s integrity. An acceptance test is implemented to report to the application layer when an ST frame will miss its deadline and hence be rejected, i.e., prevented from entering the switch. To further improve the efficiency of bandwidth usage, we investigate how to improve the performance of the more important Class A frames in an audio-video-bridging (AVB) switch, which adopts a credit-based shaper mechanism, and we propose a constant bandwidth server to replace the credit-based shaper while taking fairness into consideration. Evaluation with extensive experiments shows that both resilience and efficiency of the TSN are significantly enhanced compared with a credit-based shaper, especially when the traffic load is relatively high. For the delayed ST frames and event-triggered traffic, our approach is able to schedule more than the solution using the AVB switch even with a high network load.

Auxiliary graph-based energy-efficient routing resource allocation algorithm in space division multiplexing elastic optical networks

Space division multiplexing elastic optical networks (SDM-EONs) are considered as a efficient solution for increasing the link’s capacity. However, the SDM-EONs brought challenges to be explored, notably the routing, core and spectrum assignment (RCSA) problem. For the real network, another important issue should be considered is the problem of energy consumption. The auxiliary graph (AG) can abstract node devices into elements in it and support various routing policies for different purposes by adjusting the weights of AG edges. Therefore, this paper proposes an auxiliary graph-based energy efficient routing resource allocation algorithm (AG-EERRA) to jointly optimize the energy consumption and RCSA. Each node of the auxiliary graph is abstracted from the devices of physical node. The intra edge weights of nodes set the function associated with the energy consumption model. The inter edge weights of node are composed link length, free frequency slots, modulation efficiency and the value of core crosstalk innovatively. Meanwhile, the impact of inter-core crosstalk on request is reduced by crosstalk avoidance strategy and crosstalk aware strategy. In the proposed auxiliary graph, the shortest weight path is selected to transmit requests that can reduce network energy consumption and blocking probability. Under the high-load network, request splitting is performed for the requests that cannot successfully find a route to improve energy efficiency. Simulation results show that the proposed algorithm can improve the network energy efficiency while reducing request blocking probability compared to comparison algorithms.

Analysis of the impact of radio interference on the quality of service in 5G networks and the possibility of their minimization

This article is devoted to the analysis of the impact of radio interference on fifth-generation mobile communication networks and options for minimizing their impact. 5G networks are considered high-speed and high-load networks, which is why stability and performance are critical for effective service delivery in next-generation mobile networks, which in turn is achieved by combating or minimizing the effects of radio interference on the network. The article presents an overview of modern research and technologies aimed at detecting and eliminating radio interference in fifth-generation mobile communication networks. The article also considers the prospects and future directions of research in this field and is a valuable source of information for professionals in the field of wireless communications.

The Implementation of a Hybrid Router and Dynamic Switching Algorithm on a Multi-FPGA System

The multi-FPGA system known as, the Flow-in-Cloud (FiC) system, is composed of mid-range FPGAs that are directly interconnected by high-speed serial links. FiC is currently being developed as a server for multi-access edge computing (MEC), which is one of the core technologies of 5G. Because the applications of MEC are sometimes timing-critical, a static time division multiplexing (STDM) network has been used on FiC. However, the STDM network exhibits the disadvantage of decreasing link utilization, especially under light traffic. To solve this problem, we propose a hybrid router that combines packet switching for low-priority communication and STDM for high-priority communication. In our hybrid network, the packet switching uses slots that are unused by the STDM; therefore, best-effort communication by packet switching and QoS guarantee communication by the STDM can be used simultaneously. Furthermore, to improve each link utilization under a low network traffic load, we propose a dynamic communication switching algorithm. In our algorithm, each router monitors the network load metrics, and according to the metrics, timing-critical tasks select the STDM according to the metrics only when congestion occurs. This can achieve both QoS guarantee and efficient utilization of each link with a small resource overhead. In our evaluation, the dynamic algorithm was up to 24.6% faster on the execution time with a high network load compared to the packet switching on a real multi-FPGA system with 24 boards.

Research on Generalized Intelligent Routing Technology Based on Graph Neural Network

Aiming at the problems of poor load balancing ability and weak generalization of the existing routing algorithms, this paper proposes an intelligent routing algorithm, GNN-DRL, in the Software Defined Networking (SDN) environment. The GNN-DRL algorithm uses a graph neural network (GNN) to perceive the dynamically changing network topology, generalizes the state of nodes and edges, and combines the self-learning ability of Deep Reinforcement Learning (DRL) to find the optimal routing strategy, which makes GNN-DRL minimize the maximum link utilization and reduces average end-to-end delay under high network load. In this paper, the GNN-DRL intelligent routing algorithm is compared with the Open Shortest Path First (OSPF), Equal-Cost Multi-Path (ECMP), and intelligence-driven experiential network architecture for automatic routing (EARS). The experimental results show that GNN-DRL reduces the maximum link utilization by 13.92% and end-to-end delay by 9.48% compared with the superior intelligent routing algorithm EARS under high traffic load, and can be effectively extended to different network topologies, making possible better load balancing capability and generalizability.

Experimental demonstration of hidden node problem in visible light communication networks

In this paper, hardware investigations show the effects of the hidden and exposed node problems in visible light communication (VLC) networks. Furthermore, the request to send/clear to send (RTS/CTS) mechanism, as a physical layer-independent solution, is used for solving the hidden node problem in VLC networks. A VLC hardware system, called VLCIoT, which was designed and implemented in our laboratory based on PHY I of the IEEE 802.15.7 standard, is used for the physical layer. The medium access control layer of the IEEE 802.15.7 is considered for the network. We implement the multiple access protocol of the IEEE 802.15.7 and the RTS/CTS mechanism on the microcontroller of each device using the C programming language. Goodput, message loss ratio, fairness, average delay, energy efficiency, network load, frame size, number of hidden nodes, and number of contending nodes are the evaluated parameters. The results show that for the IEEE 802.15.7 increase of the hidden nodes decreases the goodput and energy efficiency and increases the message loss ratio and the average delay. But, using the RTS/CTS mechanism, the hidden nodes do not affect the goodput, the message loss ratio, the energy efficiency, and the average delay because this mechanism solves the hidden node problem. The results show that the RTS/CTS mechanism increases the saturation goodput by 300%, decreases the message loss ratio by 94%, decreases the average delay by 50%, and increases the energy efficiency by 300%. The results show that at higher network load, higher data frame size, and more contending nodes, the performance of the RTS/CTS mechanism is better. Also, in symmetric networks, in which all nodes are hidden, or no hidden node exists in the network, fairness is better than the asymmetric networks, in which some nodes are hidden while others are not.

Enhanced request adaptive scheme for information-centric vehicular networks

In this paper, we propose an enhanced request adaptive (EReA) scheme for content service provision in dynamic information-centric vehicular networking (ICVN) environment. The EReA scheme manages the cache update at the roadside units (RSUs) and provides low-latency content delivery to the end-users. User requests trigger the cache update at the RSU for a particular content item on the basis of its update−probability. We introduce an additional parameter tolerancei for each content item i. This parameter suppresses the frequency of content updates triggers at the RSU cache during high network load for the non-critical contents. This provides an additional saving in bandwidth. The simulation setup is created by ndnSIM 2.0 to validate the performance of this proposed scheme. The working and performance of the other content delivery schemes are also discussed in details. Results show that the proposed scheme provides a better hit-ratio compared to the existing ReA scheme under high network load, which results in an additional saving in bandwidth. Additionally, it also inherits the advantages of the ReA scheme.

Determination of optimized sleep interval for 10 gigabit-passive optical network using learning intelligence

<span>The overall aim of this project is to investigate the application of a machine learning method in finding the optimized length of asleep time interval (T<sub>AS</sub>) in a cyclic sleep mechanism (CSM). Since past decade, the implementations of CSM in the optical network unit (ONU) to reduce the energy consumption in 10 gigabit-passive optical network (XG-PON) were extensively researched. However, the newest era sees the emergence of various network traffic with stringent demands that require further improvements on the T<sub>AS</sub> selection. Since conventional methods utilize complex algorithm, this paper presents the employment of an artificial neural network (ANN) to facilitate ONU to determine the optimized T<sub>AS</sub> values using learning from past experiences. Prior to simulation, theoretical analysis was done using the M/G/1 queueing system. The ANN was than trained and tested for the XG-PON network for optimal T<sub>AS</sub> decisions. Results have shown that towards higher network load, a decreasing T<sub>AS</sub> trend was observed from both methods. A wider T<sub>AS</sub> range was recorded from the ANN network as compared to the theoretical values. Therefore, these findings will benefit the network operators to have a flexibility measure in determining the optimal T<sub>AS</sub> values at current network conditions.</span>

On the benefits of point-to-multipoint coherent optics for multilayer capacity planning in ring networks with varying traffic profiles

Point-to-multipoint (P2MP) coherent optics using digital subcarrier multiplexing have recently been proposed as a promising new technology to reduce the cost and complexity of optical transport networks, particularly those in metro aggregation scenarios with hub-and-spoke (H&S) traffic patterns. In this paper, we perform a case study on fault tolerant ring networks with more general traffic profiles to assess the benefits of P2MP optics for alternative network scenarios such as those appearing in metro rings. We propose mixed integer linear programming formulations for multilayer capacity planning to minimize total transceiver costs, considering both point-to-point and P2MP transceivers, as well as a heuristic approach for the more complex P2MP case. Results indicate that P2MP optics can bring several advantages for all tested traffic profiles, ranging from pure uniform flows to mostly H&S. Transceiver cost savings are achieved in various cases, greater at higher network loads and for traffic patterns that are mostly H&S. Results also show that P2MP optics prompt a multilayer redesign of the network with significantly reduced spectrum usage and a reduced number of IP hops, resulting in lower optical–electronic–optical and IP processing latency and lower IP layer equipment costs for all traffic profiles.

5G multi-tier radio access network planning based on voronoi diagram

Network planning of multi-layer heterogeneous mobile networks with complex topology is an important task. In this paper the spectral and energy efficiency of integrated LTE/Wi-Fi technologies for 5G are improved. A method of adaptive formation of the structure of radio access level (RAL) with the provision of the required quality of service (QoS) and the possibility of broadband data transmission is proposed. The Voronoi tessellation is used for designing the RAL of 5G mobile networks for the placement of base stations, which allowed optimum delimiting the coverage area for each base station and provide users the cell interface services. To minimize interference there was proposed a method of dynamic frequency reuse for different sizes of Voronoi cells. Modelling shows the developed method is effective both at low and at high network load, however, at high load there is a slightly smaller gain in energy efficiency than at low load.

ALBLP: Adaptive Load-Balancing Architecture Based on Link-State Prediction in Software-Defined Networking

Load-balancing optimization in software-defined networking (SDN) has been researched for a long time. Researchers have proposed many solutions to the load-balancing problem but have rarely considered the impact of transmission delay between controllers and switches under high-load network conditions. In this paper, we propose an adaptive load-balancing architecture based on link-state prediction (ALBLP) in SDN that can solve the influence of transmission delay between controllers and switches on network load balancing. ALBLP constructs the prediction model of the network link status, adopts the long-term and short-term memory neural network (LSTM) algorithm to predict the network link-state value, and then uses the predicted value as the Dijkstra weight to calculate the optimal path between network hosts. The proposed architecture can adaptively optimize network load balancing and avoid the empty window period, in which the switch flow table does not exist by actively issuing the flow table. Under the network architecture, we collect the data set of the network link-state by simulating the GÉANT network, and we verify the effectiveness of the proposed algorithm. The experiment results show that the ALBLP proposed in this paper can optimize load balancing in SDN and solve the problem of transmission delay between controllers and switches. It has a maximum load-balancing improvement of 23.7% and 11.7% in comparison with the traditional Open Shortest Path First (OSPF) algorithm and the reinforcement learning method based on Q-Learning.

Wireless Music Playing Buzzer Sensor-Assisted Music Tone Adaptive Control

Aiming at the problem of adaptive change of auxiliary music tones, this paper proposes a MAC protocol with a common music tone listening/sleeping type based on a wireless music buzzer sensor. First of all, the new MAC protocol adopts network-wide synchronization, and all sensor nodes in the entire network use the same scheduling table, so that the entire network nodes enter the music tone listening period and the sleep period at the same time. Secondly, the node adaptively adjusts the duty cycle of the node according to the number of data packets in the sending queue, increases the node’s music tone listening time, reduces the end-to-end delay of data packets, and improves the throughput of the network. Then, the experiment adopts a new backoff strategy to adjust the contention window according to the backoff times and collision times of data packets sent by nodes in the last five working cycles, increase the backoff time of sending data packets under high network load, and reduce the appearance of data packets. We build four simulation experiments on the NS2 simulation platform: unassisted music tone adaptive network, single auxiliary music tone adaptive network, auxiliary music tone adaptive convergence network, and random deployment network, which will be based on the auxiliary music tone adaptive MAC protocol, and IEEE802.11 protocol and SMAC protocol are run in four simulation experiments, respectively, and the performance of the three protocols is analyzed according to the tracking files in the simulation experiment. The analysis results show that the simulation wireless sounding buzzer sensor network is adaptive to different auxiliary music tones and different topologies.

A survey on recent contention-free MAC protocols for static and mobile wireless decentralized networks in IoT

Medium Access Control (MAC) protocols for wireless decentralized networks in IoT have attracted a lot of attention in both academic and industrial fields. They aim to coordinate access among IoT devices to the shared wireless medium. More specifically, contention-free MAC protocols are known to be more efficient than contention-based ones in the case of high traffic load and dense networks. To face the main communication challenges between IoT devices such as collisions and conflicts as well as mobility, it is crucial to design an efficient MAC protocol. To do so, several solutions have been proposed in the literature. Among recent proposed solutions, Machine Learning (ML) algorithms and game theory models were used to revolutionize the communication in IoT networks, especially for devices with high mobility degree. In our survey, we first start by studying the challenges and requirements of communication in wireless networks. Then, we provide a comprehensive survey on recent contention-free MAC protocols existing in the literature. Next, we compare these solutions based on important metrics such as QoS, robustness, fairness etc., and discuss the relationship between MAC protocols, network type and network mobility. Finally, we investigate a future research direction to solve a major problem, which is the network disruption and delay tolerance in IoT wireless mobile networks.

ADVICE: Towards adaptive scheduling for data collection and DDoS detection in SDN

Distributed Denial of Service (DDoS) is one of the most rampant attacks which threaten network security. To overcome DDoS in Software-Defined Networking (SDN), many DDoS detection methods have been presented, among which periodic detection approaches with a specific interval are widely utilized. However, periodic data collection and DDoS detection may result in high network load occupancy between SDN controller and switches, high overhead of SDN controller, and long response time to DDoS attacks. Hence, in order to address those issues above, an ADaptiVe schedulIng for data Collection and DDoS dEtection (ADVICE) mechanism is proposed in this work, to flexibly adjust the data collection and detection interval and decrease the workload of the SDN controller. Instead of detects all flow entries at one period, ADVICE collects the flow statistic information with dynamic intervals and finely-grained initiate DDoS detection for each flow entry. Based on the survival time and credence degree of each flow entry, ADVICE can reduce the network load occupancy and assure the rapid detection of DDoS. Experimental results indicate that ADVICE can effectively minimize the controller's workload and optimize the usage of the limited switch-controller connection bandwidth, shorten the response time of DDoS attacks compared with state-of-the-art methods, and thus protect the network from various DDoS attacks.

A study of human motion in computer vision systems based on a skeletal model

Methods of studying human motion in computer vision systems can be divided into two types. These are analysis in two-dimensional and three-dimensional space. The former uses a single camera image and/ or multiple body sensors. Such an approach leads to a rapid accumulation of error and, consequently, low accuracy of the figure representation. Multiple cameras are usually used in the case of three-dimensional space analysis, while the objects are represented as sets of volumetric elements. Despite the high accuracy of this method, it is associated with high computational complexity and internal network load. The purpose of the paper is to develop a model using a single camera, while approaching three-dimensional space analysis methods in terms of accuracy. In this paper a human figure is represented as a skeleton. The skeleton is described by an acyclic connected graph. The general structure of a human figure is analyzed. Fifteen basic points are selected. Physical and logical connections between them were studied and mathematically described. The velocity and spatial characteristics of the points and connections outline the general dynamics of motion. The study describes a model of human motion and gives the option for model construction on the example of a particular image. The developed algorithm for collection and analysis of information estimates relative locations and velocity characteristics of the graph elements. The model can be used for acquisition of information about the reference dynamics of human movements. In case of detecting major differences between the reference and the reality, the behavior is defined as deviant. Thus, the obtained algorithm can be applied in computer vision systems for detection and analysis of human movements.

Vision-Based Sensing Systems for Autonomous Driving: Centralized or Decentralized?

The perception of the surrounding circumstances is an essential task for fully autonomous driving systems, but its high computational and network loads typically impede a single host machine from taking charge of the systems. Decentralized processing is a candidate to decrease such loads; however, it has not been clear that this approach fulfills the requirements of onboard systems, including low latency and low power consumption. Embedded oriented graphics processing units (GPUs) are attracting great interest because they provide massively parallel computation capacity with lower power consumption compared to traditional GPUs. This study explored the effects of decentralized processing on autonomous driving using embedded oriented GPUs as decentralized units. We implemented a prototype system that off-loaded image-based object detection tasks onto embedded oriented GPUs to clarify the effects of decentralized processing. The results of experimental evaluation demonstrated that decentralized processing and network quantization achieved approximately 27 ms delay between the feeding of an image and the arrival of detection results to the host as well as approximately 7 W power consumption on each GPU and network load degradation in orders of magnitude. Judging from these results, we concluded that decentralized processing could be a promising approach to decrease processing latency, network load, and power consumption toward the deployment of autonomous driving systems.