Using Indirect Routing to Recover from Network Traffic Scheduling Estimation Error

2017 ◽  
Author(s):  
Conglong Li ◽  
Matthew K. Mukerjee ◽  
David G. Andersen ◽  
Srinivasan Seshan ◽  
Michael Kaminsky ◽  
...  
Keyword(s):  
Network Traffic ◽  
Estimation Error ◽  
Traffic Scheduling

scholarly journals A DDoS Attacks Detection Based on Conditional Heteroscedastic Time Series Models

2015 ◽  
Vol 20 (1) ◽  
pp. 23-33 ◽  
Author(s):  
Tomasz Andrysiak ◽  
Łukasz Saganowski ◽  
Mirosław Maszewski ◽  
Piotr Grad
Keyword(s):  
Time Series ◽  
Maximum Likelihood ◽  
Present Article ◽  
Network Traffic ◽  
Statistical Models ◽  
Likelihood Function ◽  
Estimation Error ◽  
Time Series Models ◽  
Ddos Attacks ◽  
Network Infrastructure

Abstract Dynamic development of various systems providing safety and protection to network infrastructure from novel, unknown attacks is currently an intensively explored and developed domain. In the present article there is presented an attempt to redress the problem by variability estimation with the use of conditional variation. The predictions of this variability were based on the estimated conditional heteroscedastic statistical models ARCH, GARCH and FIGARCH. The method used for estimating the parameters of the exploited models was determined by calculating maximum likelihood function. With the use of compromise between conciseness of representation and the size of estimation error there has been selected as a sparingly parameterized form of models. In order to detect an attack-/anomaly in the network traffic there were used differences between the actual network traffic and the estimated model of the traffic. The presented research confirmed efficacy of the described method and cogency of the choice of statistical models.

Virtually Zero Delay Interaction Between Online Game Players Using Kalman Filter-Based Dead Reckoning with Density and Distance Gain Control Adaptation

2019 ◽  
Vol 29 (08) ◽  
pp. 1091-1101
Author(s):  
Jung-Yoon Kim ◽  
Seong-Whan Kim
Keyword(s):  
Kalman Filter ◽  
Network Traffic ◽  
Traffic Congestion ◽  
Prediction Accuracy ◽  
Estimation Error ◽  
Dead Reckoning ◽  
Gain Control ◽  
Estimation Accuracy ◽  
Game Engine ◽  
Network Game

Online 3D games require fast and efficient user interaction support over the network environments, and the networking support is usually implemented by the use of a network game engine. The network game engine should minimize the network delay and mitigate the network traffic congestion. To minimize the network traffic between game users, a client-based prediction (dead reckoning (DR) algorithm) is used. Each game entity uses the algorithm to estimate its own movement as well as the others’. In case the estimation error exceeds the threshold, the entity sends an UPDATE packet which includes velocity, position and the like to other entities. As the estimation accuracy is increased, each entity can minimize the transmission of the UPDATE packet. In this paper, a Kalman filter-based approach is proposed in order to improve the prediction accuracy and an adaptive Kalman gain control in order to minimize the number of UPDATE packets to distant devices. The BZFlag game was used in the experiment in order to verify the proposed approach and the results have shown that it is possible to increase prediction accuracy and reduce the network traffic by 12%.

A Traffic Matrix Recovery Algorithm via Low-Dimension Nature in Smart Grid

2013 ◽  
Vol 392 ◽  
pp. 593-597
Author(s):  
Fan Bo Meng ◽  
Hong Hao Zhao ◽  
Qing Qi Zhao ◽  
Wei Zhe Ma ◽  
Zhi Gang Qi ◽  
...  
Keyword(s):  
Smart Grid ◽  
Network Traffic ◽  
Estimation Error ◽  
Network Tomography ◽  
Recovery Algorithm ◽  
Matrix Recovery ◽  
Traffic Matrix ◽  
Low Dimension ◽  
Simulation Results ◽  
Reconstruction Model

In this literature, we explore the solution of network traffic recovery in smart grid. Taking account of dimensionality of network traffic in smart grid, we propose a novel reconstruction model via network tomography. In our algorithm, we use the low-dimension nature of traffic matrix and the greedy adaptive dictionary algorithm to convert the network tomography into the problem of sparse reconstruction at first. Then we solve network traffic by an iterative greedy algorithm. Simulation results indicate that proposed algorithm exhibits noticeably improvement in estimation error comparing with previous work.

scholarly journals Termite inspired algorithm for traffic engineering in hybrid software defined networks

2020 ◽  
Vol 6 ◽  
pp. e283
Author(s):  
R Ananthalakshmi Ammal ◽  
Sajimon PC ◽  
Vinodchandra SS
Keyword(s):  
Network Traffic ◽  
Traffic Engineering ◽  
Test Bed ◽  
Network Resources ◽  
Traffic Scheduling ◽  
Physical Test ◽  
Multi Protocol Label Switching ◽  
Link Availability ◽  
Sdn Controller

In the era of Internet of Things and 5G networks, handling real time network traffic with the required Quality of Services and optimal utilization of network resources is a challenging task. Traffic Engineering provides mechanisms to guide network traffic to improve utilization of network resources and meet requirements of the network Quality of Service (QoS). Traditional networks use IP based and Multi-Protocol Label Switching (MPLS) based Traffic Engineering mechanisms. Software Defined Networking (SDN) have characteristics useful for solving traffic scheduling and management. Currently the traditional networks are not going to be replaced fully by SDN enabled resources and hence traffic engineering solutions for Hybrid IP/SDN setups have to be explored. In this paper we propose a new Termite Inspired Optimization algorithm for dynamic path allocation and better utilization of network links using hybrid SDN setup. The proposed bioinspired algorithm based on Termite behaviour implemented in the SDN Controller supports elastic bandwidth demands from applications, by avoiding congestion, handling traffic priority and link availability. Testing in both simulated and physical test bed demonstrate the performance of the algorithm with the support of SDN. In cases of link failures, the algorithm in the SDN Controller performs failure recovery gracefully. The algorithm also performs very well in congestion avoidance. The SDN based algorithm can be implemented in the existing traditional WAN as a hybrid setup and is a less complex, better alternative to the traditional MPLS Traffic Engineering setup.

Sign in / Sign up

Export Citation Format

Share Document