A MapReduce Approach for Traffic Matrix Estimation in SDN

Queiroz, Wander; Capretz, Miriam A. M.; Dantas, Mário A. R.

doi:10.1109/access.2020.3016249

Cited by 8 publications

(6 citation statements)

References 40 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…TM estimation often includes aggregation of large traffic flows which is a difficult task at real time. In [30], authors proposed distributed Map Reduce approach for aggregation of large traffic flows. The Map Reduce approach uses topology information to identify origin-destination links, for easier aggregation of traffic flows.…”

Section: Related Workmentioning

confidence: 99%

Estimation of Traffic Matrix from Links Load using Genetic Algorithm

Pachuau

Krishna

Saha

2021

SCPE

View full text Add to dashboard Cite

Traffic Matrix (TM) is a representation of all traffic flows in a network. It is helpful for traffic engineering and network management. It contains the traffic measurement for all parts of a network and thus for larger network it is difficult to measure precisely. Link load are easily obtainable but they fail to provide a complete TM representation. Also link load and TM relationship forms an under-determined system with infinite set of solutions. One of the well known traffic models Gravity model provides a rough estimation of the TM. We have proposed a Genetic algorithm (GA) based optimization method to further the solutions of the Gravity model. The Gravity model is applied as an initial solution and then GA model is applied taking the link load-TM relationship as a objective function. Results shows improvement over Gravity model.

show abstract

Section: Related Workmentioning

confidence: 99%

Estimation of Traffic Matrix from Links Load using Genetic Algorithm

Pachuau

Krishna

Saha

2021

SCPE

View full text Add to dashboard Cite

show abstract

“…In SDN, the network control is separated from the infrastructure. All logic is transferred to the SDN controller, which is directly programmable [11]. Therefore, the network devices are only responsible for forwarding packets based on the controller's directions [12].…”

Section: Introductionmentioning

confidence: 99%

A Survey of Graph Techniques Applied in Software Defined Networking

2023

IJCCCE

View full text Add to dashboard Cite

Modelling computer networks in general, particularly Software Defined Networking (SDN) as a graph, is beneficial in network planning and design, configuration management, traffic analysis, and security. According to the dynamic nature of SDN, it needs a fast response due to the rapid changes in the network state. The SDN network topology can be modelled as a graph and stored in a graph database, and the traffic load of each switch is stored in the created graph. Consequently, a graph processing framework can be used to process the stored traffic data, and the results are utilized in traffic engineering to assist the SDN controller in network management. This paper provides a comprehensive literature survey involving graph techniques applied to SDN. Then, a summary of graph algorithms is presented. In addition, an overview of graph databases and graph processing frameworks is displayed. Finally, a model is suggested to integrate the graph database and graph processing framework in SDN traffic analysis. Index Terms— Graph algorithms, Graph database, Graph processing frameworks, Network topology modelling, SDN, Traffic Analysis.

show abstract

“…SDN provides two methods for collecting traffic information from the forwarding devices: (1) indirectly using the SDN controller and (2) directly transmitting particular messages to the switches [10]. The common method uses the controller that provides various configurations and APIs besides collecting counter values [8].…”

Section: Introductionmentioning

confidence: 99%

“…In SDN, the network control is separated from the infrastructure. All logic is migrated to the SDN controller, which is directly programmable [8]. Therefore, the network devices are only in charge of forwarding packets according to the controller directions [9].…”

Section: Introductionmentioning

confidence: 99%

BG-SDN: A Big Graph Framework for Topology Modeling and Traffic Analysis in Large-Scale SDN

Samaan

Jeiad

2022

Preprint

View full text Add to dashboard Cite

This paper presented a new method of modeling a Software Defined Networking (SDN) network as a graph and importing traffic data into the created graph using graph techniques to speed up the computations required in traffic analysis and management. For this purpose, a specialized application named TMTA (Topology Modeling and Traffic Analysis) is created to model the SDN network topology as a graph and feed the created graph with traffic data. To achieve high performance, the TMTA application is integrated with the Neo4j graph database utilizing its ability to store and process big graphs consisting of massive amounts of highly interconnected data in a scalable and reliable manner. For performance evaluation, various sizes of network topologies ranging from 41 to 62,586 devices are modeled using the TMTA application and stored in Neo4j. Evaluation results show that the proposed big graph framework can handle large-scale topologies effectively and perform traffic analysis tasks, i.e., calculating the control score of each device in the network and finding the shortest path between devices in a very reasonable time.

show abstract

A MapReduce Approach for Traffic Matrix Estimation in SDN

Cited by 8 publications

References 40 publications

Estimation of Traffic Matrix from Links Load using Genetic Algorithm

Estimation of Traffic Matrix from Links Load using Genetic Algorithm

A Survey of Graph Techniques Applied in Software Defined Networking

BG-SDN: A Big Graph Framework for Topology Modeling and Traffic Analysis in Large-Scale SDN

Contact Info

Product

Resources

About