Model-based reinforcement learning for router port queue configurations

Kattepur, Ajay; David, Sushanth; Mohalik, Swarup Kumar

doi:10.23919/icn.2021.0016

Cited by 7 publications

(2 citation statements)

References 20 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…According to Kattepur et al [8] Virtual local area network (VLANs) is the process of partitioning a single physical LAN into smaller logical LANs, VLAN reduces the broadcast domain, improve security and performance and are ideal for separating and managing large networks. A VLAN is a switched network that is logically segmented by function, project team, or application, without regard to the physical locations of the users.…”

Section: Introductionmentioning

confidence: 99%

Technical Supplement for Network Control

Layefa

Richard²,

Egbure³

et al. 2023

CJAST

View full text Add to dashboard Cite

Networking in computer science is the connection of two or more computers electronically with the sole aim of data or peripheral sharing. As the computers and peripherals connected to a network increases, using routers and switches alone is no longer reliable to contain broadcast traffic. Therefore this research employed VLAN technology to improve the performance of an extended star network topology by dividing a single switching network into a number of overlaying virtual networks called VLANs that can accommodate various functionalities, security needs and easy administration. For Inter-VLAN routing, a layer 3 switch that support 802.1Q trunking on a fast Ethernet or GigabitEthernet interface was used. This study is conducted using a white-box approach and the waterfall methodology to design, simulate and compare. After configuration, it was seen that the presence of failure in the main network is abstracted from the network users as the VLANS is configured logically to access and share resource with systems on all blocks. Subsequently, A ring topology can be introduced to the system were each block will have a backup switch and can serve as a backup for the block nearest to it. Also a Dual ISP can be introduced to provide steady internet access.

show abstract

Section: Introductionmentioning

confidence: 99%

Technical Supplement for Network Control

Layefa

Richard²,

Egbure³

et al. 2023

CJAST

View full text Add to dashboard Cite

show abstract

“…As analyses have shown, the main technological means of ensuring the differentiated quality of service are congestion management mechanisms, which usually include FIFO, PQ, CQ, FQ/WFQ, CBQ, LLQ, and their numerical modifications and combinations [1][2][3][4][8][9][10][11][12][13]. At the moment, the perfect mechanism does not exist.…”

Section: Introductionmentioning

confidence: 99%

Hierarchical Queue Management Priority and Balancing Based Method under the Interaction Prediction Principle

et al. 2023

View full text Add to dashboard Cite

This work is devoted to improving a two-level hierarchical queue management method based on priority and balancing under the interaction prediction principle. The lower level of calculations was connected with the problem optimization solution and was responsible for two tasks. Firstly, the packet flow aggregation and distribution among the macro-queues and sub-queues organized on the router interface must solve the congestion management problem. Secondly, the resource allocation problem solution was related to the balanced allocation of interface bandwidth among the sub-queues, which were weighted relative to their priorities under the traffic-engineering queues. The method’s lower-level functions were recommended to be placed on a set of processors of a routing device responsible for servicing the packets of individual macro-queues. At the same time, the processor coordinator could perform the functions of the upper-level calculations, providing interface bandwidth allocation among the macro-queues. The numerical research results of the proposed two-level hierarchical queue management method confirmed its effectiveness in ensuring high scalability. Balanced, priority-based packet flow distribution and interface bandwidth allocation among the macro-queues and sub-queues were implemented. In addition, the time was reduced for solving tasks related to queue management. The method demonstrated high convergence of the coordination procedure and the quality of the centralized calculations. The proposed approach can be used in various embedded systems.

show abstract