DQR: An Efficient Deep Q-Based Routing Approach in Multi-Controller Software Defined WAN (SD-WAN)

2020 ◽  
Vol 20 (04) ◽  
pp. 2150002
Author(s):  
MANEL MAJDOUB ◽  
ALI EL KAMEL ◽  
HABIB YOUSSEF
Keyword(s):  
Load Balancing ◽  
Resource Utilization ◽  
Large Scale ◽  
Network Technology ◽  
Network Resources ◽  
Routing Optimization ◽  
Multi Agent ◽  
Large Scale Networks ◽  
Network Scalability ◽  
Better Than

Software Defined Networking (SDN) is a promising paradigm in the field of network technology. This paradigm suggests the separation between the control plane and the data plane which brings flexibility, efficiency and programmability to network resources. SDN deployment in large scale networks raises many issues which can be overcame using a collaborative multi-controller approaches. Such approaches can resolve problems of routing optimization and network scalability. In large scale networks, such as SD-WAN, routing optimization consists of achieving a trade-off between per-flow QoS, the load balancing in each domain as well as the resource utilization in inter-domain links. Multi-Agent Reinforcement Learning paradigm(MARL) is one of the most popular solutions that can be used to optimize routing strategies in SD-WAN. This paper proposes an efficient approach based on MARL which is able to ensure a load balancing among each network as well as optimized resource utilization of inter-domain links. This approach profits from our previous work, denoted SPFLR, and tries to balance the load of the whole network using Deep Q-Networks (DQN) algorithms. Simulation results show that the proposed solution performs better than parallel solutions such as BGP-based routing and random routing.

Routing Metric Based on Node Degree for Load-Balancing in Large-Scale Networks

2011 ◽  
Author(s):  
Hitomi Tamura ◽  
Masato Uchida ◽  
Masato Tsuru ◽  
Jun'ichi Shimada ◽  
Takeshi Ikenaga ◽  
...  
Keyword(s):  
Load Balancing ◽  
Large Scale ◽  
Node Degree ◽  
Routing Metric ◽  
Large Scale Networks

scholarly journals An Improved Rasa for Load Balancing in Cloud Computing

2021 ◽  
pp. 52-66
Author(s):  
Oyekanmi Ezekiel Olufunminiyi ◽  
Oladoja Ilobekemen Perpetual ◽  
Omotehinwa Temidayo Oluwatosin
Keyword(s):  
Load Balancing ◽  
Resource Utilization ◽  
Performance Metrics ◽  
Scheduling Algorithm ◽  
Resource Provisioning ◽  
The Common ◽  
Resource Aware ◽  
Cloud Users ◽  
Better Than

Cloud is specifically known to have difficulty in managing resource usage during task scheduling, this is an innate from distributed computing and virtualization. The common issue in cloud is load balancing management. This issue is more prominent in virtualization technology and it affects cloud providers in term of resource utilization and cost and to the users in term of Quality of Service (QoS). Efficient procedures are therefore necessary to achieve maximum resource utilization at a minimized cost. This study implemented a load balancing scheme called Improved Resource Aware Scheduling Algorithm (I-RASA) for resource provisioning to cloud users on a pay-as-you-go basis using CloudSim 3.0.3 package tool. I-RASA was compared with recent load balancing algorithms and the result shown in performance evaluation section of this paper is better than Max-min and RASA load balancing techniques. However, it sometimes outperforms or on equal balance with Improved Max-Min load balancing technique when using makespan, flow time, throughput, and resource utilization as the performance metrics.

Multiple Path Load Balancing for TCP Large Scale Networks

2019 ◽  
Author(s):  
Konstantinos Paximadis ◽  
Giannis Tzimas ◽  
Anna Galanopoulou ◽  
Pavlos Kalpakioris ◽  
Zlatan Sabic
Keyword(s):  
Load Balancing ◽  
Large Scale ◽  
Multiple Path ◽  
Large Scale Networks

scholarly journals Multi-agent based load balancing in cloud computing

2017 ◽  
Vol 7 (1.1) ◽  
pp. 689
Author(s):  
S Sandeep Kumar ◽  
S Tarun Kumar ◽  
G Sathraja ◽  
J Nagatejasri
Keyword(s):  
Cloud Computing ◽  
Load Balancing ◽  
Resource Utilization ◽  
Data Centers ◽  
Agent Based ◽  
Multiple Servers ◽  
Work Distribution ◽  
Multi Agent ◽  
Area Unit

Cloud knowledge centers area unit usually comprised of multiple servers with doubtless completely different specifications and unsteady resource usages. The challenge for these data centers is how to handle and service the millions of Requests such that the Quality of the ser-vice (Qos) is not compromised. Load balancing is an important aspect in cloud computing that involves an even work distribution among the available machines such that no machine is overloaded. This work discusses on numerous Agent primarily based load equalization tech-niques capableof equalization workloads across multiple servers to provide customer Satisfaction and economical Resource Utilization. We propose a Dynamic Multi-Agent based algorithm to address the load balancing issue .In this, we describe about Agents and how they can be used to solve load balancing in cloud computing.  

scholarly journals NVP: A Network Virtualization Proxy for Software Defined Networking

2016 ◽  
Vol 11 (5) ◽  
pp. 697 ◽  
Author(s):  
Billy Pinheiro ◽  
Eduardo Cerqueira ◽  
Antonio Abelem
Keyword(s):  
Large Scale ◽  
Network Virtualization ◽  
Software Defined Networking ◽  
Network Function Virtualization ◽  
Load Testing ◽  
Proof Of Concept ◽  
Network Resources ◽  
Network Function ◽  
Large Scale Networks ◽  
Proper Solutions

The combination of Network Function Virtualization (NFV) and Software Defined Networking (SDN) can improve the control and utilization of network resources. However, this issue still requires proper solutions to virtualize large-scale networks, which would allow the use of SDN and Virtualization in real environments.Thus, this paper proposes a virtualization architecture for SDN that relies on a proxy-based approach. The NVP (Network Virtualization Proxy) is a virtualization proxy that intercepts messages exchanged between controllers and switches SDN enabling network virtualization. An implementation of the proposal was developed as a proof of concept and load testing was performed showing that the solution can provide network virtualization in a scalable manner, using less than 2.5 MB of memory to manage 100 switches performing simultaneous requests, whereas FlowVisor requires more than 200 MB.

scholarly journals Hybrid load balance based on genetic algorithm in cloud environment

2021 ◽  
Vol 11 (3) ◽  
pp. 2477
Author(s):  
Walaa Saber ◽  
Walid Moussa ◽  
Atef M. Ghuniem ◽  
Rawya Rizk
Keyword(s):  
Genetic Algorithm ◽  
Load Balancing ◽  
Resource Utilization ◽  
Load Balance ◽  
Large Scale ◽  
Fitness Function ◽  
Optimal Load ◽  
Metaheuristic Techniques ◽  
Heterogeneous Resources ◽  
Cloud Resources

Load balancing is an efficient mechanism to distribute loads over cloud resources in a way that maximizes resource utilization and minimizes response time. Metaheuristic techniques are powerful techniques for solving the load balancing problems. However, these techniques suffer from efficiency degradation in large scale problems. This paper proposes three main contributions to solve this load balancing problem. First, it proposes a heterogeneous initialized load balancing (HILB) algorithm to perform a good task scheduling process that improves the makespan in the case of homogeneous or heterogeneous resources and provides a direction to reach optimal load deviation. Second, it proposes a hybrid load balance based on genetic algorithm (HLBGA) as a combination of HILB and genetic algorithm (GA). Third, a newly formulated fitness function that minimizes the load deviation is used for GA. The simulation of the proposed algorithm is implemented in the cases of homogeneous and heterogeneous resources in cloud resources. The simulation results show that the proposed hybrid algorithm outperforms other competitor algorithms in terms of makespan, resource utilization, and load deviation.

PENGEMBANGAN MEDIA BIG BOOK UNTUK MENINGKATKAN HASIL BELAJAR PECAHAN SENILAI SISWA SD

2018 ◽  
Vol 16 (1) ◽  
pp. 67-76
Author(s):  
Disyacitta Neolia Firdana ◽  
Trimurtini Trimurtini
Keyword(s):  
Learning Outcomes ◽  
Large Scale ◽  
Fourth Grade ◽  
Learning Activities ◽  
Equivalent Fractions ◽  
Fourth Grade Students ◽  
Teacher Needs ◽  
The Media ◽  
Test Result ◽  
Better Than

This research aimed to determine the properness and effectiveness of the big book media on learning equivalent fractions of fourth grade students. The method of research is Research and Development  (R&D). This study was conducted in fourth grade of SDN Karanganyar 02 Kota Semarang. Data sources from media validation, material validation, learning outcomes, and teacher and students responses on developed media. Pre-experimental research design with one group pretest-posttest design. Big book developed consist of equivalent fractions material, students learning activities sheets with rectangle and circle shape pictures, and questions about equivalent fractions. Big book was developed based on students and teacher needs. This big book fulfill the media validity of 3,75 with very good criteria and scored 3 by material experts with good criteria. In large-scale trial, the result of students posttest have learning outcomes completness 82,14%. The result of N-gain calculation with result 0,55 indicates the criterion “medium”. The t-test result 9,6320 > 2,0484 which means the average of posttest outcomes is better than the average of pretest outcomes. Based on that data, this study has produced big book media which proper and effective as a media of learning equivalent fractions of fourth grade elementary school.

Sign in / Sign up

Export Citation Format

Share Document