A hybrid NSGA-II for solving multiobjective controller placement in SDN

Ahmadi, Vahid; Jalili, Ahmad; Khorramizadeh, S. M.; Keshtgari, Manijeh

doi:10.1109/kbei.2015.7436122

Cited by 14 publications

(9 citation statements)

References 10 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…They use Nash bargaining to realize the trade‐off between these two latencies while considering controller load balancing as a constraint. In a multiobjective, combinatorial optimization (MOCO) formulation, researchers employ meta‐heuristic approaches such as NSGA‐II [3, 63], and MOGA [94] to obtain a Pareto‐optimal solution. Several researchers [75, 81, 123] view the problem as one of partitioning, intending to improve both the metrics.…”

Section: Cpp Solution Methodsmentioning

confidence: 99%

“…The different proposed approaches differ in the representation of the chromosome and the design of fitness function. Several papers have encoded genes either as controller locations [3, 4, 55, 63, 64] or as both controller locations and switch assignments [94]. The other tactic is to code the genes as switch locations and use the EAs to partition the switches into k SDN domains [20, 124].…”

Section: Cpp Optimization Strategiesmentioning

confidence: 99%

See 1 more Smart Citation

Controller placement problem in software‐defined networking: A survey

Kumari

Sairam

2021

Networks

View full text Add to dashboard Cite

A distinctive feature of software‐defined networking (SDN) is a logically centralized control plane realized using multiple physical controllers. The placement of the controllers, the so‐called controller placement problem (CPP), is a crucial design issue. It influences network performance parameters such as latency, flow setup time, network availability, load balance of the controllers, and energy consumption. In this article, we illustrate the formulation of these CPP objectives. We categorize the CPP design solutions as either static or adaptive. In adaptive CPP, the solutions proposed dynamically adapt to the number of controllers required and the switch to controller mapping to varying network traffic. We further differentiate adaptive CPP as wired or wireless. The optimization strategies adopted by the papers are analyzed and grouped into five categories: exact, heuristic, meta‐heuristic, clustering, and game theory. The merits and demerits of each approach are discussed. In conclusion, we outline the research challenges worth investigating.

show abstract

Section: Cpp Solution Methodsmentioning

confidence: 99%

Section: Cpp Optimization Strategiesmentioning

confidence: 99%

Controller placement problem in software‐defined networking: A survey

Kumari

Sairam

2021

Networks

View full text Add to dashboard Cite

show abstract

“…To verify the performance of PSOAP, we set up a series of simulation experiments, and PSOAP was compared with the AP and Genetic Algorithm (GA) algorithms. A heuristic algorithm was used to solve the multi-controller deployment problem in [22] and [23], and obtained a good performance.…”

Section: B Simulation Analysismentioning

confidence: 99%

“…Then, a dynamic deployment method based on the Pareto optimal controller was proposed [20] and a heuristic algorithm based on Pareto simulated annealing was proposed [21]. Ahmadi et al [22] proposed a heuristic algorithm called hybrid NSGA-II, which can get faster computation times and need much less memory to perform. Jalili et al [23] considered the latency between nodes and load balancing as important metrics, NSGA-II was introduced to solve the multi-objective model of control placement problem.…”

Section: Introductionmentioning

confidence: 99%

SDN-Based Load Balancing Scheme for Multi-Controller Deployment

Wang

Zhang

2019

IEEE Access

View full text Add to dashboard Cite

Software-defined networking (SDN) separates the control plane from the data forwarding plane and realizes the flexible management of the network resources. With the explosive growth of network traffic and scale, multi-controllers need to be deployed to improve the scalability and reliability of the control plane. However, unreasonable subdomain partitioning of SDN controllers may cause the unbalanced distribution of controller loads and reduces the communication performance of the network. Therefore, in this paper, a dynamic multi-controller deployment scheme based on load balancing is proposed. We transform the flow requests into a queuing model and consider the traffic propagation delay and the capacity of controllers as two main factors affecting the deployment of the multi-controllers. In the initial static network, a modified affinity propagation algorithm (PSOAP) based on particle swarm optimization is proposed to solve the problem of clustering performance being affected by the initial values of the bias parameters and convergence coefficients, getting the reasonable network planning. With the dynamic traffic network, switches in different sub-domains are reassigned by breadth-first search (BFS) algorithm to achieve controller load balancing. The extensive evaluations demonstrate that the scheme can provide better stable, accurate, and load balancing multi-controller deployment when compared with affinity propagation (AP) and genetic algorithms.INDEX TERMS Affinity propagation algorithm, load balancing, queuing model, software-defined networking.

show abstract

“…These evolutionary algorithms have the capability to obtain Pareto optimal solutions in relatively few iterations. A number of researchers [119], [120], [121], [122] have tried to solve the problem of minimizing latency and capacity management using a modified version of NSGA-II [123](non-dominating sorting genetic algorithm). The modification is in terms of greedy initialization among others.…”

Section: B Latency and Capacitymentioning

confidence: 99%