Bekhit, M, Abolhasan, M, Lipman, J, Liu, RP & Ni, W 2018, 'Multi objective resource optimisation for network function virtualisation requests', 26th International Conference on Systems Engineering, ICSEng 2018 - Proceedings, International Conference on Systems Engineering, IEEE, Sydney, Australia.View/Download from: Publisher's site
© 2018 IEEE. Network function vitalization (NFV) as a new research concept, for both academia and industry, faces many challenges to network operators before it can be accepted into mainstream. One challenge addressed in this paper is to find the optimal placement f or a set of incoming requests with VNF service chains to serve in suitable Virtual Machines (VMs) such that a set of conflicting objectives are met. Mainly, focus is placed on maximizing the total saving cost by increasing the total CPU utilization during the processing time and increasing the processing time for every service request in the cloud network. Moreover, we aim to maximize the admitted traffic simultaneously while considering the system constraints. We formulate the problem as a multi-objective optimization problem and use a Resource Utilization Multi-Objective Evolutionary Algorithm based on Decomposition (RU-MOEA/D) algorithm to solve the problem considering the two objectives simultaneously. Extensive simulations are carried out to evaluate the effects of the different network sizes, genetic parameters and the number of server resources on the acceptable ratio of the arrival chains to serve in the available VMs. The empirical results illustrate that the proposed algorithm can solve the problem efficiently and compute the optimal solution for two objectives together within a reasonable running time.
Gamal, M, Abolhasan, M, Jafarizadeh, S, Lipman, J & Ni, W 2019, 'Mapping and Scheduling of Virtual Network Functions using Multi Objective Optimization Algorithm', Proceedings - 2019 19th International Symposium on Communications and Information Technologies, ISCIT 2019, International Symposium on Communications and Information Technologies, IEEE, Ho Chi Minh City, Vietnam, pp. 328-333.View/Download from: Publisher's site
© 2019 IEEE. Within the context of Software-Defined Networking (SDN), the problem of resource allocation for a set of incoming Virtual Network Functions (VNF) service requests has been the focus of many studies. In this paper, a new optimization model has been developed to find the near to optimal mapping and scheduling for the incoming VNF service requests. This model while considering delay, aims to achieve three objectives functions, namely, minimizing the transmission delays occurring in every link, minimizing the processing capacity for every Virtual Machine (VM) and minimizing the processing delay at every VM. The resultant problem is formulated as a multi-objective optimization problem and the developed solution is based on a multi-objective evolutionary algorithm utilizing the decomposition algorithm. Simulation results illustrate that the resulting algorithm is scalable while considering delay and it outperforms the genetic bandwidth link allocation (GA-BA) and genetic non-bandwidth link allocation (GA-NBA) algorithms.
Gamal, M, Jafarizadeh, S, Abolhasan, M, Lipman, J & Ni, W 2019, 'Mapping and scheduling for non-uniform arrival of virtual network function (VNF) requests', IEEE Vehicular Technology Conference, Vehicular Technology Conference, IEEE, Honolulu, HI, USA.View/Download from: Publisher's site
© 2019 IEEE. As a new research concept for both academia and industry, there are several challenges faced by the Network Function Virtualization (NFV). One such challenge is to find the optimal mapping and scheduling for the incoming service requests which is the focus of this study. This optimization has been done by maximizing the number of accepted service requests, minimizing the number of bottleneck links and the overall processing time. The resultant problem is formulated as a multi- objective optimization problem, and two novel algorithms based on genetic algorithm have been developed. Through simulations, it has been shown that the developed algorithms can converge to the near to optimal solutions and they are scalable to large networks.