A model based on genetic algorithm for service chain resource allocation in NFV
Ningning Ma, Jiao Zhang, Tao Huang · 2017
In traditional network, network functions are implemented by proprietary hardware. Its deployment, update and upgrade are complicated and costly. With the rapid growth of the number of end users and Internet services' diversification recently, it is a huge challenge for operators to deliver network services. In recent years, the emergence of Network Function Virtualization technology effectively alleviates this problem. However, the problem of resource allocation for service chains comes up successively. This paper presents an effective model on resource scheduling for service chain, and aims to explore a user-friendly Pareto optimal target. The probability of nodes' breakdown is also concerned. An improved genetic algorithm is used to search the best scheduling result. The model is proved to be effective by measuring the last service completion time and compute performance in simulations.