Resource multi-objective mapping algorithm based on virtualized network functions: RMMA

Existing radio access network systems are static and rigid; they cannot easily satisfy the increasingly large volume of mobile traffic. A new multi-objective optimization approach was developed to leverage the complexity and scalability of radio resource allocation in large-scale radio access networks. A mathematical model of virtualized resource mapping in a heterogeneous radio access network is proposed in this study. We expanded the dynamic differential evolutionary algorithm by regulating the weight parameters of each objective with machine learning to solve the mathematical model. Our approach is evaluated comprehensively in terms of complexity and convergence, and simulations are conducted to verify the proposed approach and demonstrate that the unilateral value of our multi-objective optimization can mirror the results of single-objective optimizations.