为了实现网络资源的动态分配,提高网络资源利用率,满足用户业务多样性带来的切片网络差异需求,该文提出一种基于网络效用最大化的虚拟资源分配算法。该算法采用商业化模式将频谱资源作为收益载体,并对不同切片网络进行差异化定价。同时将计算资源和回程链路作为开销,还考虑了切片网络对计算资源和频谱资源的差异性需求,最后以最大化网络收益建立效用模型。并通过拉格朗日对偶分解设计了分布式迭代算法对效用模型进行求解。仿真结果表明,该算法提高了服务用户比例,并使得网络资源获得最大收益。
To realize the dynamic allocation of network resources, improve the network resources utilization and meet the demand of the diverse networks, this paper proposes a virtual resource allocation algorithm based on network utility maximization. The spectrum resource is used as the revenue and the differentiated price is commercialized according to slicing networks. It also takes the computing resources and the backhaul as the cost, and also takes into account the different demands of the slicing network on the computing resources and spectrum resources. Finally, the utility model is established to maximize the network revenue. A distributed iterative algorithm is designed to solve the utility model by Lagrangian dual decomposition. The simulation results show that the algorithm improves the percentage of service users and maximizes the network revenue.