摘要
The fifth-generation(5 G)network cloudification enables third parties to deploy their applications(e.g.,edge caching and edge computing)at the network edge.Many previous works have focused on specific service strategies(e.g.,cache placement strategy and vCPU provision strategy)for edge applications from the perspective of a certain third party by maximizing its benefit.However,there is no literature that focuses on how to effciently allocate resources from the perspective of a mobile network operator,taking the different deployment requirements of all third parties into consideration.In this paper,we address the problem by formulating an optimization problem,which minimizes the total deployment cost of all third parties.To capture the deployment requirements of the third parties,the applications that they want to deploy are classified into two types,namely,computation-intensive ones and storage-intensive ones,whose requirements are considered as input parameters or constraints in the optimization.Due to the NP-hardness and non-convexity of the formulated problem,we have designed an elitist genetic algorithm that converges to the global optimum to solve it.Extensive simulations have been conducted to illustrate the feasibility and effectiveness of the proposed algorithm.
The fifth-generation(5 G) network cloudification enables third parties to deploy their applications(e.g.,edge caching and edge computing) at the network edge. Many previous works have focused on specific service strategies(e.g., cache placement strategy and vCPU provision strategy) for edge applications from the perspective of a certain third party by maximizing its benefit. However, there is no literature that focuses on how to effciently allocate resources from the perspective of a mobile network operator, taking the different deployment requirements of all third parties into consideration. In this paper, we address the problem by formulating an optimization problem,which minimizes the total deployment cost of all third parties. To capture the deployment requirements of the third parties, the applications that they want to deploy are classified into two types, namely, computation-intensive ones and storage-intensive ones, whose requirements are considered as input parameters or constraints in the optimization.Due to the NP-hardness and non-convexity of the formulated problem, we have designed an elitist genetic algorithm that converges to the global optimum to solve it. Extensive simulations have been conducted to illustrate the feasibility and effectiveness of the proposed algorithm.
基金
the National Natural Science Foundation of China(No.61972026)。
