End-to-end network resource allocation method based on Bayesian learning

Abstract

The invention discloses an end-to-end network resource allocation method based on Bayesian learning, and belongs to the technical field of calculation, reckoning or counting. According to the method, a solution based on a game framework is adopted, a network resource pricing strategy reflecting service provider demand information is provided, and more reasonable network resource pricing and distribution are realized through analysis of network resource states and service provider behaviors. The method comprises the following steps: firstly, establishing game models of a plurality of service providers and operators, and selecting a unit network slice price of a first stage by the operators according to limited demand information; and updating the demand information in the second stage to determine the price. A simulation result shows the existence of a Nash equilibrium point of the system model, and a response type network resource slice allocation scheme is provided, so that resource shortage and network congestion caused when a service provider intensively requests resources are avoided, and reasonable allocation of network resources is realized.

Description

technical field [0001] The invention relates to communication technology, specifically discloses an end-to-end network resource allocation method based on Bayesian learning, and belongs to the technical field of calculation, calculation or counting. Background technique [0002] In recent years, it is difficult for existing networks to support application scenarios in 5G networks. It is unrealistic to deploy multiple different communication networks for different application scenarios, and the huge costs incurred are beyond the reach of operators. Therefore, the Next Generation Mobile Network Alliance has proposed the concept of network slicing, that is, on a physical infrastructure, different network slices are built on demand by using software-defined networks and technologies such as network function virtualization. Various resource allocation schemes aim to achieve resource isolation between slices, so that the failure of one slice in a dynamic network environment will n...

AI Technical Summary

Problems solved by technology

At present, most of the resource scheduling studies continue the idea of ​​traditional network resource scheduling algorithms, such as realizing virtual network resource allocation through network mapping based on load balancing, ignoring the actual needs of service providers.

Images