Network function virtualization intelligent scheduling method based on flow identification

Abstract

A network function virtualization intelligent scheduling method based on flow identification is realized based on a confidence region strategy optimization algorithm, and comprises the following operation steps: (1) classifying network data flows: (2) setting a state space of the confidence region strategy optimization algorithm; (3) setting an action space of a confidence region strategy optimization algorithm; (4) setting a reward function of a confidence region strategy optimization algorithm; (5) setting a loss function of the confidence region strategy optimization algorithm; and (6) on the basis of the previous steps, realizing intelligent scheduling of the virtual network function by utilizing a confidence region strategy optimization algorithm. According to the method, the intelligent scheduling of the virtual network function can be realized according to the service quality requirements of different flow data, and the network utilization rate and the network service performance are improved.

Description

technical field [0001] The invention relates to an intelligent scheduling method for network function virtualization based on traffic identification, which belongs to the field of information technology, in particular to the technical field of network function virtualization. Background technique [0002] With the rapid growth of network traffic data and the continuous popularization of network applications, network users have higher and higher requirements for service quality, including low delay, low packet loss rate and high throughput. Therefore, network equipment not only needs to increase bandwidth and Capacity requires an appropriate resource scheduling mechanism. Network Function Virtualization (NFV), that is, network function virtualization technology, separates network functions from traditional hardware devices, and can flexibly arrange network functions and share network resources. NFV technology uses virtualized network functions (VNF), namely virtual network f...

AI Technical Summary

Problems solved by technology

However, most of these works are limited to identifying a small number of specific network applications or traffic data in a specific network environment. It is difficult to play a role after the traffic packets are encrypted, and the scope of application is relatively narrow.

However, the dynamic orchestration scheme of virtual network functions is regarded as an online decision-making problem. Existing research is limited to modeling the environment to obtain the optimal scheme, ignoring the optimal long-term benefits extracted from historical experience.

Images