Wireless network resource allocation method based on generative adversarial reinforcement learning

Abstract

The invention discloses a wireless network resource allocation method based on generative adversarial reinforcement learning, and belongs to the field of wireless resource allocation and reinforcementlearning. The method comprises the following steps: initializing a generator network G and a discriminator network D, executing resource allocation, training the weights of the generator network G and the discriminator network D, and finally realizing wireless network resource allocation. Compared with a DQN-based resource allocation method and an average resource allocation method, the resourceallocation strategy obtained by the invention can obtain a higher system return value, i.e., higher spectral efficiency and better user experience.

Description

technical field [0001] The present invention relates to the field of wireless network resource allocation and reinforcement learning, and more particularly, to a wireless network resource allocation method based on generative confrontation reinforcement learning. Background technique [0002] The 5G network will support a large number of diverse business scenarios from vertical industries, such as smart security, high-definition video, telemedicine, smart home, autonomous driving, and augmented reality, etc. These business scenarios usually have different communication requirements, such as augmented reality technology requires more Low latency, autonomous driving technology requires the network to provide higher reliability. However, traditional mobile networks are mainly designed to serve a single mobile broadband service and cannot adapt to the diversified service scenarios of 5G in the future. Building a dedicated physical network for each business scenario will inevita...

AI Technical Summary

Problems solved by technology

[0007] However, the traditional method based on action value learning (such as deep Q network) is difficult to cope with the disturbance and the uncertainty of immediate reward in the environment. Therefore, distribution reinforcement learning is introduced. The main change is to directly estimate the action value distribution. Instead of estimating the expectation of the action value as traditional methods

Images