Resource allocation method under H-CRAN based on multi-time scale collaborative optimization

Abstract

The invention discloses a resource allocation method under H-CRAN based on multi-time scale collaborative optimization. Resource allocation is taken as a two-layer model of large time scale-small timescale, from a large time scale perspective, according to long-term average information of a user moving track and the like, a large time scale resource allocation model based on an expert system is constructed by querying and matching priori knowledge of the expert system; from a small time scale prospective, according to short-term instant information of channel quality and the like, self-learning and self-optimizing characteristics of deep reinforcement learning are utilized to construct a small time scale resource allocation model based on the deep reinforcement learning; from a collaborative optimization prospective, based on a master-slave game view, a collaborative optimization resource allocation model taking the large time scale as the principle factor and the small time scale asthe secondary factor is constructed, and the resource allocation of a heterogeneous cloud wireless access network is achieved.

Description

technical field [0001] The invention relates to a resource allocation method based on multi-time scale collaborative optimization under H-CRAN, and belongs to the technical field of network resource allocation. Background technique [0002] The mobile Internet is constantly changing all aspects of human life. Smart terminals have almost become a necessity for human life. Mobile traffic is growing at an alarming rate. Humans have increasing demands for high-speed applications, such as ultra-high-definition video and driverless driving. 5G came into being. Compared with the 4G system, 5G aims to achieve at least a 100-fold increase in peak rate, at least a 10-fold increase in spectral efficiency and energy efficiency, and at least a 25-fold increase in coverage capacity. In order to truly achieve the above goals, the network architecture is the first issue that needs to be considered. The network architecture dominates the research and implementation of technologies to a cert...

AI Technical Summary

Problems solved by technology

However, due to the introduction of macro base stations, the interference between macro base stations and RRH has become the main factor limiting the improvement of network performance

At the same time, in the traditional orthogonal multiple access technology, only one user is allowed to be allocated to one orthogonal resource, which severely limits the network throughput and the number of user connections.

Images