A Cooperative Caching Method Based on Graph Theory in Fog Wireless Access Network

Abstract

The invention discloses a graph theory based cooperative caching method in an F-RAN (Fog-Radio Access Network). The cooperative caching method comprises the steps of 1, building a connecting side setof F-AP (Fog-Access Points) of an access layer based on distances and load differences among the F-AP; 2, building an adjacent table set of the access layer; 3, building a maximum complete sub-graph gm; 4, dividing candidate cooperative caching clusters for the F-AP of the access layer and calculating a forward transmission unloading increment value Ain of each candidate cooperative caching cluster; 5, regarding each candidate cooperative caching cluster as a vertex, regarding the forward transmission unloading increment value as a weight value, and building sides between the corresponding vertexes of two different candidate cooperative caching clusters which contain same F-AP and then building a graph; and 6, solving the maximum weight independent sub-set of a vertex set, wherein the candidate cooperative caching clusters corresponding to the vertexes in the maximum weight independent sub-set are the cooperative caching cluster set with the maximum forward transmission unloading quantity. The graph theory based cooperative caching method in the F-RAN can perform reasonable cluster division on the F-AP in the F-RAN; and the calculation complexity is low, and the global performanceis relatively high.

Description

Technical field [0001] The invention relates to a cooperative caching method based on graph theory in a fog wireless access network, and belongs to the technical field of caching in a 5G mobile communication system. Background technique [0002] Fog-Radio Access Network (F-RAN, Fog-Radio Access Network), as a new system architecture of 5G communication systems, has received extensive attention in the communication field with its huge advantages in alleviating fronthaul congestion and reducing communication delays. Fog-Access Point (Fog-Access Point) in F-RAN can cache popular content in advance at off-peak hours, which is the key to realizing fronthaul offloading and delay reduction. However, the storage space of F-AP is limited, and the content popularity distribution in the coverage area of ​​F-AP is relatively scattered, and the edge cache performance gain of F-RAN is therefore greatly restricted. Divide F-AP into clusters, and deploy and share cached content between nodes in...

AI Technical Summary

Problems solved by technology

[0003] In actual scenarios, due to the complex distribution of content popularity among F-APs, and the dense deployment of F-APs, the arrival rate of content requests in the coverage area of ​​F-APs also changes with time and is unevenly distributed. These factors lead to It is very difficult to quickly and accurately select a suitable cooperative node cluster from a large number of fog wireless access nodes, and an inaccurate clustering scheme can easily lead to local performance improvement of cooperative caching and poor global performance.

[0004] In the fog wireless access network, it was previously assumed that the distribution of content popularity in the coverage area of ​​different fog wireless access nodes is the same, that is, subject to the global content popularity distribution. Under this assumption, if the number of nodes in the cluster is the same , the choice of which nodes to cooperate in clusters has no effect on the performance gain of collaborative caching, but in reality, due to the unique content preferences of users, there are complex distributions of content popularity in different regions After considering the differences in the distribution of local content popularity, the construction method of the cooperative cluster directly affects the distribution of content popularity in the cluster, thereby affecting the deployment of cache content and the amount of fronthaul offload that can be obtained by the cooperative cache of all nodes. The deployed fog wireless access nodes are quickly and accurately constructed as a suitable cooperative cluster, so that the concentration of content popularity distribution within the cluster and the difference of content popularity distribution between clusters are optimized, thereby maximizing the improvement of all fog wireless access nodes. The total fronthaul offload of the ingress cooperative cache, which is a very complex problem

Images