Research On Edge Caching Technology For Video Content In Small Cell Networks

With the rapid development of wireless communication technology,exponential growth of mobile video traffic makes the traditional cellular network bandwidth face enormous pressure.In response to this challenge,small cell technology was presented.As small cell with high capacity,easy to deploy,low power consumption and low cost advantages,so it was widespread deployment in the 4G mobile network,which not only to relieve the pressure of traditional cellular network bandwidth,but also can greatly enhance the system capacity.However,due to the limited bandwidth of the wireless backhaul link,the performance of small cell network is expected to be improved.Edge caching technology can make up for the shortcomings of small cell network system,that is,to reduce the network bandwidth occupancy rate of the wireless link by placing the network video content in the edge base station,thus the overall performance of the small cell network will be enhanced.In the current network traffic,the proportion of video traffic is over seventy percent,As the video file caching has the following characteristics:high storage resource consumption,difficulty in file popularity prediction,high dependency on the user access time,frequency and other access characteristics.So this article will focus on the edge caching technology based on video content in the network of small cell.In this paper,the above problems were studied,the main work is as follows:First,we present a scheme of the placement of video content in small cell network.Firstly,the effects of different popularity and diverse file sizes on edge caching efficiency is analyzed.Then combined with the characteristic of the actual file size exists in different,design the partial content caching scheme,establish a partial content placement scheme model under the small cell network scenario,and consider the user's tolerance of delay,aim to maximize the cache hit ratio.finally,introduced the between-cluster variance method to obtain the best threshold cache.Simulation results show that the proposed scheme can improve the cache hit ratio while ensuring user delay.Second,we propose an edge cache replacement strategy based on cache partition.Firstly,analyze the importance of the popularity prediction in the cache replacement strategy and the influence of the speed of popularity change on the estimation error of popularity.Then based on the different rate of popularity change between the dated video content and the fresh video content,a partitioned strategy of edge caching replacement is proposed.establish the model structure of the video content replacement strategy under the small cell network scenario,in modeling,we consider using different popularity estimation algorithm for fresh and dated video to make partitioned caching.introduce the content relevance,user access behavior characteristics to estimate the fresh video popularity,then based on the popularity estimate value to execute replacement.Simulation results show that the proposed scheme can not only improve the cache hit ratio,but also reduce the number of replacement.