Research On Multimedia Transmission And Caching On Smart Media Transport

In recent years,we are experiencing the explosive growth of multimedia consumption.The proportion of multimedia content in Internet service is getting bigger and bigger.Furthermore,the proportion will continue growing in the foreseeable future.With the rapid growth of radio access network,it hosts a significant portion of multimedia consumption on the Internet.In fact,every update of mobile communication network brings significant improvement of bandwidth and service quality.The emergence of immersive media technology,such as Augment Reality,Virtual Reality,Point Cloud and etc.leads to the growth of multimedia data on the order of magnitude.These various changes and developments have put tremendous pressure on the Internet to provide reliable services.Content delivery network provides a viable solution to alleviate this pressure.Since its birth,content delivery network has made fruitful achievements in theoretical research and practical application,innovative research results continue to emerge.Over the years,content popularity prediction using large data technology and machine learning methods is widely studied and applied,which improved caching efficiency.New methods proposed by academics,such as code caching,provide new research directions for caching.These research and applications promote the development of content delivery networks.In the past few years,content delivery networks on radio access network,quality of experience as a multimedia service evaluation index and immersive multimedia technology has received widespread attention.This article is focusing on these aspects and carrying out the following research.First,design a caching stratey in wireless network multicast scenario aiming at optimizing QoE of all users.Second,based on the idea of tile transmission,propose a caching stratey which taken QoE and saliency into consideration.The research results and contributions of the full text can be summarized as the following aspects.1)Propose a caching stratey in wireless network multicast scenario aiming at optimizing QoE of all users.In this part,we divide all uses into groups based on their network speed.Under the restrictions of the cache size,we optimize the grouping(number of groups,users in each group)and the corresponding media bitrate to be pushed to each group to achieve the maximization the total QoE of all users.A caching strategy is designed according to the optimization results.The establishment of the optimization model and the design of the algorithm is decribed in detail in this part,as well as the efficiency of the different algorithms.The model is simulated by matlab and the simulation results are compared with the theoretical analysis.2)Propose a caching stratey which taken QoE and saliency into consideration based on the idea of tile transmission.In a multicast scenario,we optimize the bandwidth resource allocation and user grouping according to the different saliency value of different tiles,and design a caching strategy according to the optimization results.Its advantage lies in improving the utilization efficiency of bandwidth resources.In the paper,we focus on how to set the appropriate factors to build an effective optimization model.Dynamic programming and greedy algorithm are used to solve this optimization model.