Research On Opportunistic Content Recommendation Algorithm For Mobile Communication Network

Since media and mobilization are two notable trends in contemporary Internet development,the demand for content in mobile communication network growth rapidly.So the importance of the MCDN(Mobile Content Delivery Network)is becoming increasingly prominent,which can cater the trends of the internet simultaneously.On the other hand,there are a lot of redundant resources in the wireless access network,because the base stations are more and more intensive and small,and also because the mobile traffic has an asymmetrical distribution across space and time.So how to make use of this part of the redundant resources reasonably has become more and more important.This paper proposes a new opportunistic content recommendation program based on redundant resources of mobile networks,which conforms to the network development trends.This program combine the content recommendation systems with the mobile network traffic awareness,so that it can implement content recommendations adaptively based on the local redundant resources.Our program uses the redundant resources of the wireless access network to meet the huge needs of the mobile networks contents distribution,and has great application value.This article have studied the algorithm design,performance analysis and demonstration verification of the opportunistic content recommendation system.First of all,for different scenarios and needs,this article presents a variety of cross-layer design with different content recommendation algorithm,and the performance and complexity of different algorithms are complementary.Secondly,we propose a new performance evaluation framework.The difficulty of designing the opportunistic content recommendation system is how to let the opportunity content distribution business coexist with the regular network business.Therefore,our analytical framework takes into account the compromise between the two key indicators:the recommended content quality(quantified for user interest)and the quality of coexistence with the wireless network(quantified as capacity spill rate).Based on the above algorithm analysis framework,this article derives the limits of some performance indicators in theory,and systematically analyzes and compares the performance of various algorithms through simulation.We also compare the proposed algorithms with the traditional content recommendation algorithm,and clarify the advantages of the proposed algorithms.At the end of the article,we build a simple content recommendation platform with JAVA.This platform is a local web application that adapts its recommended content adaptively according to the current input simulation flow parameters.Our research results provide useful theoretical guidance and an engineering design reference for the development of new opportunistic content recommendation systems.