An Improved Cache Strategy Based On WSP Cooperation

  • 格式:docx
  • 大小:120.62 KB
  • 文档页数:3

AnImprovedCache Strategy Based On WSP Cooperation Abstract: The rapid development of Video On Demand (VoD) service has brought huge bandwidth costs. In addition ,With the rapid advances in wireless network data services , the request of VoD in wireless network is growing at the same time. Which aggravate the crisis of the core network bandwidth requirements. To solve this problem, Wireless Service Providers (WSP) began deploying cache in the Mobile Switching Center (MSC). It can effectively reduce the bandwidth consumption of the core network and improve the user experience. We find the top 10% popular video get 80% traffic of the core network and the popularity of the most of popular video every day change stably by analyzing the video viewing behavior of the SINA mobile video user. It proved that the deployment of cache in MSC by WSP can effectively reduce the core network bandwidth consumption. On this basis, we propose two optimization models. One is MSC cooperate in WSP, The other one is MSC cooperate between WSPs. And then we propose an improved heuristic cache algorithm. We evaluate the proposed models and the improved heuristic cache algorithm with the methods –trace-driven. Results show that both the cooperation optimization models and the improved heuristic cache algorithm are very effective. And the latter is more efficient. In addition, we research on the update cycle of the cache. And we propose a method of determining the update cycle. Keywords: Collaborative Caching algorithm; Wireless Service Provider; Integer(0-1) Programming; Mobile VOD System; update cycle Ⅰ. INTRODUCTION With the increasing network bandwidth, the development of Video On Demand service get rapid. Up to 2012, the global video traffic account for 57% of all the internet traffic. However, it is expected to reach 69% in 2017. At the same time, as the mobile device get more and more popular and the wireless communication technology development and promotion, the traffic of mobile video data will account for two-thirds of all the core network bandwidth. To enhance the user experience,There are serious problems: avoiding network congestion and reducing network bandwidth load and WSP server load. Wireless Service Providers using a variety of methods to deal with the above problems. On the one hand, they store the content on the edge nodes of the network by deploying the Content Delivery Network (CDN). It makes that data get close to users. In addition, P2P technology is used to make the content distributed near the users. Both of the above methods save the core network bandwidth and enhance the user experience; on the hand, cache is used to reduce the core network bandwidth caused by getting video from the source server. With the rapid development of wireless networks, Wireless Service Providers have begun deploying cache servers on the Mobile Switching Center (MSC) to store content in a place closer to users. Which can reduce bandwidth costs, while improving the user experience. The most commonly used cache algorithm is divided into two categories: one is distributed cache algorithm, common distributed cache algorithm include Least Recently Used (LRU), Least Frequently Used (LFU) and some improved algorithms such as LRU-K and LRFU. Among those cache algorithms cache algorithms based on LRU focus on temporal locality of request access, however, the other cache algorithms based on LFU focus on frequency characteristics. The other is collaboration cache. Most of collaborative cache algorithms solve cache problem through regarding it as a relay model. Some others cooperate cache with game. They achieve optimal resource allocation by balancing several competing factors. Bao Chun Li propose a cache strategy between ISP. They cooperate peers in different ISP to accomplish collaboration cache. In which they take inter-ISP traffic model and resource allocation of cache and protocol between ISP peers into consideration. Thereby, they build a cooperation cache network between ISP. In the wireless network, both distributed cache algorithm and collaboration cache algorithm are applicable. In the distributed cache algorithm take measurement and analysis about user viewing behavior into consideration to make it more effective. However, considering that collaboration cache algorithm can make wireless network cache resources integrated together, thereby, it can avoid the waste of cache resources and reduce the cache redundancy, by which we can have more space to cache more content so that make cache have better performance in wireless network. Jian Hao Hu propose that building a two-level cache on MSC and Base Station restore packet together to avoid QoS Dropping caused by losing packets. M.Dahlin research on cache technology between Internet and World Wide Web. However, due to the dynamic topology in wireless network and limited resources and other problems, those technologies are not well applicable. Jie Dai assume thateach WSP is selfish, then propose a collaboration cache algorithm between WSP, it can improve video QoS by maximizing social welfare.Qijun Zhu propose a collaboration cache algorithm based on mobile P2P network. This paper focuses on video cache algorithm based WSP cooperation. The main contents are as follows: First,building models for the following three cases: 1.without MSC collaboration. 2. MSC collaboration inter-WSP. 3. MSC collaboration between WSP. It convert resource allocation problem into an Integer (0-1) Linear Programing model. Then we get the optimal solution through the branch and bound method. Then, since this problem is a typical NP-hard (Non-deterministic Ploynomial hard) problem, the solution time is very long. This paper proposes a heuristic algorithm got a nearly optimal solution but reduces the time overhead. In this paper, the Ⅰ chapter is introduction, it presents the meaning and background of the topic and reviews the research status of video delivery network and describes the main contents of this article. The Ⅱ chapter propose the optimal models. The Ⅲ chapter details the proposed heuristic cache algorithm. The Ⅳ chapter is evaluation, we show all the results in this chapter. The Ⅴ chapter determines the update cycle. The Ⅵ chapter make a conclusion about the research work of this article and describes the future work. Ⅱ Model With the development of wireless technology, the video traffic in wireless network grow rapidly. Wireless Service Providers have begun deploying cache servers in MSC to reduce the load on the core network. The cache strategy based on collaboration between WSPs is shown in Figure 2-1.