Research On Near-Filed Content Delivery System Based Cloud And Its Key Technology

With the advent of Mobile Internet, we can forecast that there will be a popular content consuming behavior that users watch the video online or download video on mobile phone. The varieties of users’demand of services make the existing wireless network face with tremendous data traffic. Current focal and difficult points on effective network control mainly lies in how the network infrastructure can adapt to the dramatic increase of network traffic, and how the user experience can be guaranteed and optimized.To address to the question, the paper proposes a content delivery system, which is based on the framework of Cloudlet, and takes advantage of the near-field data transmission capacity based on multiple terminals collaboration. The system can index the content completely through WAN and LAN, share the communication capability and content of the near-field terminal quickly and effectively, guaranteeing the continuity of the service process with a series of adaptive technical measures when the user is in a mobile environment. The main objects are making full use of the near-field network and computing resources, relieving the pressure on the backbone network, optimizing the user experience in a mobile environment.The paper also introduces group recommendation into the object system by analyzing the user context data from the system and the social-networking data from sina micro-blog. With the guidance of recommendation, the total out-band throughput on traffic will decrease, and efficiency of resource utilization can be achieved. The paper verifies the feasibility and potentiality of applying group recommendation to the proposed system based on the experimental evaluation.