Prediction-Based Cache Scheme for Named Data Networking

• Main Authors: Hsiang-Yin Fang, 方相尹
• Other Authors: Han-Chieh Chao
• Format: Others
• Published: 2015
• Online Access: http://ndltd.ncl.edu.tw/handle/98716420403121656845

碩士 === 國立東華大學 === 電機工程學系 === 103 === Internet technologies are developed in tremendous speed nowadays that causing many people are very dependent on these technologies. Such technologies include huge multimedia traffic which has gone beyond the general network traffic. The reason is that there are more and more mobile devices are used, such as tablet computers, smart phones and so on. However, they will result in insufficient bandwidth due to they have very big chance to be filled by huge multimedia traffic. Therefore, Named Data Networking (NDN) has been proposed that it is a name-based architecture rather than IP-based architecture. In addition, NDN has content cache for content delivery routing scheme so that it is able to reduce more unnecessary traffic consumption as well as the transmission speed will be increased. In this way, the network resource can be used efficiently. NDN has been a popular topic in direction of network technologies. There are many scholars studied for how to increase the hit rate of cache. However, they did not consider the wastage ratio of network resource. In this thesis, we combine the NDN and virtualization mechanism as a novel network transmission scheme. The main ideas are that we propose SCA and DCA algorithm which can predict the expected storage size to adjust the cache so that every router have an ability to adaptively use and release the cache space thus solve the wastage ratio of cache space. Finally, the simulation results show that our proposed method will reduces more unnecessary traffic and wastage ratio of cache.