Predictive Data Allocation and Cache Replacement Mechanisms for Wireless Multimedia Information Systems

• Main Authors: Kuo-Chen Wu, 吳國禎
• Other Authors: Shin-Mu Tseng
• Format: Others
• Language: zh-TW
• Published: 2003
• Online Access: http://ndltd.ncl.edu.tw/handle/90948719410432570264

碩士 === 國立成功大學 === 資訊工程學系碩博士班 === 91 === Accessing multimedia data under wireless communication network is more and more popular because of the increased bandwidth. However, with the bandwidth is getting wider, the multimedia data size is growing even faster. When users request multimedia data at the peak time, even the fixed network translation between base-stations will be overloaded. In this research, we propose a data mining-based method, namely PPCR（Predicted Preload and Cache Replacement）, to predict the users’ future locations and requested multimedia services under wireless communication environments. Based on the mined user access patterns, the predicted service contents will be allocated and pre-loaded into the cache in the predicted locations such that the response time for user’s requests can be reduced. Meanwhile, we also propose a cache replacement mechanism that considers more issues than traditional cache replacement policies, like the multimedia data size and the accessing probability of the data. Though experimental evaluation, PPCR was shown to outperform other methods in reducing the uses’ response time when accessing multimedia data under wireless environments.