Affiliations: Department of Computer Engineering, Hankyong
National University, Anseong, Republic of Korea | School of Computer Science and Engineering, Pusan
National University, Busan, Republic of Korea | Department of Computer Science, Texas A&M
University – Commerce, TX, U.S.A
Abstract: Mobile computing is currently becoming more common. This trend
emphasizes the necessity of accessing database systems on fixed networks from
mobile platforms via wireless networks. However, it is not an appropriate way
that applies the database access methods for traditional computing onto mobile
computing because of their essential restrictions. In this paper, we suggest an
agent-based mobile data access model and also two functions calculating data
significance scores to choose suitable data items for cache replacement and
coherence policies. These functions synthetically reflect access term, access
frequency and tendency, update frequency and tendency, and data item size
distribution. As the result of simulation experiment, our policies outperform
Least Recently Used (LRU), LRU Inverses Frequency (LIX), Stretch Access-rate
Inverse Update-frequency (SAIU) policies in aspects of decrement of access
latency, improvement of cache byte hit ratio, and decrease of cache byte pollution ratio.
Keywords: cache replacement, coherence, data significance, mobile, database
