Data caching framework and strategies for integrity constraint checking in mobile database system

• Main Author: Dzolkhifli, Zarina
• Format: Thesis
• Language: English
• Published: 2012
• Online Access: http://psasir.upm.edu.my/id/eprint/31410/1/FSKTM%202012%209R.pdf; http://psasir.upm.edu.my/id/eprint/31410/

In a mobile environment, due to the various constraints inherited from limitations of wireless communication and mobile devices, checking for integrity constraints to maintain the consistent state of mobile databases is an important issue that needs to be addressed. Due to limited storage capabilities, a mobile host is not capable of storing all data items in the network, thus it must share some data item with a database in the fixed network. Data caching technique is used to cache frequently accessed data from the base station into mobile hosts. By caching the needed data items, it allows the mobile host to continue processing without worrying about disconnection. The main aim of this research is to propose a framework of data caching for the purpose of checking integrity constraints in a mobile environment. Four caching strategies have been proposed, namely: Blind Strategy (BS), Quantifier Strategy (QS), Strict Strategy (SS), and Relax Strategy (RS) which are embedded in the proposed framework. The BS strategy blindly caches data items into mobile host while the QS strategy caches data items based on the quantifier of the integrity test. The SS and RS strategies analyze the relationships between integrity tests in identifying the data items to be cached. These four strategies adopted the simplified form of integrity tests, which are sufficient and complete tests. The RS and SS strategies improve the checking mechanism by not only preventing delays during the process of checking constraints and performing the update but also reducing the amount of data items to be cached. The result shows that both the strategies managed to cache small amount of data items and at the same time achieved nearly 100% of hit ratio.