Fclos: A Client-server Architecture For Mobile Olap

摘要

Mobile online analytical processing (mOLAP) encompasses all necessary technologies for information systems that enable OLAP data access to users carrying a mobile device. This paper presents FCLOS, a complete client-server architecture explicitly designed for mOLAP. FCLOS founds on intelligent scheduling and compressed transmissions in order to become a query efficient, self-adaptive and scalable mOLAP information system. Scheduling exploits derivability between data cubes in order to group related queries and eventually reduce the necessary transmissions (broadcasts). Compression is achieved by the m-Dwarf, a novel, compressed data cube physical structure, which has no loss of semantic information and is explicitly designed for mobile applications. The superiority of FCLOS against state of the art systems is shown both experimentally and analytically.