Processing aggregate queries on spatial OLAP data

Spatial OLAP refers to the confluence of two technologies: spatial data management and OLAP for non-spatial data. An example query on spatial OLAP data is to find the total amount of gas sold within a period of time, of certain type of gas, to a certain type of customers, by gas stations located within a query window on a map. A framework has been established in [8], which converts this query into a set of queries for a general-purpose ROLAP system. However, there has been little done at the query optimization level, once they are submitted to the OLAP system. In this paper, we develop and implement 3 query processing strategies of very different nature on our experimental MO-LAP system. Detailed experimental performance data are presented, using a real-life spatial database with 1/3 million of spatial objects.