Cardinality estimation of distributed join queries

Estimation of selectivities or intermediate result sizes are useful in query optimization, as a means of determining the feasibility of queries, and as a quick way of answering queries for which the size of the answer is of interest in its own right. We propose a statistical method for estimating the cardinality of the resulting relation after an equi-join and projection. Under various "uniformity" assumptions on the data, this method relies on statistical information maintained by the system. The result of some experiments on random generated queries and databases are shown. Our method achieves low relative errors, typically about 10%.