There have been many works on indexing techniques to enhance retrieval performance for complex objects in object-oriented databases. Though some of the techniques support efficient evaluations for single-key queries, they are inappropriate for multikey constriant queries. In this paper, we propose multikey indexing techniques which efficiently evaluate multikey constrint queries for complex objects. We evaluate the storage, retrieval, and update costs of the proposed multikey indexing techniques by analytic models. This analytic models are used to determine the most efficient multikey indexing technique for complex objects and to provide significant performance gains.
