Evaluation of middleware architectures in achieving system interoperability

The current state-of-the-art for integration of heterogeneous systems involves manually resolving differences in data modeling and mapping for each interface between systems, in an inherently customized manner. The first step in advancing the state-of-the-art is to develop a general model of the interoperation among systems. Keeping this as the broad objective, in this paper we have worked on classification of modeling differences among autonomously developed, heterogeneous systems as differences in what is being modeled (view) and differences in how the modeled information is represented (representation). Further to this, a set of criteria was selected for conducting an evaluation of existing interoperability approaches in order to compare their success in resolving such heterogeneities. These criteria were used to evaluate seven of the leading approaches for achieving interoperability among independently developed systems. The limitations of these approaches against the criteria provided the motivation for subsequent development of a general model of the interoperation among systems, the Federation Interoperability Object Model (FIOM) used in the Object-Oriented Method for Interoperability (OOMI), to address these shortcomings (Young et al., 2002).