Decision support systems or mission control systems for network-centric operations typically need input from multiple heterogeneous information sources. While the number and size of information sources increase, information integration and its semantic interoperability are becoming a growing concern and a major challenge to information management. In this chapter, we will share our experience of enabling semantic-based dynamic information integration across multiple heterogeneous information sources. While data is physically stored in existing/legacy data systems across the networks, the information is integrated based upon its semantic meanings. Informally, it sounds like a virtual data warehousing technique without any physical data conversion required in advance. Ontology is used to describe the semantics of global information content, and semantic enhancement is achieved by mapping the local metadata onto the ontology. For better system reliability, a unique mechanism is introduced to perform appropriate adjustments upon detecting environmental changes.