This paper investigates the use of advanced communication services in grid environments. Such services can include augmented communication semantics (e.g., filtering), collective operations, content-based and policy-based routing, and managing communication scope to manage feasibility. These services could be implemented and deployed in variety of ways, such as a traditional network of servers, or as middleware forwarding and routing layer, or even in an active network. In any of these approaches, topology- awareness can play major role in their performance and scalability. As case study, we demonstrate here communication service to support time management in distributed simulations that is managed using grid computing toolkit. We also present emulation and simulation results to demonstrate the scalability that topology-awareness enables for services such as time management. Since the design space for communication services offers so many possibilities and alternatives, we argue for the definition of proper high-level models and APIs such that the underlying implementations and scope of deployment can be developed and improved with minimal impact on applications.