In this paper we are interested in the early development phases of a distributed software system, especially in the case of obtaining a system design from its global requirement specification. This is an important step in the development of complex systems and their automated derivation. The Model-Driven Engineering (MDE) approach allows the designers to model their systems at different abstraction levels. Such approach provides automatic model transformations to incrementally refine abstract models into concrete ones. This paper presents a MDE approach to derive the behavior of a given system component from its global requirements. The requirements are specified using UML activity diagram extended with collaborations. The derived behavior of the components is in the form of distributed UML state machines. The suggested approach is based on the definition of an appropriate source meta-model (requirements meta-model) and the definition of a target design meta-model. A set of model transformation rules has been designed to govern the derivation process. A real application of telediagnosis in neuroscience has been developed using this approach.