Tuple centres are a coordination model that is key in the development of multi-agent systems and pervasive computing infrastructures, and is the basis of TuCSoN middleware. Tuple centres do not dictate a specific coordination model, but can rather be programmed using any language (though currently only ReSpecT language is supported). The goal of this thesis is to re-engineer tuple centres so as to support effective and efficient injection of new coordination models into tuple centre.
