Achieving control architecture flexibility in multi-vendor networking architectures

Community networking infrastructures are expected to share with other evolving multimedia networking architectures a number of common characteristics. Notable among these are: 1) They will involve a number of key enabling technologies (e.g., processor power, memory capacity and speed, interconnection bandwidth). 2) They will largely consist of a number of architectural components with open intercomponent interfaces envisioned, and more often than not with multiple vendors involved. 3) The envisioned applications and services typically will require fairly sophisticated control functionalities to effect inter-component interactions and coordination and to provide the efficient resource usage needed for economics. For these and other multimedia networking architectures to be successful, a great deal of architectural flexibility will be needed, to accommodate the inevitable architectural redirections and evolutions that occur as the underlying technologies advance, and as the applications and services that will provide real value (and, that can be offered economically!) evolve and become better understood. In particular, flexibility in the underlying control architecture will greatly increase the chances of success of the applications and services, and of the associated delivery mechanisms, i.e., the underlying architecture infrastructures. This paper focuses on identifying basic principles for achieving the flexibility, evolvability, and controllability that will be needed