Fractal: A Mobile Code Based Framework for Dynamic Application Protocol Adaptation in Pervasive Computing

The rapid growth of heterogeneous devices and diverse networks in our daily life, makes it is very difficult, if not impossible, to build a one-size-fits-all application or protocol, which can run well in such a dynamic environment. Adaptation has been considered as a general approach to address the mismatch problem between clients and servers; however, we envision that the missing part, which is also a big challenge, is how to inject and deploy adaptation functionality into the environment. In this paper we propose a novel application level protocol adaptation framework, Fractal, which uses the mobile code technology for protocol adaptation and leverages existing content distribution networks (CDN) for protocol adaptors (mobile codes) deployment. To the best of our knowledge, Fractal is the first application level protocol adaptation framework that considers the real deployment problem using mobile code and content distribution networks. To evaluate the proposed framework, we implement an adaptive communication optimization protocol by dynamically selecting four communication protocols, including Direct sending, Gzip, Bitmap, and Vary-sized blocking. In the comparison with the static and centralized protocol adaptation approaches, evaluation shows good results on both the client side and server side. For some clients, the total communication overhead reduces 41% compared with no protocol adaptation mechanism, and 14% compared with the static protocol adaptation approach.