Towards a layer-less network architecture — A case from Wireless Sensor Networks

Ad hoc and Wireless Sensor Networks (WSNs) form a challenging domain for existing network protocols and paradigms to fit into. The traditional models conceived for wired networking and adapted for wireless and mobile environments provide an inefficient and ungraceful outcome when exercised against highly dynamic and highly mobile nature of ad hoc wireless sensor networks. Efforts are being made to tweak existing (and sometimes create from scratch) principles and models that can provide a concrete framework for such newer paradigms. The momentum of the traditional layered architecture carried over from the wired networks to infrastructure based wireless networks pushes the same approaches as the de facto standard for the new ad hoc WSNs. We argue that such an approach conflicts with the long term prospects of the new paradigms. Using a network architecture designed for static nodes, albeit with tweaks (such as cross-layering), to address a completely different model of communications with modalities beyond the scope of both wired and infrastructure based wireless networks is not the right path to follow. We present a new way of thinking about the network abstractions, away from the hierarchical layers, towards network functionality as layer-less services. Further, we propose the approaches that can be utilized to standardize this approach in the long term.