CODIPHY: Composing on-demand intelligent physical layers

In this paper we present CODIPHY or Composing On-Demand In-telligent Physical Layers that aims to solve two fundamental prob-lems in practical cognitive radio networks: Collaboration between two radio physical layers (PHY) with varying capabilities to agree on a common communication protocol and provide a method to compose a functioning software defined radio (SDR) from a set of pre-compiled libraries. Both solutions use an ontology based de-scription of the internal structure of the radio subsystems and use the high-level dataflow represented by the ontology to target het-erogeneous platforms. CODIPHY isolates the various domains of radio engineering but still allows them to share domain knowledge to achieve a common goal of radio adaptation. Automating this process through declarative specification and collaborative learn-ing is the goal of this paper. We present a generic methodology to facilitate the concept of CODIPHY and present examples from the radio PHY domain