Sector-based policy generation and enforcement for cognitive radios

Abstract

The policy-managed radio model offers a way to force cognitive radios to comply with transmission restrictions and regulations. Policies that implement radio regulations are written in a policy language, and used by the policy engine to approve or deny transmission parameters. Since decision speed is one of the primary design requirements of cognitive radios, policy engines and reasoners can easily bottleneck the radio's operation. In this paper, we propose a policy generation and enforcement model for policy-managed cognitive radio systems designed with speed as its top priority. The model is based on the idea of splitting a resource into minimum-sized usable sectors. Policy is preprocessed and compiled for each of the sectors, thus creating a single static policy for each sector of the resource. This model drastically decreases runtime policy comparisons, thus speeding up the policy engine's operation. Another effect of this design is a simplified policy language model. We discuss design considerations and challenges for policy languages and policy engines, ways in which a slow policy engine can bottleneck a cognitive radio's operation, and assert that operational speed should be the top design priority when considering policy engines and reasoners. In addition, we analyze the on-disk storage required by this model, and suggest possible optimizations.