A MUDdy water filling algorithm for cochannel power allocation for cognitive coexistence radios

We consider the problem of over/underlay communications in the presence of an already established network of rate-adaptive nodes. Existing interference channel models are insufficient for describing such a scenario, especially when explicit coordination between the interfering networks is prohibited. Here, our focus is in eliciting models and deriving cochannel allocation algorithms for the case when nodes on the infringing network are capable of performing on-the-fly optimizations (cognition) and are equipped with receivers enabling multi-user detection (MUD). Two strategies are studied: selfish coexistence which is indifferent to the performance of the existing network, and considerate coexistence that factors performance of the existing network into its own decision making.