Smirnov, M.M.Smirnov2022-03-042022-03-042010https://publica.fraunhofer.de/handle/publica/224255The purpose of this chapter is to report very recent results on policy-based selfmanagement towards self-adaptive and Cognitive Radio Systems (CRS). The focus will be not on self-* properties of cognitive radio systems but on selflearning policy and cognition; the chapter shows that self-* properties appear as the result of cognition. The chapter will elaborate on radio domain-specific cognition model and on the role of context and policy in that model; specific emphasis will be given to policy multiplexing (this is where a policy disappears becoming embedded in to the fabric of other functionalities), process correctness (cross disciplinary approach based on information modelling and theory, control theory) as a part of assessment of CRS (like a "Turing test" for cognitive radio), and emerging design methodology of CRS.en004book article