Abstracl-A novel framework for the optimization of the transmission/interference strategy of secondary users in cognitive networks is proposed. The framework models the dynam ics of the transmission protocols implemented by the wireless terminals, as well as the dynamics of important characteristics of the transmission circuitry such as on-board heat generation and temperature. By generating interference to the primary user's receiver, and cooperating with primary user's packet delivery, the secondary user biases the statistics of the primary user's packet service. As the activation of the transmission circuitry generates heat, the activity of the secondary user impacts the temporal evolution of the primary user's internal temperature. The framework, based Markov Decision Process theory, enables the optimization of the transmission strategy of the secondary user under constraints on the primary user's average throughput.Numerical results show that the activity of the secondary user is beneficial to the hardware state even when the secondary user only aims at maximizing its achieved throughput.