Dynamic Spectrum Leasing (DSL) in Dynamic Channels

Dynamic Spectrum Leasing (DSL) was recently proposed in [1]–[4] as a new way to achieve dynamic spectrum sharing (DSS). Unlike previously considered dynamic spectrum access (DSA) proposals, DSL allows for the spectrum owner, called the primary user, to dynamically adjust the amount of interference it is willing to tolerate from secondary users. In response, the secondary users update their transmit powers to maximize a suitably chosen reward function. In previous work, it has been shown that the best response adaptations will converge to a unique Nash equilibrium under the assumptions of a quasistatic channel conditions and for a fixed number of secondary users. In this paper, we investigate the convergence and equilibrium performance of the proposed DSL-game in the presence of slow time-varying fading and time-varying secondary system size. Our results show that while the DSL best response adaptation algorithm is reasonably robust against these dynamics, there is a trade-off between performance and the CSI update rate.