An Improved Dynamic Spectrum Access Scheme in Cognitive Networks

Abstract

Dynamic spectrum Access is the key technology in cognitive networks. And spectrum auction has been considered as an effective way for dynamic spectrum Access. In this paper, we propose a spectrum allocation scheme with the goal to maximize the revenue of primary user (PU) and then try to minimize the interference among secondary users (SUs). There are two steps during the scheme: auction and allocation. During the auction phase, winner determination problem (WDP) is modelled as a knapsack problem, and then Genetic Algorithm is selected to solve it in polynomial time. After that, a payoff function is proposed to ensure that all the SUs can get non-negative utilities only with their truthful information. During the allocation phase, we try to minimize the interference while allocating the released channels to all the winners. Finally, simulation experiments show that the proposed auction mechanism has a better performance in solving the problem of WDP.