A multi-winner cognitive spectrum auction framework with collusion-resistant mechanisms

Dynamic spectrum access, enabled by cognitive radio technologies, has become a promising approach to improve efficiency in spectrum utilization, and the spectrum auction is one approach in which unlicensed wireless users lease some unused bands from spectrum license holders. However, spectrum auctions are different from traditional auctions studied by economists, because spectrum resources are interference-limited rather than quantity-limited, and it is possible to award one band to multiple secondary users with negligible mutual interference. Due to its special feature, the multi-winner auction is a new concept posing new challenges in the existing auction mechanisms such as the Vickery-Clarke-Groves (VCG) mechanism. Although widely employed in other auctions, the VCG mechanism does have serious drawbacks when applied to the multi-winner auction, such as unsatisfactory revenue and vulnerability to collusive attacks. Therefore, in this paper, we propose a multi-winner spectrum auction framework, and develop suitable mechanisms for this kind of auction. In specific, the mechanism awards the bands in such a way that the spectrum efficiency is maximized, and determines prices based on the Nash bargaining solution to improve revenue and prevent collusion. We further analyze that secondary users do not have incentives to manipulate information about mutual interference which is essential to the auction. Finally, simulation results are presented to evaluate our proposed auction mechanisms.