Combinatorial spectrum auction with multiple heterogeneous sellers in cognitive radio networks

Spectrum auction has been considered as an economically incentive way to motivate both primary spectrum owners (POs) and secondary users (SUs) to participate in dynamic spectrum access (DSA). In this paper, we propose a new combinatorial spectrum auction framework for the scenarios that each PO has multiple channels to sell and each SU demands multiple channels. Moreover, the heterogeneity in terms of POs' channel bandwidths and SUs' demands is also considered. The winner determination problem (WDP) in the proposed auction framework can be formulated as a multiple multidimensional knapsack problem (MMKP) and both upper bound and an approximation algorithm with polynomial time are developed. A tailored pricing mechanism is adopted in the payment design to ensure truthfulness and individual rationality. Numerical results show that our proposed auction algorithm can improve the spectrum allocation efficiency compared to counterparts.