Combinatorial auction based spectrum allocation under heterogeneous supply and demand

In this paper, we apply the combinatorial auction scheme to solve the spectrum allocation problem under heterogeneous supply and demand in cognitive radio networks. The heterogeneity of spectrum is embodied via exploiting multiple attributes, based on which a valuation function is devised to evaluate the preference of an SU over a spectrum band. We first propose an auction scheme consisting of a greedy-like winner determination algorithm and a critical value based discriminatory pricing policy. We then extend the proposed auction scheme to a more challenging scenario by considering spectrum sharing among SUs. Theoretical analysis demonstrates that our auction schemes achieve individual rational, budget balance, value-truthfulness of SUs, and weak value-truthfulness of PUs. Our simulation results verify the advantage of combinatorial auction, the functionality of spectrum sharing and the economic robustness of our auction schemes.