Radio Resource Allocation and Pricing: Auction-Based Design and Applications

The "unlimited" performance and machine-centric architecture visions for future wireless networks transform the fundamental task of allocating radio resources into a complex optimization problem that is not quickly solvable. Inspired by the increasing intelligence of connected machines, and the prosperity of auctions as efficient allocation mechanisms in the economic sector, this paper provides an alternative perspective to the problem of optimal spectrum assignment for the fifth generation (5G) of wireless networks. In a systematic approach to deal with this problem, an efficacious allocation mechanism is characterized by six axioms: incentive compatibility, individual rationality, fairness, efficiency, revenue maximization, and computational manageability. The first three are incorporated into the allocation mechanism through a nonlinear spectrum pricing. By inducing incentive compatibility through these prices, revelation of the true valuations becomes the Nash Equilibrium and puts the mechanism in the class of revelation mechanisms. The latter fact triggers the realization of the last three axioms, whereby an optimization problem is formed to find the optimal mechanism in the class of revelation mechanisms, which, by the virtue of the revelation principle, is the optimal mechanism among all auction classes. Further, it is shown that the proposed mechanism is highly scalable, as the solution to the optimization problem is obtained by root-finding operations and solving almost linear system of equations. These properties make the proposed resource allocation mechanism an ideal candidate for deployment in 5G networks.