We propose in this paper a radio resource allocation scheme in the context of Cloud-based Radio Access Networks (C-RAN) run by a single operator. We specifically leverage bandwidth calendaring, a technique that allows shifting bulk data transfers, typically of large size with less stringent real-time constraints, to future epochs when the network is less congested. In particular, we propose an auction-based framework for bandwidth calendaring, where the C-RAN operator, as the spectrum auctioneer, runs an auction with its users, with the aim of maximizing its revenue. The auction-based mechanism takes as input the set of the users’ bids and outputs the calendaring and pricing decisions. We first formulate the calendaring problem using Integer Linear Programming (ILP) and the pricing problem using the Vickrey Clarke Groves (VCG) pricing scheme. We further make use of the Bayesian settings to compute the optimal revenue. Due to the exponential time induced by the NP-hardness of the ILP formulation, we propose an effective approach that satisfies desired auction properties such as individual rationality and truthfulness while achieving a sub-optimal revenue, in polynomial time. We explicitly evaluate the impact of mobile systems features such as spatial frequency re-use and interference among mobile users, and study their impact on the overall system’s performance. Extensive simulations, conducted in representative network scenarios, demonstrate the effectiveness of our proposal in improving the performance of C-RAN scheduling.