PreDA: Preference-based double auction for spectrum allocation in heterogeneous DSA networks

Abstract

In this paper, we propose PreDA– a preference-based truthful double auction for dynamic spectrum access (DSA) networks where multiple heterogeneous spectrum bands are sold by the primary users and bought by the secondary users. Unlike existing double auctions, we not only consider channels’ heterogeneity and multi-bids from buyers, but also consider buyers’ preferences for the channels. We use the signal to interference and noise ratio (SINR) as a metric for the preference; channels with higher SINR are preferred and hence bids are more compared to the bids for channels that offer lower SINRs. In order to maximize the spatial and temporal reuse of channels, we use the concept of virtual grouping of conflict-free buyers. Virtual groups allow us to transform multi-unit bids to single-unit bids. We propose a novel winner determination and pricing mechanism to allocate the unused spectrum bands to the most appropriate buyers. We prove that the proposed double auction achieves the three economic properties of truthfulness, individual rationality, and ex-post budget balance. We validate PreDA through simulation experiments and show its performance in terms of the number of allocated bands, utilization, revenue, and fairness.