Dynamic Matching Based Distributed Spectrum Trading in Multi-Radio Multi-Channel CRNs

Jingyi Wang,Wenbo Ding,Yuanxiong Guo,Chi Zhang,Miao Pan,Jian Song

DOI: https://doi.org/10.1109/glocom.2016.7841761

2016-01-01

Abstract:Spectrum trading not only improves spectrum utilization but also benefits both secondary users (SUs) with more accessing opportunities and primary users (PUs) with monetary gains. Although existing centralized designs consider the special features of spectrum trading (e.g., frequency reuse, interference mitigation, multi-radio multi- channel transmissions, etc.), they have to deploy new infrastructure, deal with extra control overhead, have scalability issues, and may miss many instantaneous opportunities. To address those issues, in this paper, we propose a novel dynamic matching based distributed spectrum trading (DMDST) scheme in multi-radio multi- channel cognitive radio (CR) networks. We employ conflict graph to characterize interference relationship among SUs with multiple CR radios, and formulate the centralized PUs' revenue maximization problem under multiple constrains. In view of the NP- hardness of solving the problem and no existence of centralized entity, we develop the DMDST algorithms based on conflict graph observed by PUs, solve the problem via dynamic matching with evolving preferences, and prove its stability. Through extensive simulations, we show that the results of proposed DMDST algorithm is close to the optimal one and outperforms other distributed algorithms without considering spectrum reuse.