In cognitive radio networks, service providers with spare channels can sell the spectrum to those in need of them. The redistribution of the spectrum among service providers reduces the waste of spare spectrum, therefore enhances the spectrum utilization. It also provides the service provider more revenue from the sail of spectrum. Traditional method of spectrum trading mainly based on double auction, which requires an auctioneer, is a trustworthy third-party authority, to centrally enforce a certain spectrum allocation policy. In this chapter, we take a different and new perspective, proposing to use matching as an alternative tool to realize spectrum trading in a distributed way for a free market, which consists of only buyers and sellers, without a trustworthy third-party authority. In this chapter, we will first give a brief introduction of the whole chapter in the first section and then present the fundamentals of the matching theory in the second section. In the third section, matching theory is leveraged in spectrum trading among service providers to decide the spectrum allocation and trading price, the distinctive challenge of spectrum matching compared with conventional matching is analyzed, and a two-stage distributed algorithm is proposed to solve the spectrum matching problem. In the fourth section, we considered a more general case, where multiple channels can be bought by the same service provider, and the spectrum matching algorithm for combinatorial spectrum trading is proposed to enable the spectrum allocation. For both algorithms, the proposed algorithm can achieve a Nash-stable matching, and the simulation shows that the proposed algorithms achieve good performance compared with centralized schemes.
CITATION STYLE
Zhang, J., Jiang, L., Cai, H., & Chen, Y. (2019). Many-to-many matching for distributed spectrum trading. In Handbook of Cognitive Radio (Vol. 2–3, pp. 1379–1412). Springer Singapore. https://doi.org/10.1007/978-981-10-1394-2_40
Mendeley helps you to discover research relevant for your work.