Graph-based spectrum sharing for multiuser OFDM Cognitive Radio Networks

Spectrum sharing in OFDM-based Cognitive Radio Networks (CRN) is investigated by using graph theory. We consider the networks in which a set of secondary links (SLs) make underlay access to support single-hop ad hoc transmission. Spectrum sharing must be carried out in SLs so that no excessive interference is caused to links of the primary network and maximizes the total achievable rates of SLs. We divide the spectrum sharing into two stages: coarse-scale interference mitigation and SINR-based channel mapping. Interference mitigation is done through constructing weighted undirected interference graph and applying k-max cut method. Then we perform channel-cluster mapping, which takes instantaneous channel quality into consider, on the basis of interference mitigation. Low-complexity heuristic algorithms are proposed to effectively solve the problems of each stage. Numerical simulations for various scenarios are executed to show the superior performance of the proposed schemes.