摘要： The latest trends in spectrum trading allow secondary users (SUs) to employ hybrid access models and exploit bandwidth employing either opportunistic spectrum access or exclusive spectrum access of vacant frequency bands (FBs) leased for exclusive usage. In this paper, the spectrum trading problem is addressed in an operational framework, where a primary spectrum owner (PSO) allocates specific FBs to a number of primary users (PUs) and at the same time allows SUs belonging to multiple service classes to operate opportunistically in a part of the spectrum primarily allocated to the PU operation. By offering exclusive bandwidth-price contracts to SUs in FBs, which are excluded from the PU access, the PSO expects additional revenue. In this framework, the PSO aims at maximizing its revenue through optimal contract design that considers the multiple SU service classes, about which the PSO may have at its disposal either complete or incomplete information. In the first case, it is proven that only one type of contract destined for the highest SU service class is profitable to the PSO. In the second case, two heuristic algorithms are proposed for the optimal contract design corresponding to offering a single contract or multiple contracts to various service classes. The simulation results for the complete and incomplete information cases have been examined with regard to the SU service classes and the number and transmission specifications of the PUs. The two proposed algorithms manage to address the incomplete information case and be close to the optimal values obtained when complete information is available. Finally, the social welfare, as the aggregate expected utility of all parties involved, is examined.