The acquisition of reliable spectrum data is a prerequisite for spectrum sharing. However, current spectrum sensing faces serious Spectrum Sensing Data Falsification (SSDF) attacks. To mitigate this challenging issue, we propose a reward-based joint spectrum sensing and channel allocation scheme, where secondary users are required to submit estimated revenue values along with their sensing results to the fusion center, the center then allocates channels based on the revenue results. The channel allocation problem is formulated as an optimization problem to maximize revenue, which is then solved using the Hungarian algorithm. Simulation results show that the obtained revenue of the secondary user with normal sensing is significantly higher than that of when launching SSDF attacks, demonstrating the effectiveness of the proposed scheme in mitigating SSDF attacks. Moreover, the proposed scheme outperforms existing spectrum sensing and channel allocation schemes when the number of trusted users is small.