Light Field Super-Resolution (LFSR) endeavors to reconstruct high-resolution (HR) light field images from their low-resolution (LR) counterparts by capitalizing on multi-view image information. This process not only enables a more efficient restoration of high-frequency details but also preserves the geometric structure of the scene. Nevertheless, prevailing methods encounter difficulties in capturing the long-range spatial and angular dependencies inherent in light field data, as well as their high-frequency spectral characteristics. Moreover, the availability of high-quality paired training data for real-world scenarios remains limited. To address these challenges, this paper presents DFFIT (Dual learning and Fractional Fourier Image Transformer), a novel LFSR framework that seamlessly integrates frequency-domain analysis with a dual-learning strategy grounded in degradation modeling. We introduce the Fractional Fourier Image Transformer (FrIT), which ingeniously combines the fractional Fourier transform (FrFT) with Transformer-based long-range dependency modeling. This integration effectively captures frequency-specific features while guaranteeing cross-view consistency. Additionally, our dual-learning framework generates a variety of LR training samples by emulating real-world degradation processes, thereby narrowing the domain gap between synthetic and real-world data. Experimental results verify that the proposed method exhibits remarkable performance in enhancing the resolution of light field images.