2013 Volume 4 Issue 2 Pages 172-182
We generally analyze data measured at equal time intervals using fast Fourier transforms (FFTs). However, when for some reason we can only obtain unequally spaced measurement points, then the data become non-uniform and cannot be analyzed directly using FFTs. In this paper, the non-uniform trapezoidal integral discrete Fourier transform (TINUDFT) is proposed to replace the FFT, and is implemented in non-harmonic analysis (NHA) to develop non-uniform NHA (NUNHA) for non-uniform data. NUNHA improved the SNR by over 70 dB compared with the simple non-uniform sinusoidal signal frequency. NUNHA's effectiveness for swept source optical coherence tomography (SS-OCT) is shown by simulation.