Fast Fourier Transform (FFT) Using Flash Arrays for Noise Signal Processing

Abstract

We propose an In-Memory Computing (IMC) scheme to implement Fast Fourier Transform (FFT) for noise signal processing using flash memory. By taking advantage of the algorithms of complex number calculation and butterfly operation in FFT, the analyses of low-frequency noise (LFN) and random telegraphy noise (RTN) signals using flash-based IMC were demonstrated successfully, showing the frequency properties of 1/ <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">${f}$ </tex-math></inline-formula> and 1/ <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">${f}$ </tex-math></inline-formula> <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</sup> , respectively, and matching well with the Matlab numerical calculation. An energy efficiency of 2.7 TOPS/W has been achieved with a dynamic power of 27.7 mW. Our proposed IMC scheme provides an efficient way to enable FFT using a hardware solution and is extendable for other important applications such as image processing and voice recognition.