Stage-Dependent DSP Operation Range Clipping-Induced Bit Resolution Reductions of Full Parallel 64-Point FFTs Incorporated in FPGA-Based Optical OFDM Receivers

Abstract

The fast Fourier transform (FFT) is the fundamental algorithm at the heart of an optical OFDM (OOFDM) transceiver. The high digital signal processing (DSP) complexity has become one of the most significant obstacles to experimentally demonstrating real-time high-capacity OOFDM transceivers. In this paper, stage-dependent clipping of FFT DSP operation dynamic range is proposed and extensively explored, for the first time, based on which an improved stage-dependent minimum bit resolution map is numerically identified by taking into account the DSP operation dynamic range-clipping and precision of FFT DSP operations. The validity and high accuracy of the identified minimum bit resolution map is experimentally verified in 25 km SSMF OOFDM transmission systems based on intensity modulation and direct detection. Experimental results show that, compared to a previously reported FFT stage-dependent bit resolution map including the FFT DSP operation precision only, the improved minimum bit resolution map offers significant bit resolution reductions of up to 3-bits for full-parallel pipelined 64-point FFT architectures. As a direct result, an approximately 30% reduction in FPGA logic resource usage is achievable compared to the spiral FPGA design.