Wideband photonic A/D conversion using 2-D spatial oversampling and spectral noise shaping

The analog-to-digital (A/D) interface is generally considered to be the most critical part of any signal acquisition and processing system. Because of the difficulty in achieving high-resolution and high-speed A/D converters, this A/D interface has been and continues to be a barrier to the realization of high-speed, high-throughput system. Recently, there has been a renewed interest in new and innovative approaches to A/D conversion, with a significant emphasis on photonic techniques. Interleaving is a common approach applied to high-speed photonic A/D conversion which reduces the wide-bandwidth input signal to one which can be converted using conventional high-speed A/D converters. The high-speed sampled input is interleaved to N individual channels with each channel operating at 1/N of the sampling rate. These channelization techniques are known to suffer from performance degradations due to channel-to-channel mismatch. Within the electronic A/D converter community, temporal oversampling and spectral noise shaping have become common practice in high-fidelity audio applications. Here, a low-resolution quantizer is embedded in a feedback architecture in an effort to reduce the quantization noise through spectral noise shaping. A large error associated with a single sample is diffused over many subsequent samples and then linear filtering techniques are applied to remove the spectrally-shaped noise thereby improving the overall SNR of the converter. The approach to wideband photonic A/D conversion described here leverages the 2-D nature of an optical architecture to extend the concept of spectral noise shaping to include 2-D spatial noise shaping. The proposed approach uses a mode-locked laser to generate the optical sampling pulses, an interferometer to modulate the electronic analog signal onto the optical pulses, and a 2-D smart pixel hardware implementation of a distributed error diffusion neural network.