A low power prediction SAR ADC integrated with DPCM data compression feature for WCE application

Abstract

The main challenge in wireless capsule endoscopic system is to reduce the area and power consumption while maintaining acceptable image reconstruction. In this paper, a low power successive approximation register (SAR) analog-to-digital converter (ADC) with a prediction scheme integrated with a differential pulse coded modulation (DPCM) data compression feature for wireless capsule endoscopy (WCE) application is presented. The analysis of endoscopy images shows that neighbor pixels have strong correlations. By predicting the value of the selected pixel, A/D conversions steps for MSBs are significantly reduced as well as power consumption. Considering that DPCM also makes use of the likelihood of neighbor pixels and its low computation and implementation complexity, it is merged into the SAR ADC to perform data compression during A/D conversion. The analog input of the ADC is digitized and compressed simultaneously, which results in a seamless connection between A/D data compression and data compression. Data compression algorithm is integrated on sensor node which achieves saving of area and power consumption. Simulation results show the proposed ADC performs strongly with a compression ratio of 62% and a high average PSNR of 38.9dB. It consumes only 10% of the switching energy of the conventional SAR ADC.