JSSC 2020第2期RF & Wireless28nmNeural Network Accelerator

An Event-Driven Quasi-Level-Crossing Delta Modulator Based on Residue Quantizati

提出一种用于物联网无线网络的事件驱动准电平交叉Δ调制器ADC，具有自适应分辨率。

28nm CMOS, 1.42MHz信号带宽, 53dB峰值SNDR, 205µW功耗, 0.0126mm²面积

事件驱动准电平交叉Δ调制器自适应分辨率物联网

▸采用4位异步SAR子ADC量化残差电压信号

▸通过数字多级比较窗口实现自适应分辨率

▸全局信号依赖的平均采样率实现数据压缩

Abstract

This article introduces a digitally intensive event- driven quasi-level-crossing (quasi-LC) delta-modulator analog- to-digital converter (ADC) with adaptive resolution (AR) for Internet of Things (IoT) wireless networks, in which minimiz- ing the average sampling rate for sparse input signals can signiﬁcantly reduce the power consumed in data transmission, processing, and storage. The proposed AR quasi-LC delta mod- ulator quantizes the residue voltage signal with a 4-bit asynchro- nous successi