Title | A 148-nW Reconfigurable Event-Driven Intelligent Wake-Up System for AIoT Nodes Using an Asynchronous Pulse-Based Feature Extractor and a Convolutional Neural Network |
Authors | Wang, Zhixuan Liu, Ying Zhou, Peng Tan, Zhichao Fan, Haitao Zhang, Yihan Shen, Linxiao Ru, Jiayoon Wang, Yangyuan Ye, Le Huang, Ru |
Affiliation | Peking Univ, Sch Integrated Circuit, Beijing Lab Future IC Technol & Sci, Beijing, Peoples R China Peking Univ, Adv Inst Informat Technol, Hangzhou 311200, Peoples R China Nano Core Co Ltd, Hangzhou 311200, Peoples R China Zhejiang Univ, Coll Informat Sci & Elect Engn, Hangzhou 310058, Peoples R China |
Keywords | LEVEL-CROSSING ADC FRONT-END CHIP DESIGN COMPARATOR PROCESSOR CLOCK |
Issue Date | Nov-2021 |
Publisher | IEEE JOURNAL OF SOLID-STATE CIRCUITS |
Abstract | This article presents a 148-nW always-on wake-up system that drastically reduces the system power consumption of Internet of Things (IoT) sensor nodes while oftentimes operating in random-sparse-event (RSE) scenarios. To significantly reduce the long-term average (LTA) power consumption and realize multiapplication and intelligent event detection, three techniques are proposed: 1) In a three-stage pipelined event-driven architecture, a frame generator and a convolutional neural network intelligent inference engine (CNN IIE) in stage III are event-driven by the preliminary detectors in stage II, and the detectors are triggered by a level-crossing (LC) analog-to-digital converter (ADC), i.e., stage I, dramatically reducing the overall power consumption. 2) The clock-free pulse-based instant rate of change (IROC) feature extractor directly processes the asynchronous pulses of the LC-ADC outputs in the temporal domain instead of utilizing a conventional power-hungry frequency-domain method. 3) A reconfigurable IROC, the frame generator, and the CNN IIE provide adaptive intelligence for various IoT events, enhancing the accuracy of multipurpose detection with ultralow power. We demonstrate two artificial intelligence IoT (AIoT) applications at 0.6-V V-DD. For electrocardiogram (ECG) recognition, one example works at a typical event rate (ER) of similar to 4800/h, with an active power of 1.68 mu W and a precision of up to 99%; the other is used for keyword spotting (KWS), where the chip achieves 378 nW at similar to 720/h ER and 94% accuracy. The LTA power is bounded to 148 nW, while the event-driven chip is on call and waiting for events; this chip dominates the AIoT device battery life in RSE scenarios. |
URI | http://hdl.handle.net/20.500.11897/628822 |
ISSN | 0018-9200 |
DOI | 10.1109/JSSC.2021.3113257 |
Indexed | EI SCI(E) |
Appears in Collections: | 待认领 |