← 返回 JSSC 论文列表JSSC 2024第6期RF & Wireless40nmPhased Array
A Multimode 157 µW 4-Channel 80 dBA-SNDR Speech Recognition Frontend With Direct
一种多模式157µW语音识别前端系统，通过创新贪婪阻塞矩阵和自动方向误差补偿提升性能。
40nm CMOS, 157µW/72µW高低功耗模式, 0.94mm²面积
语音识别波束成形贪婪阻塞矩阵方向误差补偿低功耗
▸采用贪婪阻塞矩阵(GBM)降低DSP功耗
▸自动方向误差补偿(DOA)技术
▸多模式混合ADC适应信号条件
▸多模式波束成形利用高信噪比降低总功耗54%
Abstract
This work introduces a speech recognition fron-
tend system that solves the problems of conventional adaptive
beamforming (ABF) with: 1) low digital signal processing (DSP)
power consumption (3× lower than state-of-the-art ABF) thanks
to an innovative greedy blocking matrix (GBM) employing sim-
ple calculations; 2) automatic direction-of-arrival (DOA) error
compensation with direction tracking delay-and-sum beamformer
aided by the GBM; 3) a multimode hybrid analog-to-digital
converter (ADC) adap