Activity-Scaling SAR with Direct Hybrid Encoding for Signed Expressions for AIoT Applications

Designing an AIoT system with low standby power and high efficiency has become increasingly challenging. The AIoT system is an IoT device with artificial intelligence. A typical AIoT system is an always-on portable ECG monitoring system with AI algorithm to detect irregular event. ADC is the bottleneck of the current AIoT systems as it bridge the gap between analog world and digital computation. To address the challenge, this thesis presents a SAR ADCs with two modes, activityscaling and direct hybrid Encoding for signed expressions. In the activity-scaling mode, the proposed ADC can finish the conversion in just one cycle in the optimal case compared with N cycles for typical SAR. In the direct hybrid encoding for signed expressions (HESE) mode, it directly provides hybrid encoding for signed expressions which paves the way for high efficient digital inference. The proposed ADC has two thresholds. The activity-scaling mode has an initial guess and takes two steps per cycle to approach the sampled input until overshoot. After that, it performs a ternary search to the LSB. The direct hybrid encoding for signed expressions mode places one of the thresholds at the normal binary conversion threshold and the other for two bits look ahead to produce one-pass encoding. A proof-of-concept SAR ADC has been designed in 65nm CMOS technology.

Name thesis:degree_name

Master

Department dc:contributor.department

Massachusetts Institute of Technology. Department of Electrical Engineering and Computer Science

Grantor dc:publisher

Massachusetts Institute of Technology

Year dc:date.issued

2021

Author dc:creator

• Chen, Ruicong

Advisors dc:contributor.advisor

• Anantha P. Chandrakasan
• Hae-Seung Lee

Statement dc:rights

• In Copyright - Educational Use Permitted
• Copyright MIT

Licence dc:rights.uri

• http://rightsstatements.org/page/InC-EDU/1.0/

Handle dc:identifier.uri

https://hdl.handle.net/1721.1/139133

OAI identifier oai:identifier

oai:dspace.mit.edu:1721.1/139133