A Compressive Neural Telemetry With Address-Event-μ-Packet Encoding and Duty-Cycling IR-UWB Transmitter

A Compressive Neural Telemetry With Address-Event-μ-Packet Encoding and Duty-Cycling IR-UWB Transmitter

Abstract

DOI: 10.1109/TMTT.2026.3698352

IEEEXplore: https://ieeexplore.ieee.org/document/11554176

Phrase Summary:

Networked wireless telemetry for 384-sensor-pixel neural data acquisition.

Sentence Summary:

This article presents a networked wireless telemetry for 384-sensor-pixel neural data acquisition with high throughput, low power consumption, and meter-scale communication distance.

Paragraph Summary:

This article presents a networked wireless telemetry for 384-sensor-pixel neural data acquisition with high throughput, low power consumption, and meter-scale range.  The challenging trade-off between throughput-power, transmission range and signal fidelity is addressed with a novel event-based Address-Event-µ-Packet (AEµP) encoder with differential binary phase shift keying (DBPSK) modulation at a 125 MHz symbol rate that compresses data and minimizes protocol overhead and memory access to achieve a compression ratio of 11.52x.  The communication link for the encoder is a duty-cycling impulse-radio ultra-wideband (IR-UWB) transmitter with a proven 6.6-m communication range.  The telemetry with a compact active area of only 0.12 mm2 consumes about 1 mW and is fabricated in 65 nm CMOS.