A Neuromorphic RF Power Sensor for Inductive Coupling IoT Systems

Read full text Summarise with VIRAT
Authors: Rafael Cantalice, Gutemberg Goncalves dos Santos Junior, Marcos R. A. Morais, Antonio M. N. Lima, Ângelo Perkusich, Pietro Maris Ferreira
Cited by: 0
References: 26

Summary

No summary yet. VIRAT can read this paper and write a plain-language summary of its aim, methods, key findings and limitations.

Summarise with VIRAT Free to try with your Science Catalogs account.

Abstract

This work present a design and validation of a neuromorphic RF power sensor for inductively coupled IoT systems (e.g., RFID tags and implantable devices) that operates with ultra-low power consumption and integrates seamlessly with Spiking Neural Networks (SNNs) for event-driven processing.The proposed sensor employs a Pulse Width Modulation (PWM) mechanism to encode the received RF power into a duty cycle. This PWM signal drives a charge pump, generating an analog voltage proportional to the RF input power. The analog voltage is converted into spike trains using a voltage-to-spike converter based on a relaxation oscillator. The sensor was designed using a 180 nm CMOS process, with key components including low-power comparators, a charge pump, and a relaxation oscillator. Electrical simulations were conducted to evaluate the sensor’s performance under process, voltage, and temperature (PVT) variations.The sensor operates at 1 V supply with a current consumption of 1 µA. It achieves a frequency range of approximately 17.4 kHz for antenna input voltages varying between 3 V and 8 V, with theoretical performance closely matching simulation results. The architecture demonstrated stable operation under PVT variations, achieving ±5% frequency variation after trimming. The sensor supports both static RF power measurements and amplitude modulation (AM) demodulation through spike counting, highlighting compatibility with neuromorphic processors.The proposed PWM-based RF power sensor enables efficient, ultra-low-power, event-driven signal processing for inductively coupled IoT systems. Its spike-based output format facilitates direct integration with neuromorphic processors, providing a viable solution for power-constrained edge-AI applications.

Fields of Study & Subjects

Radio Frequency Integrated Circuit Design Advanced Memory and Neural Computing Analog and Mixed-Signal Circuit Design

Publication details

Journal / source
IEEE International Conference on RFID
Pages
1–6
Publication date
1 June 2026
Language
EN