Project Objective
We envision a future where every object in our living and working environment will carry one or more RF tags. Such tags will also be embedded into physical infrastructure around us – such as building doors, windows or walls. These tags will be tiny (dime-sized), thin and light allowing them to blend into the object they are on. These tags will be able to sense activities and interactions around them thus forming context-aware, interactive smart environments possessing unprecedented ambient intelligence. The objective of this project is development of batteryless, RF-powered tags that can communicate to each other and perform activity recognition via analysis of reflected RF signals in order to fullfill this vision.

Project Description
We are developing RF tags that operate on harvested RF energy, have a passive envelope detector based receiver and a modest computing ability. They communicate directly with each other as they can resolve a reflected external RF excitation signal from a transmitting tag. This eliminates the need for the centralized device like RFID reader or on-tag high power radio. Prior to this work, the passive receiver based on envelope detection was not able to perform the wireless channel estimation as the receiver was not capable of IQ demodulation. Our novel technique to overcome this limitation is based on multiphase probing of the backscatter channel between a pair of tags, wherein the transmitting tag sends out the systematically designed probing signal which in turn enables the receiving tag to estimate channel parameters. This empowers an autonomous network of radio-less RF-powered backscattering tags with the ability to recognize and localize activities in the surrounding environments. Compared to centralized RFID reader based systems, our approach has (i) an order of magnitude lower cost due to the elimination of the high cost reader (ii) much greater scalability due to the ability of the tags to communicate with each other and form multihop networks and (iii) much higher precision, granularity and diversity in the measured RF fingerprint of the space, since for N tags our system establishes O(N2) distributed sensing channels as opposed to O(N) channels centralized around a single reader. By providing RF-powered passive RF tags the ability to ’fingerprint’ the surrounding wireless channels without any need for external active elements, we bring about a paradigm shift in passive RF tag networking. They are no longer limited to just traditional IoT like systems communicating ID information between each other, but can also sense and learn activities and interactions in their vicinity and propagate this knowledge all over the passive network.

Results
We have conducted an initial study in which we evaluated the feasibility of our proposed channel estimation technique for activity recognition. We developed an instrumented tag prototype with discrete components that records the demodulated signal at the receiving tag. In the study, we have collected training and testing samples for 9 participants that performed 10 different daily activities in a lab setting. The classification rates for some events are high even with single tag-to-tag link, while in others we have significant improvements in classification rate as the number of links increases. The results are comparable to those obtained by active radio-based techniques proposed in the recent literature.

Open-source RF tag platform
As a part of the project, we have developed an open-source RF tag. RF tag is an instrumented tag prototype that can be used for the data collection in experiments with tag-to-tag link in a presence of a dedicated exciter (RF source generator and antenna). RF tag implements a multiphase modulator and input baseband signal recording after the envelope detection. The multi-phase modulator integrates a 8-port RF switch. Switch is terminated with six impedances, preselected to provide the diversity in the reflection coefficient, along with an open-circuit and 50 Ohms. The envelope detector is followed by a low-pass filter and a 16-bit 1 MSample/s analog-to-digital (ADC) converter. The digitized amplitude of the input RF signal is collected by a microcontroller unit (MCU) from the ADC via SPI communication. Data can be written into a on-board micro SD card or transferred to PC using USB connector.

Schematic of the RF tag
BOM of the RF tag

The gerber files and software are available upon request.

Project Team
Milutin Stanaćević, Associate Professor, Electrical and Computer Engineering
Samir R. Das, Professor, Computer Science
Petar M. Djurić, Distinguished Professor, Electrical and Computer Engineering
Akshay Athalye, Research Scientist, Electrical and Computer Engineering
Xiao Sha, PhD student
Abeer Ahmad, PhD student
Puyang Zheng, PhD student
Dyumaan Arvind, PhD student

Former members:
Yuanfei Huang, now with Qualcomm

Publications
[1] A. Ahmad, X. Sha, M. Stanaćević, A. Athalye, P.M. Djurić and S.R. Das, ”Enabling Passive Backscatter Tag Localization Without Active Receivers,” Proc. 19th ACM Conference on Embedded Networked Sensor Systems (SenSys 2021), November 2021.
[2] A. Ahmad, X. Sha, A. Athalye, S.R. Das, P.M. Djurić and M. Stanaćević, ”Collaborative Backscatter Based on Phase Channel Estimation in Passive RF Tag Networks,” Proc. 11th IEEE International Conference on RFID Technology and Applications (IEEE RFID-TA 2021), October 2021.
[3] M. Stanaćević, A. Ahmad, X. Sha, A. Athalye, S.R. Das, K. Caylor, B. Glisić and P.M. Djurić, ”RF Backscatter-Based Sensors for Structural Health Monitoring,” Forth International Balkan Conference on Communications and Networking (BalkanCom’21), September 2021.
[4] X. Sha, PY. Zheng and M. Stanaćević, ”1.81 kHz Relaxation Oscillator with Forward Bias Comparator and Leakage Current Compensation Based Techniques,” Proc. 34th IEEE International System-on-Chip Conference (SOCC’21), September 2021.
[5] PY. Zheng, X. Sha and M. Stanaćević, ”Analysis of the Sub-uA Fully Integrated NMOS LDO for Backscattering System,” Proc. 34th IEEE International System-on-Chip Conference (SOCC’21), September 2021.
[6] Y. Huang, A. Athalye, S.R. Das, P.M. Djurić and M. Stanaćević, ”RF Energy Harvesting and Management for Near-zero Power Passive Devices,” Proc. IEEE International Symposium on Circuits and Systems Conference (ISCAS’21), May 2021.
[7] M. Stanaćević, A. Athalye, Z.J. Haas, S.R. Das and P.M. Djurić, "Backscatter Communication with Passive Receivers: From Fundamentals to Applications," ITU Journal on Future and Evolving Technologies, vol. 1(1), pp. 1-11, 2020.
[8] A. Ahmad, Y. Huang, X. Sha, A. Athalye, M. Stanaćević, S.R. Das and P.M. Djurić, ”On Measuring Doppler Shifts between Tags in a Backscattering Tag-to-Tag Network with Applications in Tracking,” IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), pp. 9055-9059, May 2020.