Abstract
Our work considers a new generation of Internet-of-Things (IoT) technology that uses 'passive' backscattering RF tags operating in an extremely low-power regime (microwatts). Such tags can communicate by reflecting an external RF signal (excitation) and can operate using the power harvested from the same RF signal. The most common use of such tags is in the RFID technology where a separate active radio transceiver ('reader') is used to communicate with the tags thus limiting the scalability and wide applicability of such tags. Our work considers recent advances where the passive tags can communicate among themselves without needing the presence of any active radio device. We envision that such networks of passive tags can enable a range of applications including identification, sensing, monitoring and tracking and develop technologies to enable some of these applications. Our work considers a mix of modeling and experimental study using tag prototypes developed using discrete components.
In the first part of our work we develop a robust method for tag-to-tag backscatter channel estimation using entirely passive techniques deployed on-board on the tags. Previously, such estimations were possible only using active radios. The proposed channel estimation method is able to isolate the backscatter radio channel between two communicating tags without influence from other channels such as between the tags and the source of the excitation signal. We demonstrate multiple applications of such channel estimation. We show that accurate channel phase estimation between any pair of communicating tags provides a range estimate. We develop a tag localization technique using the range estimates between all communicating pairs. This approach of tag localization is highly accurate relative to the more traditional approaches used in RFID due to the availability of a large number of tag-to-tag measurements. Experimental results show <1cm localization accuracy.
We also explore other applications of channel estimation. We show that the same estimation method can be used for diverse radio-based sensing applications such as tracking of mobile tags using Doppler shift measurements, detecting human activities in the surrounding environment and monitoring of engineering structures where the tags are embedded inside the structure.
In the second part of our work (ongoing) we develop techniques to build robust and energy-efficient communications in tag networks. Here, we focus on two aspects. In the first we study mechanisms to improve the range or robustness of tag-to-tag links via collaborative backscatters from a set of helper tags by using a form of beamforming. The collaborative backscatter works by boosting the desired signal via optimally setting the backscatter reflections from the helper tags. We describe the basic idea and results from preliminary experiments. In the second, we study the impact of the physical layer encoding in the energy consumption and bit-error performance of the tag-to-tag links. We present a new encoding scheme that has potential to reduce energy consumption significantly.
