Design and Development of Backscatter-based Tag to Tag Networks

The project has previously developed a prototype RF tag platform called RIBBN (Research Infrastructure for Backscatter-Based Networks) - pronounced as 'ribbon.' RIBBN is a tag platform that uses the principle of backscatterd communication (similar to RFID), but there is no requirement of RFID readers or equivalent devices. They are powered by RF signals coming from 'exciters' specifically deployed for this purpose or ambient RF signals if present. The tags have the ability to communicate among themselves using completely passive backscatter modulation. We have developed a modular/extensible, programmable and powerful platform that will drive the future of IoT (Internet of Things).

In recently funded current effort (PNET, collaboration between Stony Brook and UT-Dallas) we are using new generations of the RIBBN tags to drive fundamental research in tag-to-tag networking paradigm, where RF tags are able to network among themselves and are able to operate in varying degrees of RF power availability with adaptive abilities. The goal is to form the foundational technology for smart spaces enabling object identification and tracking, understanding inter-object interactions and associations, sensing and distributed processing of information to support futuristic higher level applications.

We expect to make the following contributions: 1) Architectural innovations: We plan to develop an adaptive energy- aware tag architecture that will allow for high efficiency of energy harvesting over a wide range of incident RF power along with increased sensitivity and robustness of received signal demodulation in tag-to-tag link. 2) Detecting interactions and events: Measurement of channel amplitude and phase in a passive receiver tag will be enabled with novel backscatter modulation schemes and receiver architectures. This will empower the the tag network with the ability to detect dynamic interactions between tags as well as dynamic events in the environment around the tags. 3) Routing and information processing: We will develop scalable and adaptable distributed multi-hop routing for diverse topologies through novel hybrid routing approaches. We will also develop novel solutions for making inference under stressful conditions that include poor signal quality, lack of time-space coordinates, intrinsic asynchronism in receiving data, and limited computational and memory capabilities of the tags. 4) Demonstration and evaluation: A set of tags will be implemented in ASIC that will embody our research results. Few application scenarios in a realistic smart home-like setup, created by interactions between a number of objects tagged with developed prototypes, will be performed to demonstrate the robustness of the developed algorithms for network-wide inferencing.