Distributed Inference Based on 2-D Materials for Spectral Studies in the RF

Two dimensional (2-D) materials like graphene and molybdenum disulfide (MoS2) have  tremendous potential for the development of smart surfaces providing unprecedented RF sensing and communication capabilities to DoD systems. When applied at large scale, these smart surfaces will generate an immense amount of data to be processed and acted upon, creating a significant interconnection/communication bottleneck. This project will focus on developing the basic science, materials, devices and technologies needed to demonstrate two of the most important building blocks for future smart RF surfaces: high frequency electronic devices compatible with large area fabrication, and a distributed machine learning architecture to allow for real time local decision making (inference). This project will leverage the team’s extensive expertise on MoS2 devices and circuits to demonstrate the integration of GHz-class MoS2 RF receivers and transmitters with MoS2-based artificial synapses that enable the classification and prioritization of sensor data under almost-zero latency. These building blocks will be used to prototype a flexible smart surface capable of providing real-time RF sensing.