Roth, FlorianBidoul, NoemieRosca, TeodorDoerpinghaus, MeikFlandre, DenisIonescu, Adrian M.Fettweis, Gerhard2022-08-292022-08-292022-08-292022-01-0110.1109/JCS54387.2022.9743501https://infoscience.epfl.ch/handle/20.500.14299/190342WOS:000835384400005Highly energy-efficient wireless sensor nodes are a prerequisite for a sustainable operation of the Internet of things. Therefore, classical approaches for system design based on digital signal processing are not a viable solution, but system design has to follow entirely new paradigms. In this regard, we present a sensory system with analog spike-based signal processing for sensing and communication, encoding the sensory information in the pulse repetition frequency (PRF), getting rid of energy hungry A/D and D/A conversion. Our spiking sensory system can generate spikes from any conventional analog output sensor using a compact, highly tunable voltage-controlled oscillator based on vanadium dioxide, and an analog differentiator circuit performing the transmit pulse shaping. The sole conversion from analog to digital takes place at the base station followed by the estimation of the PRF, for which we compare a conventional receiver design consisting of an analog-to-digital converter (ADC) with the use of an integrate-and-fire time encoding machine (IFTEM). Results show the successful communication of sensory information from the edge node over an additive white Gaussian noise channel to the base station, with the IF-TEM outperforming the conventional ADC for a signal-to-noise ratio above 0 dB.Engineering, Electrical & ElectronicTelecommunicationsEngineeringspike-based signalinganalog processingmetal-to-insulator transitiona/d conversionfrequency estimationtext::conference output::conference proceedings::conference paper