Spectral Efficiency Analysis for a Spike-Based Pulse Repetition Frequency Modulation Scheme

The massive employment of wireless sensor systems and an increasing number of mobile sensors in the Internet of things (IoT) calls for highly energy-efficient designs for sensor edge nodes. In prior work a sensory system operating in the analog spike domain for sensing and communication was presented where the sensed information is encoded entirely in the pulse repetition frequency (PRF). Such a PRF modulation is also envisioned for joint communication and radar sensing. In this regard, the question on the spectral efficiency arises that is achievable with an end-to-end spike-based PRF modulation and how it compares to the related frequency-shift keying (FSK). Thus, in the present paper we perform a numerical evaluation of the spectral efficiency of PRF modulation considering two receiver architectures. One being a receiver design using a conventional analog-to-digital converter (ADC) while the other one uses an integrate-and-fire time encoding machine (IF-TEM) for sampling. Furthermore we describe an algorithm that selects PRFs that are easily distinguishable at the receiver. Our results show that PRF modulation provides the low spectral efficiencies required for the considered sensor node application, while allowing for end-to-end spike-based signaling covering the whole processing chain from the sensor edge node to the base station receiver. Moreover, the gained insight can be used to identify which spectral efficiency requirements spike-based PRF modulation can meet and for which applications this modulation scheme is a new contender.