Reliable and Low-Complexity Chirp Spread Spectrum-Based Aerial Acoustic Communication

Abstract

Aerial acoustic communication (AAC) has been developed for a variety of Internet-of-things (IoT) applications, thanks to the merit of embedding small amounts of data directly into audio contents and transmitting them through commercial off-the-shelf (COTS) devices without additional infrastructure. However, AAC has clear limitations due to ambient noise, severely frequency-selective acoustic channels, and high power consumption of COTS devices. We adopt chirp spread spectrum (CSS) modulation and design new quaternary symbols to overcome frequency selectivity of audio interfaces of COTS devices. We also develop a computation-efficient method to demodulate the proposed symbols, aiming to reduce the power consumption of COTS devices. We employ a frame combining technique without increasing computational complexity, to mitigate the effects of multipath fading and ambient noise in acoustic channels. To evaluate the proposed methods, we conduct extensive experiments with several smartphones in various environments. The experimental results and evaluation demonstrate that the proposed symbols and frame combining method contribute to improving the frame reception ratio by up to 59.8%p (267.9%) and 14%p (84.3%), respectively. The frame combining increases the possibility of receiving a frame within two attempts at extremely low SNR, by up to 107.9%, resulting in reduced excessive delay. The method to demodulate the proposed symbols lowers power consumption by tens to hundreds of milliwatts, depending on the device. Our proposal significantly helps to achieve high reliability and low power consumption of COTS devices in AAC. INDEX TERMS aerial acoustic communication, ambient noise, chirp spread spectrum, correlator, frame combining, software-based digital modem.

High Data Rate Near-Ultrasonic Communication with Consumer Devices

“…A comparison of the data rates achieved in these works with package error rates varying from 0.008 to 0.1 can be seen in Table I. References [2], [9], [24] employed chirp-based modulation, and [4], [25] used coherent modulation. None of these works included an equalizer at the receiver.…”

“…For consumer devices, the frequency response of the built-in hardware in the near-ultrasonic range is usually nonflat [2], [4]. The nonflat frequency response of the built-in hardware causes dispersion, and the symbol waveforms interfere with each other.…”

“…In applications such as network credential exchange, user authentication or proximity-based content sharing, a few hundred bits of data are needed to be transmitted quickly and locally. Airborne acoustic communication (AAC) is steadily gaining significance in those scenarios [1], [2].…”

“…Three main challenges arise from using the built-in microphones and speakers of consumer devices for high data rate NUSC: First, the frequency response of the built-in hardware in the near-ultrasonic range is usually nonflat [2], [4], [9]. Second, the indoor acoustic channel is frequency selective with severe intersymbol interference (ISI) at high rates due to the hardware characteristics and multipath.…”