Exploring the Landscape of Acoustic Authentication: Techniques, Performance and Security

The Internet of Things (IoT) is a rapidly growing network of interconnected devices, transforming our relationship with technology. The potential of the interconnected ecosystem relies on seamless device pairing. Previously, complex pairing processes hindered adoption due to user frustration. Bluetooth pairing, a standardized method for IoT device interaction, has limitations such as manual entry, error-proneness, and user intervention. This can hinder user experience and integration in multi-device ecosystems. Acoustic pairing is a potential alternative to traditional wireless communication, using sound waves as a communication channel between devices. Unlike conventional Bluetooth or Wifi pairing, it requires minimal user interaction and no additional hardware. The study result indicates that in acoustic authentication, On-Off Keying, FSK, GMSK, OFDM, and DFSK are the most common modulation techniques, while the effective range is from proximity (several mm) to maximum limited to acoustic signal propagation. Combining different modulation techniques could be a promising approach to address both high range and high success rates for secure acoustic communication. First, GMSK modulation achieved a Bit Mismatch Metric (BMM) of 0.5%-3% over 50-70cm. Second, a novel deep learning-based modulation that embeds information onto channel-specific features. This approach achieved a Bit Error Rate (BER) of less than 1% over a 25 cm² area. Lastly, OFDM achieved a success rate greater than 95% over 20 - 220cm. Those results indicate that combining different modulation techniques, such as GMSK, deep learning-based modulation, and OFDM, can be a promising approach to achieving both high range and high success rates for secure acoustic communication systems. In addition, most of the reviewed studies are resilient against common attacks.