Using 3D Printed Metamaterial Filters to Mitigate Inaudible Ultrasound Attacks on Smart Speakers

C. Shen, J. Lloyd, C. Ludwikowski, D. Phansalkar, C. Malik
Rowan University,
United States

Keywords: metamaterials, filters, acoustic, smart speakers

Summary:

Smart voice assistants have become a standard in most households in the last 5 years. Devices such as computers, smartphones, and smart speakers all have voice assistants built-in to help complete tasks. Due to the great increase in their popularity, there also grows a concern for the exploitation of these devices. For example, voice assistants like Amazon Alexa have the ability to control locks on doors and garages, alarm systems, and other electronics in someone’s home. In addition, these voice assistants can have access to one’s personal information including address, contacts, and payment information. All these variables pose a potential threat if the voice assistant was ever exploited to give another person access to these features. One way of accomplishing the exploitation of voice assistants is by using ultrasonic commands. These commands are outside of the audible range of humans but the microphones of these devices can still pick up the commands and execute them. This is because most microphones have an internal low-pass filter built into their hardware, with the cutoff usually being set at 20kHz to allow devices to operate in the audible frequency range. However, due to a “shadow” effect that occurs on the microphone diaphragm, inaudible frequencies are able to be processed as regular message signals, making any smart assistant vulnerable to inaudible attacks. In this work, we develop a metamaterial-based approach to enhance the security of voice assistants. A composite acoustic metamaterial filter composed of rigid panels and individual resonators is designed that physically modify ultrasonic attack signals so that they cannot trigger smart speakers. The proposed filters do not require any additional hardware to help prevent any attack signals. They are small and can be conveniently installed on any smart speaker in order to operate. Measurements are performed on an Amazon Echo as an example to validate the proposed approach. When the metamaterial filters are installed, ultrasound attacks are effectively mitigated while normal audible signals can still be processed without any interruptions. Thanks to the relatively simple configuration, these filters can be reliably fabricated by additive manufacturing and can be produced with reduced cost for mass production. In comparison to other defense methods, the metamaterial filters provide a much simpler and more cost-efficient option. Our approach provides a versatile solution for sound filtering in audio devices and is expected to greatly reduce the risk of smart speakers being exploited by ultrasonic commands in a versatile, cost-effective manner.