Research On Secure Speech Recognition System And Acoustic Communication Based On Nonlinearity Effect

With the rapid development of the Internet of Things(Io T)and artificial intelligence(AI),the connection between people and devices,devices and devices have become more common and intelligent.As an important information carrier,acoustic waves have been widely used in human-computer interaction(HCI)and device-to-device(D2D)communication.However,the speech recognition system brings convenience but also faces new security risks? for acoustic communi-cation,due to the characteristics of the audio interface of the smart device itself,the key to the wide application of acoustic communications is how to achieve the silent and high-speed communication.In this paper,we focus on the above problems and provide corresponding solutions.1.We design a completely inaudible attack,Dolphin Attack,that modulates voice commands on ultrasonic carriers(e.g.,f > 20 k Hz)to achieve inaudibility.By leveraging the non-linearity of the microphone circuits,the modulated low-frequency audio commands can be successfully demodulated,recovered,and more importantly interpreted by the speech recog-nition systems.We validate Dolphin Attack on popular speech recognition systems,includ-ing Siri,Alexa,etc.By injecting a sequence of inaudible voice commands,we show a few proof-of-concept attacks,which include activating Siri,Face Time to anyone,and even ma-nipulating the navigation system in an Audi car.Therefore,an attacker can imperceptibly manipulate smart device and causing a series of security issues.And,even more crucially,the attack distance can reach 20 m,which is difficult for users to perceive.2.Eliminating Dolphin Attacks is challenging if ever possible since it requires to modify the microphone hardware.In this section,we design Ear Array,a lightweight method that can not only detect such attacks but also identify the direction of attackers without requiring any extra hardware or hardware modification.Essentially,inaudible voice commands are modulated on ultrasounds that inherently attenuate faster than the one of audible sounds.By inspecting the command sound signals via the built-in multiple microphones on smart devices,Ear Array is able to estimate the attenuation rate and thus detect the attacks.We propose a model of the propagation of audible sounds and ultrasounds from the sound source to a voice assistant,e.g.,a smart speaker,and illustrate the underlying principle and its feasibility.We implemented Ear Array using two specially-designed microphone arrays and our experiments show that Ear Array can detect inaudible voice commands with an accuracy of 99 % and recognize the direction of the attackers with an accuracy of 97.89 %.The methods can be easily transplanted to smart devices to defend against such attacks.3.Acoustic communication has become a research focus without requiring extra hardware on the receiver side and facilitates numerous near-field applications such as mobile payment,data sharing.To communicate,existing researches either use audible frequency band or in-audible one.The former gains a high throughput but endures being audible,which can be annoying to users.The latter,although inaudible,falls short in throughput due to the limited available(near)ultrasonic bandwidth(20-24 k Hz).To address the conflict,we achieve both high speed and inaudibility for acoustic communication by modulating the coded acoustic sig-nal(0-20 k Hz)on the ultrasonic carrier.By utilizing the nonlinearity effect on microphone,the modulated audible acoustic signal can be demodulated and then decoded.We design and implement Ultra Comm,an inaudible acoustic communication system with OFDM scheme based on the characteristics of the nonlinear speaker-to-microphone channel.We evaluate Ultra Comm on different mobile devices and achieve throughput as high as 16.24 kbps which is 4 times the current highest communication rate.