| In recent years,with the development of Internet of Things technology and artificial intelligence has made the human-computer interaction mode developed from a single touch method to multiple modes.With the development of the voice control system,It has more and more permissions,so in order to better promote the development of voice control systems,the discussion of its security is necessary.Non-linearity as a common property in circuits has been discovered and proposed in the last century.It is called harmonic distortion and inter-modulation distortion in the field of circuit design.So,this paper will discuss an ultrasonic attack method for the voice control system and conduct in-depth research;Firstly,this paper introduces the basic ultrasonic attack model and analyzes the existing defense mechanism.It finds that the existing defense mechanism is based on the characteristics of the signal and through the detection and attack rejection mechanism of the attack command.This paper points out the vulnerability and proposes a more advanced attack model through theoretical analysis.It verifies the effectiveness of our proposed attack model experimentally.The experimental results prove that we propose that a more advanced attack model can deceive the existing defense mechanism;Secondly,because the existing signal feature-based defense mechanism can not provide effective protection for the voice control system,this paper focuses on the defense mechanism of such attacks.For this reason,this paper proposes a new defense mechanism.Different from the existing system,this paper The proposed mechanism is based on the reliable detection and adaptive cancellation of spectrum shifting: firstly,the attack signal is copied and moved to the frequency band to be detected by using the spectrum shifting property,and then extracted as a reference signal to cancel the attack command.In order to verify its feasibility,we will implement the system and the feasibility of passing various experimental test systems on mobile devices.Thirdly,although the ultrasonic attack is harmful to the speech intelligent system,its principle can also have other applications.Finally,this paper discusses how to use its principle for ultrasonic communication.The ultrasonic communication mechanism designed in this paper uses the ultrasonic probe as the transmitter and the intelligent mobile device as the receiver.At the end,the communication rate of the original commercial device ultrasonic communication mechanism is more than 10 times,and the energy consumption is as low as one-sixth compared with other radio methods.This communication has great significance for the communication of the current Io T device.In general,the main technical contributions of this paper are: firstly,the existing defense system’s loopholes are pointed out,and more threatening attack methods are proposed.Secondly,a new defense mechanism is proposed,which makes the attack model completely invalid.Finally,an effective broadcast communication mechanism is designed by the attack principle.In this paper,we have conduct in-depth research on the specific problem of ultrasonic attack.With the development of intelligent technology,not only intelligent voice control systems but also other intelligent systems will have more security problems.In the future,we will further explore the system security field of smart devices. |