CN104363092B - The device authentication based on audio physical fingerprint under the conditions of spacing - Google Patents

Abstract

Audio-based physical fingerprint-based device authentication under distance conditions. The invention discloses a fixed-distance device authentication protocol based on the physical fingerprint of the audio hardware (speaker and microphone) of the wireless device. After the audio handshake between the two parties is authenticated, the protocol first generates a mixed frequency audio signal at the sending end, and at the same time sends the mixed frequency signal to the receiving end through a loudspeaker; secondly, the receiving end uses a microphone to receive the mixed frequency signal and extracts a physical fingerprint; finally , the receiving end uses the fingerprint matching algorithm based on the deviation rate to match the collected fingerprints with the fingerprints in the fingerprint library. The authentication method designed by the invention has the characteristics of high efficiency, low energy consumption, simple and convenient operation, high portability and the like.

Description

Device authentication based on audio physical fingerprint under fixed-distance conditions

technical field

The invention belongs to the field of wireless security, and more specifically relates to equipment authentication based on audio hardware physical fingerprints under the condition of a fixed authentication distance.

Background technique

In recent years, researchers have found that physical information of wireless networks can be used to achieve various authentications. The authentication method has a wide range of application scenarios, such as information anti-counterfeiting, identity-based attack detection, access control, fault detection, and target tracking. According to the different physical information used, the method can be divided into: authentication based on software fingerprint; authentication based on channel or location fingerprint; authentication based on hardware fingerprint.

Authentication based on software fingerprints uses the inherent characteristics of software or protocols running on devices for authentication. The most common is the use of the IEEE 802.11 standard Media Access Control (MAC) protocol for device authentication. The principles include: Due to the large and complex specification of the protocol, different device manufacturers and driver developers usually implement it differently; due to different combinations of chipsets, firmware and device drivers, individual devices reflect behavior of different MAC layers. However, the disadvantages of software-based fingerprinting include: different physical devices that use the same software cannot be well distinguished; the adversary can learn the behavior of legitimate users through observation, and imitate the behavior of legitimate users by changing their own device drivers.

The authentication based on the wireless channel or location fingerprint is realized through the location characteristics of path loss (Path Loss) and channel fading (Channel Fading). The physical quantities that reflect the characteristics of the wireless channel include channel state information (Channel State Information, CSI) and received signal strength (Received Signal Strength, RSS). Among them, CSI is a fine-grained description of channel characteristics, and RSS is a coarse-grained description of channel characteristics. Since it is difficult to measure CSI under existing hardware, the authentication method based on CSI cannot be well promoted. Existing wireless devices can easily obtain the RSS value, but because the RSS description of the channel is not accurate enough, it is vulnerable to impersonation attacks. In short, the existing authentication protocols based on wireless channel or location fingerprint either have special requirements for hardware or have security holes.

Considering the increasing commercialization of audio communication (including audio-based near-field communication), and most existing wireless devices are equipped with audio hardware such as microphones and speakers, the present invention designs an authentication based on the physical fingerprint of audio hardware under the condition that the authentication distance is fixed. Device Authentication Protocol. The protocol takes advantage of the physical unclonability of the frequency response of the wireless device microphone and speaker, and extracts the frequency response of the audio RF microphone and speaker to different frequencies as a physical fingerprint. This protocol can be widely used in various devices with audio hardware in the authentication Identity authentication under the condition of fixed distance.

Contents of the invention

The purpose of the present invention is to overcome the security problem of the existing device identity authentication based on the cryptographic method. At the same time, it eliminates the strict hardware requirements of the existing device identity based on the physical layer fingerprint of the device.

In order to achieve the above object, the present invention is based on audio hardware physical fingerprint device authentication at a fixed distance, which is characterized in that it includes: a mixing audio signal generation and sending module, used for the generation of audio signals when the device is learning and authenticating; an audio device A physical fingerprint extraction module is used to preprocess the mixed signal at the receiving end and extract frequency domain features; an audio physical fingerprint matching module is used to match the acquired audio fingerprint with the data in the fingerprint library during device authentication.

Description of drawings

In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the drawings that need to be used in the description of the embodiments will be briefly introduced below. Obviously, the drawings in the following description are only some embodiments of the present invention. For those skilled in the art, other drawings can also be obtained based on these drawings without any creative effort.

Fig. 1 is the authentication flowchart of the present invention;

Fig. 2 is a protocol frame diagram proposed in the present invention.

Detailed ways

The implementation of the present invention provides a fixed-distance device identity authentication protocol based on the physical fingerprint of the audio hardware. In order to make the purpose, features, and advantages of the present invention more obvious and understandable, the authentication scheme in the embodiment of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiment of the present invention. Obviously, the described The examples given are only the main parts of the invention and do not cover all details.

Figure 1 is the authentication flow chart of this protocol.

The authentication framework of this method is as follows: the initiator of authentication Alice (A for short) and the authenticator Bob (B for short) first establish a connection through an audio handshake; Send; again, after B uses the microphone to receive the audio signal, extract the physical fingerprint of A's audio device; finally, in the learning mode, B associates and stores the acquired physical fingerprint with A's ID, and in the authentication mode, B will collect The obtained fingerprint is matched with the fingerprint of A stored locally. If the match is successful, the authentication is passed; otherwise, the authentication fails.

Figure 2 is the protocol framework diagram of this agreement. The steps performed by the sender and receiver are as follows:

Sending end: (1) Establish a connection and obtain the receiving end ID; (2) Generate a mixed frequency signal; (3) Send a signal.

Receiver: (1) establish a connection, send the local ID, and obtain the other party's ID; (2) receive the signal; (3) perform FFT transformation and logarithmic processing on the received signal; (4) determine whether it is learning or authentication. If it is authentication, jump to step 5, if it is learning, save the learning file; (5) read the fingerprint data of the sender in the fingerprint database, and calculate the deviation rate; (6) compare the deviation rate with the threshold to determine whether the authentication is successful.

The following describes each module instance in the protocol:

Mixed audio signal generation and transmission module: In order to eliminate the interference of environmental noise and improve the certification efficiency, the source signal used here is the 4000Hz-20000Hz frequency band (with ΔHz as the step size) The frequencies are mixed together, namely:

Among them, Sound(A) is a mixed frequency signal, sin(β _i ) is a sinusoidal single-frequency signal with a frequency of 4000+Δ*(i-1), β _i =2*π*(4000+Δ*(i-1) )*T. It is worth noting that in the actual system, Δ=400Hz, n=41, the transmission duration is T=2 seconds, and it is stipulated that the volume transmitted in the learning mode and the authentication mode is the same.

Audio fingerprint extraction and matching module:

1. Use FFT to convert the audio signal in the time domain into an audio signal in frequency, and perform 20log(·) numerical processing on the amplitude, as the audio fingerprint obtained this time, abbreviated as: _OA = (ξ ₁ ,… , ξ _n ).

2. In the learning mode, B associates the fingerprint with A's ID, and stores the fingerprint sample in the fingerprint database.

3. In the authentication mode, B calls out the fingerprint sample associated with A’s ID from the fingerprint database, and the fingerprint sample is recorded as: O′ _A =(ξ′ ₁ ,…,ξ′ _n ), and calls rate matching algorithm DR-MA, if DR-MA(Γ,Δ, _OA , _O'A )=1, the matching is successful; otherwise, the authentication fails.

Wherein, the DR-MA algorithm is as follows: two thresholds Γ and Δ are set (in an actual system, Γ=8, Δ=0.5). Initialize S=0; T=0, for each i (i=L..., n), do the following cycle: if |ξ _i -ξ _i '|≤Γ, then S=S+1; otherwise, T=T +1; i=i+1; the loop ends. Finally, calculate the deviation rate: DR( _OA , _O'A )=T/S. If DR(O _A , O′ _A )≤Δ, output 1; otherwise, output 0.

The invention of fixed-distance authentication based on the physical fingerprint of audio equipment has the following characteristics:

1) Utilize frequency mixing technology to generate mixed frequency audio signal;

2) Convert audio signal to frequency domain signal to extract physical fingerprint;

3) Propose an offset-based DR-MA algorithm

The device identity authentication based on the audio physical fingerprint of the present invention has been introduced in detail above. In the present invention, the protocol proposed in the invention is described in detail; based on the authentication protocol proposed in the present invention, all other audio-based device identity authentication methods obtained by those of ordinary skill in the art without creative work are all Belong to the protection scope of the present invention.

Claims (2)

1. a kind of lightweight finger print matching method based on bias ratio, suitable for being amplified under the conditions of set a distance based on audio hardware The device authentication of the physical fingerprint of device and microphone：Two threshold value Γ and Δ are set, and remembers O_A=(ξ₁,…,ξ_n) it is mode of learning The physical fingerprint sample of lower acquisition, O '_A=(ξ '₁,…,ξ′_n) for the physical fingerprint sample under certification mode, initialize S=0, T =0, for each i, i=1 ..., n, following cycle is done, if | ξ_i-ξ′_i|≤Γ, then S=S+1；Otherwise, T=T+1；I=i+ 1；Cycle terminates, and finally, calculates bias ratio DR (O_A,O′_A)=T/S, if DR (O_A,O′_A)≤Δ, then export 1；Otherwise, 0 is exported.

2. in the device authentication system of the physical fingerprint based on audio hardware loudspeaker and microphone under the conditions of set a distance, ginseng Number values are respectively Γ=8, Δ=0.5.