CN111865553B - Multi-audio encryption method based on chaos and Zigzag transform - Google Patents

Abstract

A multi-audio encryption method based on chaos and Zigzag transform belongs to the field of information encryption. At present, the activities of transmitting multimedia information on the network are becoming more and more frequent, and audio is an important part of multimedia information. How to protect the security of audio content has become one of the important research directions of researchers. The invention proposes a multi-audio encryption method based on chaos and Zigzag transform. First, use the random full arrangement generated by the chaotic sequence to scramble the connection sequence of multiple audio files; secondly, use Zigzag transform to scrambling phonemes between multiple audio files; finally, use the chaotic sequence to scrambling the phoneme matrix XOR operation to get the final encrypted audio set. Experiments show that this method can realize simultaneous encryption of multiple audio files, with good encryption effect, high security and high efficiency.

Description

Multi-audio encryption method based on chaos and Zigzag transform

technical field

This paper involves an audio encryption method, which mainly faces encryption of multiple audio files.

Background technique

In recent years, the rapidly updated mobile communication equipment and multimedia technology have attracted much attention to the security of multimedia data itself. As an important part of multimedia audio, audio has different characteristics from text and image information. Because of the large redundancy and strong correlation between its signals, the method to be taken for audio protection is different from that of images and texts. Among them, audio encryption is to use the matrix feature of digital audio, in the spatial domain of audio, according to a certain transformation rule, change the position or phoneme value of the phoneme, and make the information of the original audio "chaotic". As a new multimedia security technology, multi-audio encryption has the characteristics of high efficiency and strong confidentiality, and gradually attracts the attention of researchers.

Audio encryption technology mainly includes two means of phoneme scrambling and phoneme diffusion. The purpose of phoneme scrambling is to change the position of the phoneme; the purpose of phoneme diffusion is to change the value of the phoneme. In order to improve the security and efficiency of audio encryption and ensure the safe and efficient transmission of audio, a multi-audio encryption method based on chaos and Zigzag transform is designed by using chaos theory and Zigzag transform theory. The method utilizes the characteristics of multi-audio data, Zigzag transform has fast speed, good encryption effect and good randomness and complexity of chaos, and effectively protects the security of multi-audio file network transmission and storage.

SUMMARY OF THE INVENTION

The purpose of the present invention is to propose a multi-audio encryption method based on chaos and Zigzag transform for the problem of weak security of the existing single-audio encryption method.

Technical scheme of the present invention: in order to achieve the above purpose of the invention, the adopted scheme is a multi-audio encryption method based on chaos and Zigzag transform.

The multi-audio encryption method based on chaos and Zigzag transform is characterized in that, the encryption process comprises the following steps:

Step 1: Integerization of phonemes: let k interactive audio files be A ¹ , A ² , …, A ^k , the size of which is m × 1, where m is the number of phonemes, and 1 means that the audio file is mono; the elements in A ^j The range is [-1, 1], add 1 to all elements to make the range [0, 2], and multiply by the specified coefficient d , so that each phoneme value can be represented by n binary bits, that is

b _i ^j =( a _i ^j +1)× d , i =1, 2, …, m ; j =1, 2, …, k , (1)

Among them, a _i ^j ∈ A ^j , B ^j ={ b _i ^j } is a phoneme integer matrix; k number of m × 1 phoneme integer matrices B ¹ , B ² , …, B ^k can be obtained;

Step 2: Chaos sequence generation: randomly select the initial value x ₀ ∈ (0, 1) and the control parameter p ₁ ∈ (0, 0.5), and iterate the Piecewise Linear Chaotic Map (Piecewise Linear Chaotic Map, PWLCM) k times, can generate a chaotic sequence X ¹ ={ x _i ¹ };

， （2）

, (2)

Randomly select the initial value y ₀ ∈(0, 1) and the control parameter p ₂ ∈(0, 0.5), and iterate formula (2) k × m × n times to generate a chaotic sequence Y ¹ ={ y _i ¹ };

Step 3: Chaotic Sequence Integerization: Computation,

y _i ² = mod ( floor ( y _i ¹ ×10 ¹⁶ ), 2 ⁿ ), (3)

Among them, mod ( ) and floor ( ) are modulo and integer functions respectively, y _i ¹ ∈ Y ¹ , Y ² ={ y _i ² };

Step 4: File-level scrambling: use X ¹ to generate a full permutation P of the sequence 1, 2, …, k ; in the order of P , reshape B ¹ , B ² , …, B ^k into an m × k integer matrix C ;

Step 5: phoneme-level scrambling: utilize two-dimensional Zigzag transformation to convert C into an audio vector V with a length of mk ;

Step 6: Audio segmentation: with m phonemes as the unit, divide V into k audio vectors D ¹ , D ² , ..., D ^k equally;

Step 7: Phoneme Diffusion: Calculate:

E ¹ = D ¹ ⊕ Y ² , E ⁱ = D ⁱ ⊕ Y ² ⊕ E ^{i -1} , i =2, 3, …, k , (4)

Among them, ⊕ represents the exclusive OR operation; the diffusion results can be obtained as E ¹ , E ² , …, E ^k ;

Step 8: Phoneme decimalization: Divide each element in E ¹ , E ² , …, E ^k by d , and then subtract 1 to make its range become [-1, 1], that is

f _i ^j = e _i ^j / d -1, i = 1, 2, …, m ; j = 1, 2, …, k , (5)

Among them, e _i ^j ∈ E ^j , F ^j ={ f _i ^j } is a phoneme matrix; k encrypted audio frequencies F ¹ , F ² , …, F ^k can be obtained.

Further, in described step 5, Zigzag transformation: Zigzag transformation is a kind of common scrambling method, by starting from the upper left corner of the matrix to scan the number according to the "zigzag" shape, so as to achieve the replacement of data. Purpose. Then, the scanned elements are stored in a one-dimensional array in turn, and then converted into a two-dimensional matrix in a certain way. It is characterized by simple implementation and low time complexity.

Beneficial effects: Aiming at the problem of poor security of the single-audio encryption method, the present invention proposes a multi-audio encryption method based on chaos and Zigzag transform. The main contributions are: (1) synchronous encryption of multiple audio files; (2) file-level scrambling using chaotic sequences; (3) scrambling of phoneme matrices using two-dimensional Zigzag transform; (4) using chaotic sequences And XOR operation realizes the diffusion of phonemes; (5) This method utilizes the randomness and complexity of chaos and the simplicity and efficiency of Zigzag transform, which improves the encryption effect and encryption efficiency of audio. Therefore, the proposed multi-audio encryption method has the characteristics of high efficiency, security and good encryption effect, and can effectively protect the security of network transmission and storage of multiple audio files.

Description of drawings

Figure 1: Encryption flow chart of multi-audio encryption method based on chaos and Zigzag transform;

Figure 2: Time-domain waveform diagram of the original audio;

Figure 3: Time-domain waveform diagram of encrypted audio.

Detailed ways

The implementation process of the present invention will be further described in detail below with reference to the specific drawings and examples.

Fig. 1 is the encryption flow chart of this method.

The programming software used is Matlab R2016b, and five audio files with a size of 73113×1 are selected as the original audio. Using this method, the detailed process of encrypting the original image is described as follows.

Step 1: Phoneme integerization: The specific operation is to add 1 to all the phonemes, and then multiply by 10000, so that the range becomes [0, 20000].

Step 2: chaotic sequence generation: let 5 original audios be y ¹ , y ² , …, y ⁵ ; select the initial value y ₀ ¹ =0.78 and parameter p ¹ =0.44 of PWLCM, and iterate the map 73113×5 times, which can be Obtain a chaotic sequence Y ¹ ={ y _i ¹ } ₃₆₅₅₆₅ ; similarly, using the initial value of PWLCM x ₀ ¹ =0.32 and parameter p ² =0.44, and iterating the map 5 times, another chaotic sequence X ¹ ={ x _i ¹ } ₅ ; sort X ¹ to get the index value.

Step 3: Integerize the chaotic sequence:

y _i ² = mod ( floor ( y _i ¹ ×10 ¹⁶ ), 2 ⁸ ),

Among them, mod ( ) and floor ( ) are modulo and integer functions respectively, y _i ¹ ∈ Y ¹ , Y ² ={ y _i ² } is the integer result of Y ¹ .

Step 4: Perform file-level scrambling: use X ¹ to generate a full permutation P of the sequence 1, 2 ..., 5; according to the order of P , reshape B ¹ , B ² , ..., B ⁵ into a 73113 × 5 Integer matrix C.

Step 5: Phoneme-level scrambling: Convert C into an audio vector V of length 73113 × 5 using a two-dimensional Zigzag transform.

Step 6: Audio segmentation: With 73113 phonemes as the unit, V is divided into 5 audio vectors D ¹ , D ² ,..., D ⁵ .

Step 7: Do Phoneme Diffusion: Computation,

E ¹ = D ¹ ⊕ Y ² , E ⁱ = D ⁱ ⊕ Y ² ⊕ E ^{i -1} , i =2, 3, …, 5 ,

Among them, ⊕ represents the exclusive OR operation, and the diffusion results can be obtained as E ¹ , E ² ,…, E ⁵ .

Step 8: Phoneme decimalization: Divide each phoneme by 10000, and then subtract 1 to make the range become [-1, 1] to get 5 encrypted audios; their time domain waveforms are shown in Figure 3 .

In the decryption process, the same chaotic sequence, chaotic matrix and corresponding decryption method are used to act on the encrypted audio, and the time-domain waveform of the decrypted audio can be obtained as shown in Figure 2. The decryption process is the reverse process of encryption.

Claims (1)

1. The multi-audio encryption method based on chaos and Zigzag transformation is characterized in that the encryption process comprises the following steps:

step 1: and (3) phoneme integer quantization: order tokAn interactive audio file isA ¹, A ², …, A ^k Of a size ofm×1，mIs the number of phonemes, 1 means that the audio file is mono;A ^j middle elementIn the range of [ -1, 1]Adding 1 to all elements to make their range become [0, 2 ]]Then multiply by a specified coefficientdMaking each phoneme value availablenIs represented by a binary bit, i.e.

b _i ^j =(a _i ^j +1)×d，i=1, 2, …, m；j=1, 2, …, k， （1）

Wherein,a _i ^j ∈A ^j ，B ^j ={b _i ^j the phoneme integer matrix is used as the phoneme integer matrix; can obtain the productkAnmA phoneme integer matrix of x 1B ¹, B ², …, B ^k ；

Step 2: and (3) chaotic sequence generation: randomly selecting an initial valuey ₀E (0, 1) and control parameterp ₁E.g. (0, 0.5), Piecewise Linear chaos mapping (PWLCM) shown in iterative formula (2)m×kThen, a chaotic sequence can be generatedY ¹={y _i ¹}；

， （2）

Randomly selecting an initial valuex ₀E (0, 1) and control parameterp ₂E (0, 0.5), iterative formula (2)kThen, a chaotic sequence can be generatedX ¹={x _i ¹}；

And step 3: and (3) integer transformation of the chaotic sequence: the calculation is carried out according to the calculation,

y _i ²=mod(floor(y _i ¹×10¹⁶), 2 ⁿ )， （3）

wherein,mod() Andfloor() Respectively a modulo and an integer function,y _i ¹∈Y ¹，Y ²={y _i ²}；

and 4, step 4: file level scrambling: by usingX ¹The sequence of sequences 1, 2, …,ka full array ofP(ii) a According toPIn the order of (1) toB ¹,B ², …, B ^kIs reshaped into onem×kIs a matrix of integersC；

And 5: phoneme level scrambling: using a two-dimensional Zigzag transform, willCConverted into a length ofmkAudio vector ofV；

Step 6: audio frequency segmentation: to be provided withmTaking individual phoneme as unit, willVAre equally divided intokAn audio vectorD ¹, D ², …, D ^k ；

And 7: phoneme diffusion: and (3) calculating:

E ¹=D ¹⊕Y ²，E ⁱ =D ⁱ ⊕Y ²⊕E ^i-1，i=2, 3, …, k， （4）

wherein ≧ represents an exclusive or operation; the diffusion result can be obtained asE ¹, E ², …, E ^k ；

And 8: phonemic decimal transformation: to pairE ¹, E ², …, E ^k Each element in (1) divided bydThen decrease 1 to make its range to [ -1, 1 [)]I.e. by

f _i ^j = e _i ^j /d-1，i=1, 2, …, m；j=1, 2, …, k， （5）

Wherein,e _i ^j ∈E ^j ，F ^j ={f _i ^j the phoneme matrix is used as the phoneme matrix; can obtain the productkPersonal encrypted audioF ¹, F ², …, F ^k 。

Images