TWI823592B - System and method for performing encrypted mixing based on big number format and additive homomorphism - Google Patents

Abstract

A system and a method for performing encrypted mixing based on big number format and additive homomorphism are provided. The method includes: performing a preprocessing-operation by a second-voice-information-processing-unit to obtain a voice-information-preprocessing-result, and storing the voice-information-preprocessing-result in a big number format; performing an encryption-operation on the voice-information-preprocessing-result by an encryption-and-transmission-unit to obtain an voice-information-encrypted-result in the big number format; performing an additive-homomorphism-operation on the voice-information-encrypted-result by a mixing-unit to obtain a mixed-ciphertext in the big number format; performing a decryption-operation on the mixed-ciphertext by a decryption-unit to obtain mixed-voice-information in the big number format; and performing a post-processing-operation on the mixed-voice-information by a first-voice-information-processing-unit to obtain a mixed-voice-information-post-processing-result.

Description

System and method for encrypted mixing based on large number types and additive homomorphism

The present invention relates to a system and method for performing encrypted mixing based on large number types and additive homomorphism.

Currently, the general encrypted mixing method is to perform voice encryption on the user device, then transmit the encrypted voice to the server, and the server performs decryption and mixing. Then, the server encrypts the mixing result and transmits it to the user device, so that the user device can perform decryption and playback. In other words, in the general encrypted mixing method, the server will perform decryption to obtain the plaintext voice information. If the server is monitored by a third party, this approach will cause information security risks.

The system of the present invention for performing encrypted mixing based on large number types and additive homomorphism includes user equipment and a server. The user equipment includes a decryption unit, a first voice information processing unit, a second voice information processing unit, and an encryption and transmission unit. The server includes a mixing unit, where the server communication is connected to the user equipment, wherein the second voice information processing unit performs a pre-processing operation to obtain a voice information pre-processing result, and stores the voice information pre-processing result in a large number type; encryption and The transmission unit performs an encryption operation on the voice information pre-processing result to obtain a large number type voice information encryption result; the mixing unit performs an additive homomorphic operation on the voice information encryption result to obtain a large number type ciphertext mix; the decryption unit Perform a decryption operation on the ciphertext mix to obtain the mixed voice information of a large number type; the first voice information processing unit performs a post-processing operation on the mixed voice information to obtain a post-processing result of the mixed voice information.

The method of performing encrypted mixing based on large number type and additive homomorphism of the present invention includes: performing a pre-processing operation by the second voice information processing unit to obtain the voice information pre-processing result, and storing the voice information pre-processing in large number type As a result; the encryption and transmission unit performs an encryption operation on the voice information pre-processing result to obtain a large number type voice information encryption result; the mixing unit performs an additive homomorphic operation on the voice information encryption result to obtain a large number type encryption result. text mixing; the decryption unit performs a decryption operation on the ciphertext mixing to obtain the mixed voice information of a large number type; and the first voice information processing unit performs a post-processing operation on the mixed voice information to obtain the mixed voice information. Post-processing results.

FIG. 1 is a schematic diagram of a system 10 for performing encrypted mixing based on large number types and additive homomorphism according to an embodiment of the present invention. System 10 may include multiple user devices 100 and servers 200 . It should be noted that the present invention does not limit the number of user equipments 100. The user equipment 100 and the server 200 can support voice transmission over IP (VoIP, Voice over IP).

The user equipment 100 (or server 200) has a processing unit (such as a processor but not limited to this), a communication unit (such as various communication chips, Bluetooth chips, WiFi chips, etc. but not limited to this) and a storage unit (such as : Removable random access memory, flash memory, hard disk, etc. (but not limited to these) are necessary components for running the user device 100 (or server 200). The server 200 can be communicatively connected to each user device 100 (respectively). For example, the server 200 can be communicatively connected to each user device 100 via the Internet 300 .

The user equipment 100 may include a receiving unit 110 and a speaking unit 120. Furthermore, the receiving unit 110 may include a sound collecting unit 111, a decryption unit 112 and a first voice information processing unit 113. In addition, the speaking unit 120 may include a recording unit 121, a second voice information processing unit 122, and an encryption and transmission unit 123. In this embodiment, the user equipment 100 may include an input and output device (not shown). The input/output device is, for example, a device for inputting speech or outputting speech, such as a microphone or a speaker. On the other hand, the server 200 may include a transmitting unit 210 and a mixing unit 220.

FIG. 2 is an operation flow chart of each unit of the system 10 shown in FIG. 1 . Please refer to both Figure 1 and Figure 2. In step S201, the recording unit 121 may receive voice information through the input and output device. In detail, the user can operate the user device 100 to input voice information (in the form of analog signals) to the recording unit 121 . Then, the recording unit 121 may convert the voice information in the form of analog signals into voice information in the form of digital signals, and send the voice information (in the form of digital signals) to the second voice information processing unit 122 . It should be noted here that the voice information may include (multiple) voice samples (Samples). The user can use the input and output device to set the number of bits of the speech sample. The number of bits is, for example, 8 bits or 16 bits, but the invention is not limited thereto.

In step S202, the second voice information processing unit 122 may perform a pre-processing operation to obtain a voice information pre-processing result, and store the voice information pre-processing result in a large number type. Specifically, the second voice information processing unit 122 may perform a voice information segmentation operation on the voice information to obtain a voice information segmentation result, and perform a voice information upward translation operation on the voice information segmentation result to obtain an upward translation result of the voice information. The upward translation operation of the speech information may be related to the number of bits of the speech sample. Specifically, the second voice information processing unit 122 can perform a voice information segmentation operation on the voice information according to each voice sample. Next, if the number of bits of the speech sample is bits, the second voice information processing unit 122 can add each voice sample in the voice information respectively. , to perform the upward translation operation of the voice information. In other words, since each voice sample in the voice information (in the form of a digital signal) obtained by the second voice information processing unit 122 from the recording unit 121 in the aforementioned step S201 may be a negative number, the second voice information processing unit 122 can convert the voice information into Each voice sample of is added separately , so that each speech sample must be a positive number. Furthermore, the second voice information processing unit 122 may add an overflow buffer code to the voice information upward translation result to obtain the voice information pre-processing result. The overflow buffer code may be associated with the number of bits (of the speech sample). For example, the second voice information processing unit 122 may use the following formula 1 to obtain the number of bits (length) of the overflow buffer code. Finally, the second voice information processing unit 122 can merge these voice samples, store the voice information pre-processing results in a large number format, and send the voice information pre-processing results to the encryption and transmission unit 123 . ……… (Formula 1) where k is the number of tracks before mixing.

It should be noted that in the above formula 1 The meaning of (number of tracks before mixing) may be the number of user devices 100 shown in FIG. 1 . Furthermore, in this embodiment, the server 200 can inform each user equipment 100 of the number of user equipments 100 in advance, so that the second voice information processing unit 122 can utilize the above-mentioned The value performs the operation of adding the overflow buffer code in step S202.

In step S203, the encryption and transmission unit 123 may perform an encryption operation on the voice information pre-processing result (stored in large number type) to obtain the voice information encryption result in large number type. Specifically, the encryption and transmission unit 123 may use the Paillier algorithm to perform an encryption operation on the voice information pre-processing result. Then, the encryption and transmission unit 123 can transmit the voice information encryption result to the sending and receiving unit 210 . Specifically, the encryption and transmission unit 123 can transmit the voice information encryption result to the transmission unit 210 of the server 200 through the Internet 300 according to the UDP/RTP transmission protocol.

In step S204, the sending and receiving unit 210 may disassemble the UDP/RTP transmission protocol packet according to the UDP/RTP transmission protocol to obtain the aforementioned voice information encryption result, and send the voice information encryption result to the mixing unit 220.

In step S205, the mixing unit 220 may perform an additive homomorphic operation on the voice information encryption result to obtain a large number type ciphertext mixing operation. Specifically, the mixing unit 220 may use the Payet algorithm to perform an additive homomorphic operation on the speech information encryption result. In other words, the server 200 of the present invention does not need to perform decryption, but only needs to add the ciphertext to obtain the ciphertext mix. Then, the sending and receiving unit 210 may package the ciphertext mix into a UDP/RTP transmission protocol format, and transmit the ciphertext mix to the radio unit 111 through the Internet 300 .

In step S206, after the sound collecting unit 111 receives the packet, the sound collecting unit 111 may use the UDP/RTP transmission protocol to unpack the packet to obtain the aforementioned ciphertext mix. Then, the sound collection unit 111 can transmit the ciphertext mix to the decryption unit 112 .

In step S207, the decryption unit 112 may perform a decryption operation on the ciphertext mix to obtain the mix voice information in the large number type. Specifically, the decryption unit 112 may use the Payet algorithm to perform a decryption operation on the ciphertext mix. Specifically, the decryption unit 112 can perform a decryption operation by mixing the ciphertext using the key of the Payet algorithm to obtain the mixed voice information, and send the mixed voice information to the first voice information processing unit 113 .

In step S208, the first voice information processing unit 113 may perform a post-processing operation on the mixed voice information to obtain a post-processing result of the mixed voice information. Specifically, the first voice information processing unit 113 can perform a mixed voice information segmentation operation on the mixed voice information according to the aforementioned overflow buffer code to obtain a mixed voice information segmentation result, and perform a voice information segmentation operation on the mixed voice information segmentation result. Pan down to obtain the mixed voice information post-processing results. The downward panning operation of the voice information may be related to the number of audio tracks before mixing and the number of bits of the voice samples. Specifically, if the number of bits of the speech sample (in the mixed speech information segmentation result) is bits, and is the number of tracks before mixing (as explained in the previous embodiment, It may be the number of user equipments 100 shown in Figure 1). The first voice information processing unit 113 may subtract each voice sample (in the mixed voice information segmentation result) respectively. , to perform the downward panning operation of the voice information. Finally, the first voice information processing unit 113 can output the post-processing result of the mixed voice information through the input and output device.

FIG. 3 is a schematic diagram of a group encrypted mixed call according to an embodiment of the present invention. Please refer to Figure 1, Figure 2 and Figure 3 at the same time. In the embodiment shown in FIG. 3 , the user "Xiaohua", the user "Xiaoming" and the user "Azhong" respectively operate the user device 100 to perform the group encrypted mixed call. In other words, in the embodiment of Figure 3, the above formula 1 The value (number of tracks before mixing, i.e. number of user devices 100) is 3. It is assumed here that each user equipment 100 and the server 200 shown in FIG. 3 have exchanged homomorphic encryption keys in advance, and it is assumed that the number of bits of the voice samples set by each user equipment 100 is 8 bits.

Similar to what was explained in the previous embodiments, the user "Xiaohua", the user "Xiaoming" and the user "Azhong" can respectively start a conversation with the user equipment 100 they hold. The conversation content can first be recorded by the recording unit 121 in the speaking unit 120 to obtain a digital signal composed of a series of voice samples with a unit number of 8 bits. Then, the digital signal can be pre-processed by the second voice information processing unit 122 to obtain a voice information pre-processing result. Then, the voice information pre-processing result can be encrypted by the encryption and transmission unit 123 to obtain the voice information encryption result, and the encryption and transmission unit 123 transmits the voice information encryption result to the transmission unit 210 of the server 200 . After the sending and receiving unit 210 receives the voice information encryption results of each user equipment 100 respectively, the mixing unit 220 of the server 200 can perform an additive homomorphic operation on the voice information encryption results of each user equipment 100 to obtain ciphertext mixing. The sending and receiving unit 210 then sends the mixed ciphertext back to the radio receiving units 111 of each user equipment 100 used by the user "Xiaohua", the user "Xiaoming" and the user "Azhong". Then, each user equipment 100 can perform a decryption operation on the ciphertext mix through the decryption unit 112 to obtain a large number of mixed voice information that has not undergone post-processing operations. Finally, the first voice information processing unit 113 of each user equipment 100 can respectively perform a post-processing operation on the mixed voice information to obtain a post-processing result of the mixed voice information (that can be output/played), and through the input of each user equipment 100 The output device respectively outputs the mixed voice information post-processing results.

FIG. 4A is a schematic diagram of the first stage operation of the system 10 shown in FIG. 1 during a group encrypted mixed call. FIG. 4B is a schematic diagram of the second stage operation of the system 10 shown in FIG. 1 during a group encrypted mixed call. Following the aforementioned explanation of the embodiment of Figure 3, in the embodiment of Figure 4A and Figure 4B, the formula 1 above The value (the number of audio tracks before mixing, that is, the number of user equipment 100) is 3, and the number of bits of the voice sample set for each user equipment 100 is 8 bits (i.e. value is 8).

Please refer to Figure 1, Figure 2, Figure 3 and Figure 4A first. The following uses the user "Xiaohua" as an example to continue the explanation. As shown in Figure 4A, it is assumed that the voice information received by the recording unit 121 of the user equipment 100 held by the user "Xiaohua" includes 3 voice samples with a unit number of 8 bits [11110110,00000110, 00011110]. The second voice information processing unit 122 of the user equipment 100 held by the user "Xiaohua" can add all three voice samples. , to perform the upward translation operation of the speech information and obtain the upward translation result of the speech information (ie, 3 speech samples [118, 134, 158]). Therefore, these three speech samples must be positive numbers (positive integers from 0 to 255). Then, in order to avoid overflow in the subsequent mixing process (of the server 200), when the second voice information processing unit 122 concatenates the three voice samples to form a large number, the second voice information processing unit 122 can Add an overflow buffer code to the voice information and shift the result upward to obtain the voice information pre-processing result. In detail, according to the above formula 1, since , the second voice information processing unit 122 can add an overflow buffer code with a bit number (length) of 1 at the beginning of these three voice samples to obtain the voice information pre-processing result (ie, 1110110010000110010011110), and complete the large number type state conversion. Then, the encryption and transmission unit 123 of the user equipment 100 held by the user "Xiaohua" can perform an encryption operation (ie, homomorphic encryption) on the voice information pre-processing result (in the large number type) to obtain the large number type voice The information encryption result is E(1110110010000110010011110). Similarly, the user equipment 100 held by the user "Xiao Ming" and the user equipment 100 held by the user "A Zhong" can also obtain their respective voice information encryption results in the same manner as the user "Xiao Hua". Then, each user equipment 100 can respectively transmit the voice information encryption result to the sending unit 210 of the server 200.

Please continue to refer to Figures 1, 2, 3, 4A and 4B. After the server 200 receives the voice information encryption results from each user device 100, the mixing unit 220 can perform an additive homomorphic operation on the voice information encryption results to obtain a (large number type) ciphertext mix, that is, E ( 110110100110011100100111101). Then, the sending and receiving unit 210 may return the ciphertext mix to the sound receiving unit 111 of each user equipment 100 . Then, the decryption unit 112 may perform a decryption operation on the ciphertext mix to obtain the mix voice information (in large number type), that is, 110110100110011100100111101. Then, the first voice information processing unit 113 may perform a post-processing operation on the mixed voice information to obtain a post-processing result of the mixed voice information. Specifically, the first voice information processing unit 113 can perform a mixed voice information segmentation operation (ie, string segmentation) on the mixed voice information according to the bits of each voice sample containing the overflow buffer code, to obtain the mixed voice information. Voice information segmentation results [110110100, 110011100, 100111101]. Furthermore, the first voice information processing unit 113 can perform a voice information downward translation operation on the mixed voice information segmentation result to obtain the mixed voice information post-processing result. In other words, the first voice information processing unit 113 can subtract each of the three voice samples from the mixed voice information segmentation result. , to obtain the mixed speech information post-processing results [52, 28, -67]. Finally, the first voice information processing unit 113 can output/play the post-processing result of the mixed voice information through the input and output device.

FIG. 5 is a flowchart illustrating a method for performing encrypted mixing based on large number types and additive homomorphism according to an embodiment of the present invention, wherein the method can be implemented by the system 10 shown in FIG. 1 . In step S501, the second voice information processing unit performs a pre-processing operation to obtain a voice information pre-processing result, and stores the voice information pre-processing result in a large number type. In step S502, the encryption and transmission unit performs an encryption operation on the voice information pre-processing result to obtain a large number type voice information encryption result. In step S503, the mixing unit performs an additive homomorphic operation on the voice information encryption result to obtain a large number type ciphertext mix. In step S504, the decryption unit performs a decryption operation on the ciphertext mix to obtain mixed voice information in the large number type. In step S505, the first voice information processing unit performs a post-processing operation on the mixed voice information to obtain a post-processing result of the mixed voice information. The method has been described in the foregoing embodiments and will not be described again here.

In summary, the system and method for performing encrypted mixing based on large number types and additive homomorphism of the present invention can perform pre-processing operations such as upward translation of the voice information and addition of an overflow buffer code on the voice information by the user equipment. The user device performs an encryption operation to obtain the voice information encryption result. Then, the server does not need to perform decryption, but performs an additive homomorphic operation on the voice information encryption result to obtain the ciphertext mix. Finally, the user device can perform a decryption operation on the ciphertext mix and output/play it after the user device performs a post-processing operation. In other words, since the server does not need to perform decryption, the server will not obtain the clear text voice information. This improves the security of voice information on user devices.

10: A system that performs encrypted mixing based on large number types and additive homomorphisms 100: User equipment 110:Receiver unit 111:Radio unit 112:Decryption unit 113: First voice information processing unit 120: Sending unit 121:Recording unit 122: Second voice information processing unit 123: Encryption and transmission unit 200:server 210: Sending and receiving unit 220: Mixing unit 300:Internet S201~S208, S501~S505: steps

FIG. 1 is a schematic diagram of a system for performing encrypted mixing based on large number types and additive homomorphism according to an embodiment of the present invention. FIG. 2 is an operation flow chart of each unit of the system shown in FIG. 1 . FIG. 3 is a schematic diagram of a group encrypted mixed call according to an embodiment of the present invention. FIG. 4A is a schematic diagram of the first stage operation of the system shown in FIG. 1 during a group encrypted mixed call. FIG. 4B is a schematic diagram of the second stage operation of the system shown in FIG. 1 during a group encrypted mixed call. FIG. 5 is a flowchart illustrating a method for performing encrypted mixing based on large number types and additive homomorphism according to an embodiment of the present invention.

S501~S505: steps

Claims (9)

A system for performing encrypted mixing based on large number types and additive homomorphism, including: user equipment, including a decryption unit, a first voice information processing unit, a second voice information processing unit, and an encryption and transmission unit; and a server, including Mixing unit, wherein the server communication is connected to the user equipment, wherein the second voice information processing unit performs a pre-processing operation to obtain a voice information pre-processing result, and stores the pre-voice information in a large number format Processing results; the encryption and transmission unit performs an encryption operation on the voice information pre-processing result to obtain the large number type voice information encryption result; the mixing unit performs addition homomorphism on the voice information encryption result The operation is performed to obtain the ciphertext mix of the large number type; the decryption unit performs a decryption operation on the ciphertext mix to obtain the mixed voice information of the large number type; the first voice information processing The unit performs a post-processing operation on the mixed voice information to obtain a mixed voice information post-processing result, wherein the first voice information processing unit performs a mixed voice information segmentation operation on the mixed voice information according to the overflow buffer code. To obtain the mixed voice information segmentation result, and perform a voice information downward translation operation on the mixed voice information segmentation result to obtain the mixed voice information post-processing result, wherein the downward translation operation of the voice information is associated with the mixed voice information segmentation result. The number of pre-sound tracks and the number of bits of the speech sample. The system of claim 1, wherein the user equipment further includes an input and output device, wherein the user equipment further includes a recording unit, and the recording unit receives voice information through the input and output device. The system of claim 2, wherein the voice information includes the voice sample, wherein the second voice information processing unit performs a voice information segmentation operation on the voice information to obtain a voice information segmentation result, and performs a voice information segmentation operation on the voice information. The voice information segmentation result performs a voice information upward translation operation to obtain a voice information upward translation result, wherein the voice information upward translation operation is associated with the number of bits of the speech sample. The system according to claim 3, wherein the second voice information processing unit adds the overflow buffer code to the voice information upward translation result to obtain the voice information pre-processing result, wherein the overflow buffer code associated with the number of bits. The system of claim 4, wherein the second voice information processing unit concatenates the voice samples to become the voice information pre-processing result in the large number type. The system of claim 1, wherein the server further includes a transmitting unit, wherein the user equipment further includes a sound receiving unit, wherein the encryption and transmitting unit transmits the voice information encryption result to the transmitting unit ; The transmitting unit transmits the mixed ciphertext to the radio unit. The system according to claim 1, wherein the user equipment further includes an input and output device, wherein the first voice information processing unit outputs the post-processing result of the mixed voice information through the input and output device. The system according to claim 1, wherein the encryption and transmission unit uses a Paillier algorithm to perform the encryption operation on the voice information pre-processing result; the mixing unit uses the Paillier algorithm. The additive homomorphic operation is performed on the voice information encryption result; the decryption unit uses the Payet algorithm to perform the decryption operation on the ciphertext mixing. A method for performing encrypted mixing based on large number types and additive homomorphism, suitable for systems including user equipment and servers, wherein the user equipment includes a decryption unit, a first voice information processing unit, and a second voice information processing unit and an encryption and transmission unit, and the server includes a mixing unit, wherein the method includes: performing a pre-processing operation by the second voice information processing unit to obtain a voice information pre-processing result, and storing it in a large number type The voice information pre-processing result; the encryption and transmission unit performs an encryption operation on the voice information pre-processing result to obtain the large number type voice information encryption result; the mixing unit performs an encryption operation on the voice information The information encryption result performs an additive homomorphic operation to obtain the ciphertext mix of the large number type; The decryption unit performs a decryption operation on the ciphertext mixing to obtain the large number type of mixed voice information; and the first voice information processing unit performs a post-processing operation on the mixed voice information. To obtain the mixed voice information post-processing result, the first voice information processing unit performs a mixed voice information segmentation operation on the mixed voice information according to the overflow buffer code to obtain the mixed voice information segmentation result, and performs a mixed voice information segmentation operation on the mixed voice information according to the overflow buffer code, and The mixed voice information segmentation result performs a voice information downward translation operation to obtain the mixed voice information post-processing result, wherein the voice information downward translation operation is associated with the number of pre-mixing audio tracks and the number of bits of the voice sample. .

Images