CN101026836A - Anti vocoder compressed end-to-end voice encryption device and method - Google Patents

Abstract

The end-to-end voice encryption device and method of anti-vocoder compression relate to the field of end-to-end secure communication in the global mobile communication system, specifically an end-to-end anti-vocoder compression implemented on the traditional voice channel of the GSM mobile communication system End-to-end voice encryption device and method. It consists of an encryption module (1) at the sending end and a decryption module (2) at the receiving end. The encryption module (1) at the sending end includes a low code rate compression module (4), an encryption module (5), an orthogonal multi-carrier modulation module (6), a first carrier generator module (7), and a first key management negotiation module (8) and the first synchronization module (9); the receiving end deciphering module (2) comprises an orthogonal multi-carrier demodulation module (14), a second carrier generator module (15), a deciphering module (16), a low code rate A decompression module (17), a second key management negotiation module (18) and a second synchronization module (19). The invention performs end-to-end encrypted voice transmission through the GSM voice channel, and has the advantages of low time delay, high security, flexible and convenient implementation, and the like.

Description

End-to-end voice encryption device and method against vocoder compression

technical field

The present invention relates to the field of end-to-end secure communication in the Global System of Mobile Communication (GSM, Global System of Mobile communication), in particular to an end-to-end voice that is compressed by an anti-vocoder implemented on the traditional voice channel of the GSM mobile communication system Encryption device and method.

Background technique

GSM is a cellular mobile communication standard based on time-division multiple access mode formulated by the International Telecommunication Union Standardization Organization. It can provide services such as voice, data, short message and wireless Internet access. It was first applied in Europe and my country and achieved great success. At present, China's GSM network has covered most of the cities and towns in the country, and has become an important means of communication for all walks of life in the national economy.

With the rapid development of the GSM network, a large amount of sensitive information related to business and personal needs to be transmitted through the GSM mobile phone. For the open electronic communication system of GSM mobile phone communication, GSM mobile phone and base station are transmitted through wireless air, and the confidentiality of communication is poor. The GSM system voice communication itself provides a set of security management functions, but its security functions are incomplete and incomplete, and there are security risks such as one-way authentication, weak A5 encryption algorithm and transparent transmission of voice behind the base station; at the same time, the GSM system itself The encryption is controlled by the operator, and the system implementation is complex. In order to ensure the security of communication information, it is necessary to use additional encryption methods to encrypt and protect the transmitted information. Wireless communication network generally adopts air interface encryption and end-to-end encryption in two ways. The air interface encryption only encrypts and protects the information transmission on the wireless channel between the mobile phone and the base station, and the data is easy to be eavesdropped and stolen. The end-to-end encryption encrypts and protects all links from the source to the destination, which is suitable for end-to-end secure transmission of voice and data, and has high security and confidentiality.

The end-to-end encryption of mobile voice needs to be encrypted in the source processing part of the terminal to achieve end-to-end security. Mobile communication can carry voice and data services, so voice transmission includes voice transmission of traditional voice channel and voice transmission of data channel (via circuit switched link data channel or IP protocol channel on GPRS), so end-to-end mobile secure voice communication There are also methods for these two channels. The advantage of voice channel encryption is that the delay is small, and it can be well integrated with various services of the existing mobile network (such as phone cards and various packages), and it has good popularity among the existing users. The advantage of data channel encryption is that it is easy to realize confidential voice communication because it avoids the RPE-LTP vocoder. However, the data channel encrypted voice transmission has the following disadvantages: a) the problem of excessive delay caused by establishing a connection and using an automatic retransmission mechanism cannot be overcome; b) there is a problem of interoperability of the data channel between mobile networks in various regions; c ) The data channel method needs to open data services at the operator, and the existing phone card packages and other services cannot be used at both ends.

Performing voice signal encryption processing directly on the voice channel will make the encrypted signal not have the characteristics of voice, and it is difficult to effectively pass the Regular Pulse Excited-Long Term Prediction (RPE-LTP, Regular Pulse Excited-Long Term Prediction) vocoder of the GSM system. RPE-LTP vocoder is a compression coding algorithm based on mixed parameters. It compresses by extracting the sound model parameters of the input speech, and then transmits the compressed parameters, and restores the original speech through these parameters at the receiving end. Generally, the voice restored by the receiving end is basically the same as the original voice in terms of voice characteristics, but it is different from the original voice in waveform. It is for this reason that the bit scrambling encryption method on the general wired network is limited to be used on the GSM voice channel.

At present, there are methods for encrypting voice and data of mobile phones. For example, the invention with the Chinese patent application No. 98108859.7 proposes a new communication encryption method, but it mainly focuses on the negotiation process of the end-to-end key. The main content of this method is: the calling party terminal uses the random number sent by the network to generate an encryption key, and the encryption key is sent to the called party terminal through the call setup message, so that the users at both ends can use the same encryption key when communicating. key to encrypt/decrypt call information. The invention with the application number of 200410030771 describes a method for encrypting and transmitting voice over IP based on the GSM data channel to realize voice encryption of mobile phones. The invention with application number 200410021689.2 provides an end-to-end encryption method for GSM network voice and data services. By adding a common external terminal security device, the external terminal security device is connected with the GSM mobile station supporting the Bluetooth data communication function through Bluetooth technology, and the end-to-end encryption of GSM network voice and data services can be realized. The invention whose application number is 200410024248 only describes a model and method for voice scrambling and encryption on an analog channel, without in-depth analysis of specific anti-vocoder compression; the invention whose application number is 200410013706 describes only applicable to public The method of encrypted voice transmission using information hiding in the telephone switching network does not consider the influence factors of the vocoder compression on the GSM.

According to the survey results, most of the existing patents are focused on applying for protection of mobile phone encryption models, methods, and frameworks, and most of them focus on the mode of voice transmission through data channels. A new type of voice encryption method for voice channel anti-vocoder compression has not yet found any published patents. And the anti-vocoder compression end-to-end encryption transmission device and method of a kind of GSM voice channel described in the present invention, by the voice data after the encryption is modulated into the waveform that can pass through the GSM vocoder through the orthogonal multi-carrier modulation mode , the realized voice encryption process has low delay, high voice intelligibility, good voice quality, can adapt to the whole GSM network, there is no network compatibility problem, and can adapt to different voice tariff charging policies of operators.

Contents of the invention

Technical problem: the object of the present invention is to realize the anti-vocoder compression end-to-end voice encryption device and method through the GSM network voice channel, and propose a kind of anti-GSM RPE-LTP13Kbps full-rate voice code based on orthogonal multi-carrier modulation The end-to-end encryption device and method for vocoder compression, that is, the end-to-end voice encryption device and method for anti-vocoder compression, combined with a low bit rate compression module, a synchronization module and a key agreement module, without the need to open additional services Under this condition, a simple, fast, safe and low-cost GSM end-to-end secure voice communication device is realized.

Technical solution: A one-way encryption process of an end-to-end voice encryption device that resists vocoder compression includes a sending end encryption module and a receiving end decryption module; the sending end encryption module includes a low bit rate compression module, an encryption module, and an orthogonal multi-carrier Modulation module, first carrier generator module, first key management negotiation module and first synchronization module; receiver decryption module includes orthogonal multi-carrier demodulation module, second carrier generator module, decryption module, low code rate solution Compression module, the second key management negotiation module and the second synchronization module; wherein, the input end of the low bit rate compression module is connected to the output end of the A/D encoding module of the original sending mobile phone, and the low bit rate compression module and the key management negotiation The output terminal of the module is connected to the input terminal of the encryption module, the output terminals of the encryption module, the carrier generator module and the synchronization module are connected to the input terminal of the quadrature multi-carrier modulation module, and the output terminal of the quadrature multi-carrier modulation module is connected to the first mobile phone of the original sending The input terminal of a RPE-LTP vocoder; the decryption module at the receiving end consists of an orthogonal multi-carrier demodulation module, a second carrier generator, a decryption module, a low code rate decompression module, a second key management negotiation module and a second The synchronous module is composed of, wherein, the input terminal of the orthogonal multi-carrier demodulation module is respectively connected with the output end of the second carrier generator, the second synchronous module and the second RPE-LTP vocoder of the original receiving mobile phone, and the orthogonal multi-carrier demodulator The output terminal of the tuning module and the second key management negotiation module is connected to the input terminal of the decryption module, the output terminal of the decryption module is connected to the input terminal of the low code rate decompression module, and the output terminal of the low code rate decompression module is the D of the receiving mobile phone. /A The input terminal of the decoding module.

Use the orthogonal multi-carrier modulation method to establish an end-to-end transmission channel to realize encrypted voice communication; the linear PCM digital voice from the A/D quantization coding module of the mobile phone is standard 8KHz sampling, 13bit quantized data, and this data is compressed at a low bit rate After the module is compressed, it first becomes a compressed code stream with a low code rate; the compressed code stream is encrypted by the encryption module, and then sent to the orthogonal multi-carrier modulation module, which is processed by the carrier wave generated by the first carrier generator module. Modulate to form a transmission signal waveform that can pass through the vocoder compression, and then send it back to the first RPE-LTP vocoder of the baseband processing chip of the mobile phone for RPE-LTP compression coding, and then send it to the channel coding module for channel coding and other At the receiving end of the mobile phone, the signal from the antenna is processed by the channel decoding module and the second RPE-LTP vocoder, and then sent to the orthogonal multi-carrier demodulation module of the receiving end decryption module , together with the local second carrier generator module for demodulation processing; the demodulated signal is decrypted by the decryption module and sent to the low bit rate decompression module to restore the linear PCM 8KHz sampling and 13bit quantized digital voice data. Finally, it is sent to the D/A decoding module in the mobile phone to form a decrypted voice stream.

The low bit rate compression module and the low bit rate decompression module are used in pairs to compress the input 8KHz, 13bit PCM voice into a bit stream lower than 2.4K with the AMBE encoding algorithm, or decompress the compressed bit stream with the AMBE decoding algorithm 8KHz, 13bit PCM voice signal. The encryption module and the decryption module are used in pairs. The encryption and decryption algorithm in the module adopts the AES digital encryption algorithm of the enhanced encryption standard, or the RC4 algorithm based on bit stream encryption.

The first carrier generator module, the second carrier generator module, the orthogonal multi-carrier modulation module, and the orthogonal multi-carrier demodulation module are combined to realize the orthogonal multi-carrier modulation transmission and demodulation process of the compressed voice code stream ; This orthogonal multi-carrier transmission process has two implementation modes; one is to complete the direct modulation and demodulation with the carrier generated by the first carrier generator module or the second carrier generator module; the other method is to use orthogonal Frequency division multiplexing algorithm to achieve fast.

When the orthogonal multi-carrier modulation method is used to establish an end-to-end transmission channel, orthogonal frequency division multiplexing is used to realize fast orthogonal multi-carrier modulation and demodulation. The up modulation mode adopts QPSK modulation mode.

When adopting OFDM to realize fast orthogonal multi-carrier modulation and demodulation, the input code stream is first converted to serial and parallel, and then filled to the specified carrier point, while other carrier points are filled with 0 values, and at the same time In order to ensure that the fast Fourier transform FFT calculation is a real number, it is necessary to ensure that the data before the fast Fourier transform FFT and the fast inverse Fourier transform IFFT are based on the front and rear conjugate even symmetry of the center point; the transformation of the fast inverse Fourier transform IFFT and the fast Fourier transform FFT is used To realize the modulation and demodulation of orthogonal multi-carrier.

Both ends of the encrypted voice communication need to use the first key management negotiation module and the second key management negotiation module to realize key synchronization and key update management at both ends of the communication; Key agreement mechanism to complete.

The two ends of the encrypted voice communication also need a synchronization module to realize the synchronization of the encrypted frame; a training synchronization sequence is added at the beginning of each frame, and the synchronization sequence contains two identical parts before and after in the time domain; a synchronization symbol contains two Periodic sine wave or square wave, the frequency of sine wave or square wave is between 300Hz and 500Hz; at the receiving end, the correlation calculation is performed on the received sample points of the two halves and accumulated, and the peak point obtained is the synchronous starting point. The starting point; the sending end adds a synchronization sequence to each encrypted orthogonal multi-carrier frame, and at the receiving end, the corresponding synchronization sequence is removed.

The linear PCM digital voice from the mobile A/D quantization coding module is standard 8KHz sampling and 13bit quantized data. After the data is compressed by the low bit rate compression module, it first becomes a low bit rate compressed bit stream. at 2.4Kbps. In this way, within the 20ms GSM voice, the voice information can be effectively represented with a code smaller than 48bit. After the compressed code stream is encrypted by the encryption module, it is sent to the orthogonal multi-carrier modulation module, and modulated with the carrier wave generated by the carrier generator to form a transmission signal waveform that can pass through the vocoder compression, and then sent back to the The RPE-LTP vocoder of the mobile phone's baseband processing chip performs RPE-LTP compression coding, and then sends it to the channel coding module for channel coding and other operations, and finally sends it out through the mobile phone antenna.

At the receiving end of the mobile phone, after the signal from the antenna is processed by the channel decoding module and the RPE-LTP vocoder, it is first sent to the orthogonal multi-carrier demodulation module of the decryption module at the receiving end, and is carried out together with the local carrier generator. demodulation processing. The demodulated signal is decrypted by the decryption module and sent to the low bit rate decompression module to recover the linear PCM 8KHz sampling and 13bit quantized digital voice data, and finally sent to the D/A decoding module in the mobile phone to form a decrypted voice stream . At the sending and receiving ends of the mobile phone, the corresponding key management negotiation module and synchronization module complete the process of key negotiation and encrypted frame synchronization at both ends of the call.

The low bit rate compression module and the low bit rate decompression module are used in pairs, using the AMBE encoding algorithm with a lower compression bit rate to compress the input 8KHz, 13bit PCM voice into a bit stream lower than 2.4Kbps, or compress the compressed The code stream is decompressed into 8KHz, 13bit PCM voice signal. The AMBE coding algorithm is a kind of multi-band excitation coding algorithm, which can realize high-efficiency sound compression coding. Encryption module and decryption module must be used in pairs. Usually, the Enhanced Encryption Standard (AES) digital encryption algorithm is used, or the RC4 bit stream encryption algorithm is used. The AES algorithm is a kind of block encryption algorithm with high encryption strength. The RC4 algorithm is a kind of stream encryption algorithm, and the encryption speed is fast.

The carrier generator module is combined with the orthogonal multi-carrier modulation module and the demodulation module to realize the orthogonal multi-carrier modulation transmission and demodulation process for the compressed voice code stream. There are two implementation modes for this orthogonal multi-carrier transmission process. One is to use the direct modulation and demodulation of the carrier generated by the carrier generator module. The carrier generator module generates corresponding carrier signals. According to the voice frequency range of the GSM call (50Hz to 3400Hz), the carrier generator outputs the frequency of the relevant frequency point, and modulates and demodulates on this frequency, and it is necessary to ensure that the orthogonality between the carriers is ensured. Through our experiments, the sine wave below 1500 Hz is modulated by quadrature phase shift keying QPSK to carry the compressed bit stream data through the vocoder and the bit error rate can be less than 10 ^-3 . When a higher frequency is used for modulation and demodulation, the bit error rate of the transmission will become higher. The number of sub-carriers of the orthogonal multi-carrier is related to the frequency limitation used, the interval of the orthogonal frequency, and the transmission rate of the required modulated data. The number of subcarriers multiplied by the orthogonal frequency interval is the highest frequency used by the current modulation and demodulation, which should be kept in the frequency range with a low bit error rate (such as within approximately 1500Hz), but at the same time the number of subcarriers will be within a certain range. To a certain extent, it affects the bit rate of transmission, so the number of subcarriers needs an appropriate value. Assume that the number of subcarriers is n, the data rate is R _b b/s, and the highest signal frequency is F _m . Assuming that the subcarrier adopts the phase modulation mode of PSK, each subcarrier can modulate m bit data. Assuming that the starting frequency is equal to the frequency interval, the highest frequency of the modulated carrier signal can be derived as:

F _m =R _b /mHz

It can be seen from the above formula that the highest frequency is directly related to the input data rate requirement and the number of bits modulated on each subcarrier, but not directly related to the number of subcarriers. That is to say, we want to obtain the highest frequency with a lower bit error rate. When the input data rate requirement is determined, increasing the number of bits carried on the subcarrier can obviously reduce the highest frequency of the orthogonal multi-carrier subcarrier. The number of bits that can be modulated by each subcarrier is related to the system characteristics of the vocoder. It is proved by experiments that the sinusoidal waveform passing through the vocoder has a smaller bit error rate for QPSK modulation of 2 bits at most. Therefore, before entering the vocoder, we compress the voice to less than 2.4kb/s, and use 24 orthogonal subcarriers for QPSK modulation, which can control the highest frequency used to be less than 1500Hz to achieve a lower bit error rate.

Another way to realize the orthogonal multi-carrier modulation and demodulation process is to use the Orthogonal Frequency Division Multiplexing (OFDM Orthogonal Frequency Division Multiplexing) algorithm to quickly realize. Modulation and demodulation of orthogonal multi-carriers are realized by transforming fast inverse Fourier transform IFFT and fast Fourier transform FFT.

Orthogonal frequency division multiplexing OFDM modulation method is a kind of multi-carrier modulation technology. Its basic idea is to serially convert data streams into N sub-data streams with lower rates, use them to modulate N sub-carriers respectively, and then transmit them in parallel. The rate of the factor data flow is the original 1/N, that is, the symbol period is expanded to N times of the original, so that OFDM divides a wideband frequency selective channel into N narrowband flat fading channels, which has a strong resistance to wireless channels The ability of multipath fading and anti-pulse interference is especially suitable for high-speed wireless data transmission. OFDM selects subcarriers that are orthogonal to each other in the time domain. Although they alias each other in the frequency domain, they can still be separated at the receiving end.

In the orthogonal multi-carrier modulation module, OFDM modulation can be used to complete. The modulation principle of the OFDM sending end is as follows: the serial bit stream with a rate of R _b bps, according to the size N of the inverse Fourier transform IFFT (such as 128 IFFT), first performs serial-to-parallel conversion and converts it into N rows of parallel streams, Each row of parallel streams represents an orthogonal subcarrier. Then, according to the code stream compressed by the previous low code rate compression module (for example, 2.4Kbps, there is 48bit information every 20ms), the code of the row corresponding to the subcarrier is firstly filled. The carrier carrying this effective information still needs to be controlled within the GSM voice spectrum between 50Hz and 3400Hz, and to ensure that QPSK modulation can pass through the vocoder, the highest frequency of the subcarrier needs to be less than 1500Hz. Except for subcarriers carrying effectively compressed voice information, the codes on other subcarriers are generally filled with 0 values. It should be noted that in order to ensure that the IFFT results are real numbers, the 128-point IFFT is mirrored according to the center point, that is to say The second half of the data is the first half of the data is conjugate even symmetry. In this way, the serial-to-parallel converted code is subjected to QPSK modulation, and then undergoes inverse Fourier transform IFFT. Finally, the transformed signals are superimposed and converted to parallel and serial. After forming the modulated waveform and adding the synchronization sequence, it can be sent to the vocoder for transmission.

At the OFDM demodulation end, the OFDM modulated signal from the vocoder is firstly processed by Fourier transform FFT. Then, according to the code stream correspondence table at the time of modulation, the effective compressed voice information is extracted from the converted information, and after parallel-to-serial conversion, it is sent to the low bit rate decoding module for decoding processing, and the voice can be decoded Come.

Both ends of an encrypted call need to use a key management negotiation module to realize key synchronization and key update management at both ends of the communication. The key exchange process can be completed by the two communicating parties through the Diffie-Hellman (Diffie-Hellman) key agreement mechanism.

Both ends of encrypted voice communication also need a synchronization module to realize the synchronization of encrypted frames. Synchronization is an important link in any digital communication system. Without an accurate synchronization algorithm, it is impossible to send and receive data accurately and reliably. Orthogonal multi-carrier modulation and demodulation systems have very high requirements for synchronization, and inaccurate symbol timing will directly affect the performance of the entire system. The purpose of symbol timing is to find the starting position of the demodulation window, and a special training synchronization sequence is used for symbol timing.

In the present invention, a training synchronization sequence is added at the beginning of each frame. The synchronization sequence contains two identical parts before and after in the time domain. Taking a sine wave signal or a square wave signal as half of the synchronization symbol is equivalent to a sine wave or square wave containing two cycles in a synchronization symbol, and the frequency of the sine wave or square wave is generally between 300Hz and 500Hz. At the receiving end, correlation calculations are performed on the received two halves of sample points and accumulated, and the peak point is the synchronization start point. The sending end adds synchronization symbols to each encrypted orthogonal multi-carrier frame, but at the receiving end, the synchronization symbols must be removed before data demodulation.

In the end-to-end encryption method for anti-vocoder compression described in the present invention, all modules can be implemented on the DSP through software, and finally form a single-chip encryption and decryption module. Considering the complexity of the low-bit-rate compression module and the low-bit-rate decompression module, the low-bit-rate compression module and the low-bit-rate decompression module can also be replaced by ready-made chips. End-to-end speech encryption transmission process against vocoder.

The low code rate compression module of the present invention adopts the compression algorithm of 2.4Kbps code rate, also can adopt lower compression algorithm such as 1.2Kbps to 1.6Kbps code rate in addition, after adding the forward error correction algorithm, the error of transmitting encrypted voice The lower the bit rate, the better the voice effect.

Beneficial effects: the innovation of the present invention lies in the use of orthogonal multi-carrier modulation transmission mode, combined with low bit rate voice compression, data encryption, synchronization and key management negotiation modules, to realize end-to-end encryption based on voice channel. Its advantages are:

1) It is convenient and quick to transmit encrypted voice on the voice channel of GSM, and does not need to open additional new services.

2) The real end-to-end encryption process can pass from mobile terminal to mobile terminal or wired terminal.

3) The processing delay is small, less than 100ms. High sound intelligibility and strong confidentiality.

4) Anti-compression, anti-filtering, anti-resampling.

5) Strong adaptability, can be applied to CDMA and 3G systems through appropriate modification.

6) The tariff can use the original voice tariff discount, which is simple and cheap.

7) It does not affect the core network architecture and has little impact on operators.

Description of drawings

The present invention will be further described below in conjunction with the accompanying drawings and embodiments.

FIG. 1 is a general block diagram of an end-to-end speech encryption device and method against vocoder compression, describing a connection diagram of each module in a one-way end-to-end encryption and decryption process.

Fig. 2 is a time-domain waveform diagram of orthogonal multi-carrier modulation after synchronization is inserted.

Fig. 3 is a schematic diagram of quadrature multi-carrier modulation combined with a carrier generator module.

Fig. 4 is a schematic diagram of OFDM-based orthogonal multi-carrier modulation.

Detailed ways

In order to further illustrate the above-mentioned purpose, technical solution and effect of the present invention, the present invention will be described in detail below through examples in conjunction with the above-mentioned figures.

A one-way encryption process of the anti-vocoder compression end-to-end voice encryption device is composed of a sending end encryption module and a receiving end decryption module. They are respectively combined with the mobile phone at the sending end and the mobile phone at the receiving end to realize the encryption and decryption process of the one-way communication voice. Two-way voice encryption needs to add a decryption module on the mobile phone at the sending end of the secure call and an encryption module at the mobile phone at the receiving end. The encryption module at the sending end is composed of a low code rate compression module, an encryption module, an orthogonal multi-carrier modulation module, a carrier generator module, a key management negotiation module and a synchronization module. The receiver decryption module is composed of an orthogonal multi-carrier demodulation module, a carrier generator, a decryption module, a low code rate decompression module, a key management negotiation module and a synchronization module.

The overall block diagram of the end-to-end speech encryption device and method against vocoder compression is shown in FIG. 1 . It is divided into a sending end encryption module and a receiving end decryption module. The processing steps of the encryption process are as follows:

a) The 8KHz that is taken off from the mobile phone A/D quantization encoding module 3, the 13bit PCM format voice signal is sent to the low bit rate compression module 4 of the encryption module 1 at the sending end. This PCM stream is a signal driven by a synchronous clock and consists of a clock line, a word synchronization line and a receive data line. For reverse transmission, there is also a sending data line. The low bit rate compression module 4 compresses the 104Kbps signal into a bit stream smaller than 2.4Kbps through the AMBE compression algorithm, which can be realized by using a separate low bit rate compression chip or through a DSP. Speech processing is generally processed according to 20ms as a frame, less than 2.4Kbps is to ensure that the compressed effective information within 20ms is less than 48bit, so as to ensure that the subsequent orthogonal multi-carrier modulation module only uses low-frequency carrier with a lower bit error rate to complete modulation. When the compressed bit rate is less than 2.4Kbps, such as 1.2Kbps or 1.6Kbps, a forward error correction algorithm can be added to improve the bit error rate performance of the subsequent orthogonal modulation transmission.

b) The code stream compressed by the low code rate encoding module 4 is sent to the encryption module 5 for encryption operation. Use the enhanced encryption standard AES algorithm for encryption, or use the RC4 bit stream encryption algorithm for encryption.

c) The encoded code stream is sent to the orthogonal multi-carrier modulation module 6, and performs orthogonal multi-carrier modulation together with the carrier generator module 7. In addition, OFDM-based orthogonal multi-carrier modulation processing can also be performed on the incoming code stream. The selected sub-carrier frequencies should be less than 1500 Hz and orthogonal to each other. The modulation mode adopts QPSK modulation to carry on.

d) The waveform modulated by the orthogonal multi-carrier is finally sent back to the RPE-LTP vocoder 10 in the mobile phone, then sent to the channel coding module 11 and other parts for channel processing, and finally sent out from the air.

e) The key management and negotiation module 8 and the synchronization module complete the functions of key negotiation, key distribution management and encrypted frame synchronization at both ends of the encrypted communication. The process of key exchange is completed by the two communicating parties through the Diff-Huffman key agreement mechanism. Synchronization is performed by adding a corresponding synchronization sequence in front of the modulated waveform to synchronize each encrypted frame.

The processing steps at the decryption end are as follows:

a) After the signal received from the mobile phone in the air is processed by the channel decoding module 12, etc., and then decoded by the RPE-LTP vocoder 13, it becomes an encrypted modulated 8KHz, 13bit PCM code stream. Send this code stream into the orthogonal multi-carrier demodulation module 14 .

b) The orthogonal multi-carrier demodulation module 14 is combined with the carrier generator module 15 to perform corresponding demodulation processing corresponding to the orthogonal multi-carrier modulation processing. Or by performing corresponding OFDM demodulation processing. Finally, the encrypted code stream is demodulated.

c) The demodulated encrypted code stream is decrypted by the decryption module 16 to obtain decrypted compressed code stream information.

d) The low bit rate decoding module 17 decodes the decrypted code stream sent over, and recovers the original 8KHz, 13bit voice PCM signal, which is then sent back to the D/A decoding module in the mobile phone, and finally passed through the speaker Restore the sound of the sending end.

At the sending encryption end, the signal modulated by the orthogonal multi-carrier forms an encrypted frame structure together with the synchronous signal and is transmitted. As shown in Figure 2, in a voice frame with a processing unit of 20ms, the first 4ms time is drawn to fill the synchronization signal, and two 500Hz sine waves or square waves can be used to fill in, and the amplitude is slightly larger than the statistical maximum amplitude of the latter signal. , and the 16ms after 20ms are filled with data after orthogonal multi-carrier modulation. During demodulation, the waveform of the last 2ms and the waveform of the first 2ms are subjected to sliding correlation calculation, and finally a highest peak value in the range of 20ms is obtained, which is the start of a synchronization head. And after 4ms from the synchronization header is the transmitted modulated data.

The core of the invention is the process of orthogonal multi-carrier modulation and demodulation. When performing modulation processing with the orthogonal subcarriers generated by the carrier generator, it is necessary to follow the steps shown in Figure 3:

a) First, select the corresponding number of carriers according to the code rate requirement of the previous compression coding and the QPSK modulation mode. Since in a 20ms period, 16ms is designed to transmit the modulated signal. Then the carrier spacing of the orthogonal subcarriers is 62.5 Hz. Consider that the selected carrier frequency needs to be less than 1500Hz (modulated sine wave bit error rate is low). So first determine the initial subcarrier frequency, and then determine other subcarriers, for example, set the input code stream requirement to 2.4Kbps. It means that there is 48bit information in 20ms. When using QPSK. If 24 sub-carriers are required, 24 sub-carriers with an interval of 62.5 Hz between 62.5 Hz and 1500 Hz are selected as effective carriers. When the previous code rate requirement is less than 2.4Kbps, less carriers need to be used for modulation transmission.

b) After the carrier is selected. Perform serial-to-parallel conversion on the previously input serial code stream, then QPSK modulate it onto each effective sub-carrier, and then perform combination processing to obtain the modulated signal waveform.

c) The receiving end performs demodulation processing on the received modulated signal on each subcarrier corresponding to the modulation.

d) Parallel-to-serial conversion is performed on the demodulated code stream to obtain a code stream signal before orthogonal multi-carrier modulation.

When the OFDM-based method is used for modulation and demodulation processing, as shown in Figure 4, the steps to be performed are as follows:

On the sender side:

a) First, as with the carrier generator quadrature multi-carrier modulation mode above, the number and frequency of subcarriers need to be selected first. According to the frequency interval is 62.5Hz. According to the requirements of the previous compressed code stream, select the number and frequency of effective subcarriers according to the QPSK modulation method. Also assume that the incoming code stream is 2.4Kbps. When using QPSK modulation, the selected orthogonal subcarriers are between 62.5Hz and 1500Hz. Subcarriers spaced at 62.5Hz.

b) The code stream less than 2.4Kbps compressed by the low bit rate compression module is first converted into a parallel code stream through serial-to-parallel conversion.

c) Perform QPSK modulation on the parallel code stream, and then correspond to the vector value before 128-point IFFT processing. In order to ensure that the complex IFFT operation result is a real number, among the vector points before the 128-point IFFT, the 1st point and the 65th point take the value of 0, and the compressed code stream is filled to the 2nd to the 25th point after QPSK modulation, and the 26th to the 25th point. The 64th point is filled with 0 value, and the 128th point is filled with 0 value. The 2nd to 64th points and the 66th to 128th points are conjugate-even symmetrical about the 65th point. Perform IFFT transformation on the filled vector to obtain 128 points of OFDM modulation vector data.

d) The data of these 128 points, according to the 8KHz sampling process, actually takes up 16ms. After the synchronization module, plus the 4ms synchronization sequence described above, it can be output to the RPE-LTP vocoder module, and then the subsequent processing is carried out for wireless transmission.

At the receiving end, OFDM demodulation can demodulate the compressed code stream. Before doing OFDM demodulation, it is necessary to do synchronization related processing through the synchronization module to find out the synchronization point, and then extract the 16ms information of the modulated data from the 20ms voice frame for OFDM demodulation processing.

Finally, the encryption module and decryption module formed by the present invention can be embedded in the mobile phone to realize end-to-end encrypted transmission, and the end-to-end encrypted mobile phone can also be customized according to this method.

Claims (9)

1, a kind of end-to-end voice encryption device of anti-vocoder compression, it is characterized in that, a one-way encryption process comprises sending end encryption module (1) and receiving end deciphering module (2); Sending end encryption module (1) It includes a low code rate compression module (4), an encryption module (5), an orthogonal multi-carrier modulation module (6), a first carrier generator module (7), a first key management negotiation module (8) and a first synchronization module Module (9); receiving terminal deciphering module (2) comprises orthogonal multi-carrier demodulation module (14), second carrier generator module (15), deciphering module (16), low code rate decompression module (17), The second key management negotiation module (18) and the second synchronous module (19); Wherein, the output end of the A/D coding module (3) that the input end of the low code rate compression module (4) connects the original sending mobile phone, low The output terminal of the code rate compression module (4) and the key management negotiation module (8) is connected to the input terminal of the encryption module (5), and the output of the encryption module (5), the carrier generator module (7), and the synchronization module (9) Termination is connected to the input end of the orthogonal multi-carrier modulation module (6), and the output end of the orthogonal multi-carrier modulation module (6) is connected to the input end of the first RPE-LTP vocoder (10) of the original sending mobile phone; the receiving end deciphers Module (2) is composed of orthogonal multi-carrier demodulation module (14), second carrier generator (15), decryption module (16), low code rate decompression module (17), second key management negotiation module (18 ) and the second synchronous module (19), wherein, the input end of the orthogonal multi-carrier demodulation module (14) is respectively connected to the second carrier generator (15), the second synchronous module (19) and the first receiving mobile phone The output end of two RPE-LTP vocoders (13), the output end of the orthogonal multi-carrier demodulation module (14) and the second key management negotiation module (18) are connected to the input end of the decryption module (16), and the decryption module The output terminal of (16) connects the input end of the low bit rate decompression module (17), and the output end of the low bit rate decompression module (17) is the input end of the D/A decoding module (20) receiving mobile phone. 2. a kind of voice encryption method of the end-to-end voice encryption device of anti-vocoder compression as claimed in claim 1, is characterized in that, adopts orthogonal multi-carrier modulation method to set up end-to-end transmission channel to realize encrypted voice communication; The linear PCM digital voice from the A/D quantization coding module (3) of the mobile phone is standard 8KHz sampling, 13bit quantized data. After the data is compressed by the low bit rate compression module (4), it first becomes low bit rate Compress the code stream; the compressed code stream is encrypted by the encryption module (5), and then sent to the orthogonal multi-carrier modulation module (6), and modulated with the carrier wave generated by the first carrier generator module (7) to form The transmission signal waveform that can penetrate the vocoder compression is sent back to the first RPE-LTP vocoder (10) of the baseband processing chip of the mobile phone for RPE-LTP compression encoding, and then sent to the channel coding module (11) for channel Encoding and other operations are finally sent through the antenna of the mobile phone; at the receiving end of the mobile phone, after the signal from the antenna is processed by the channel decoding module (12) and the second RPE-LTP vocoder (13), it is first sent to the decrypting module of the receiving end On the orthogonal multi-carrier demodulation module (14) of (2), carry out demodulation processing together with the local second carrier generator module (15); The demodulated signal is deciphered through deciphering module (16), sends low The code rate decompression module (17) restores the 8KHz sampling of linear PCM, the data of 13bit quantized digital voice, and finally sends it to the D/A decoding module (20) in the mobile phone to form a decrypted voice stream. 3. The voice encryption method of the end-to-end voice encryption device of anti-vocoder compression as claimed in claim 2, characterized in that, the low bit rate compression module (4) and the low bit rate decompression module (17) are paired for use , Compress the input 8KHz, 13bit PCM voice with AMBE encoding algorithm into a code stream lower than 2.4K, or decompress the compressed code stream into 8KHz, 13bit PCM voice signal with AMBE decoding algorithm. 4. The voice encryption method of the end-to-end voice encryption device of anti-vocoder compression as claimed in claim 2, characterized in that, the encryption module (5) and the decryption module (16) are paired for use, and the encryption and decryption algorithm in the module Adopt AES digital encryption algorithm of enhanced encryption standard, or RC4 algorithm based on bit stream encryption. 5. The speech encryption method of the end-to-end speech encryption device of anti-vocoder compression as claimed in claim 2, is characterized in that, the first carrier generator module (7), the second carrier generator module (15) and The orthogonal multi-carrier modulation module (6) and the orthogonal multi-carrier demodulation module (14) are combined to realize the orthogonal multi-carrier modulation transmission and demodulation process of the compressed voice stream; this orthogonal multi-carrier transmission process There are two modes of implementation; one is to directly modulate and demodulate the carrier generated by the first carrier generator module (7) or the second carrier generator module (15); another method is to use orthogonal frequency division Reuse algorithms for fast implementation. 6. The voice encryption method of the end-to-end voice encryption device which is resistant to vocoder compression as claimed in claim 2, characterized in that, when adopting the orthogonal multi-carrier modulation method to establish the end-to-end transmission channel, the orthogonal frequency division Multiplexing is used to realize fast orthogonal multi-carrier modulation and demodulation. The sub-carrier frequency used by the orthogonal multi-carrier is lower than 1500 Hz, and the modulation method on each sub-carrier adopts the QPSK modulation method. 7. The voice encryption method of the end-to-end voice encryption device of anti-vocoder compression as claimed in claim 6, wherein, when using OFDM to realize fast orthogonal multi-carrier modulation and demodulation, Firstly, the input code stream is serial-to-parallel converted, and then filled to the specified carrier point, while the other carrier points are filled with 0 values. At the same time, in order to ensure that the fast Fourier transform FFT calculates a real number, it is necessary to ensure that the fast Fourier transform FFT and The data before the fast inverse Fourier transform IFFT is based on the front and rear conjugate even symmetry of the central point; the transformation of the fast inverse Fourier transform IFFT and the fast Fourier transform FFT is used to realize the modulation and demodulation of orthogonal multi-carriers. 8. The voice encryption method of the end-to-end voice encryption device that is resistant to vocoder compression as claimed in claim 2, wherein the two ends of encrypted voice communication need to adopt the first key management negotiation module (8), the second The key management and negotiation module (18) realizes key synchronization and key update management at both ends of the communication; the key exchange process is completed by the two communication parties through the Dif-Huffman key agreement mechanism. 9. The voice encryption method of the end-to-end voice encryption device of anti-vocoder compression as claimed in claim 2, characterized in that, the two ends of the encrypted voice communication also need a synchronization module to realize the synchronization of encrypted frames; Add a training synchronization sequence at the beginning, the synchronization sequence contains two identical parts before and after in the time domain; take a synchronization symbol containing two periods of sine wave or square wave, the frequency of sine wave or square wave is between 300Hz and 500Hz at the receiving end, perform correlation calculations on the received sample points of the two halves and accumulate them, and the obtained peak point is the synchronization start point; the transmitting end adds a synchronization sequence to each encrypted orthogonal multi-carrier frame, At the receiving end, the corresponding synchronization sequence is removed.

Images