CN113270105B - A Voice-like Data Transmission Method Based on Hybrid Modulation - Google Patents

Abstract

The invention discloses a voice-like data transmission method based on hybrid modulation, comprising the following steps: performing serial-parallel conversion on an input data stream to obtain a segmented data stream of fixed length; within the optimal transmission bandwidth of a voice channel of a vocoder The frequency bands are divided, and different modulation methods are used for different frequency bands to achieve the minimum transmission bit error rate, and the bit error rate of each frequency band is the same; the data symbols are obtained by inverse fast Fourier transform; the blank symbols, synchronization The header symbol and the data symbol are framed together to obtain a voice-like signal frame; the receiving end receives the voice-like signal frame and demodulates it to obtain a data stream. Through the present invention, data can be modulated on the amplitude or phase parameters of multiple frequency points in the sub-band to generate a voice-like signal, which has the time-frequency characteristics of general voice, can be transmitted through a voice channel with a vocoder, and can be widely used. Secure communication for end-to-end voice channels.

Description

A Voice-like Data Transmission Method Based on Hybrid Modulation

technical field

The invention belongs to the technical field of designing speech-like modulation, and in particular relates to a speech-like data transmission method based on hybrid modulation.

Background technique

Voice communication is an indispensable way of daily information exchange, and the eavesdropping of mobile phones and telephones occurs frequently. After the Trojan horse is installed on the mobile phone, it is easy to eavesdrop on the user's call through the wireless network; the monitoring of the telephone voice can be achieved by paralleling the telephone lines. The VoIP service that has appeared in recent years can achieve a certain degree of secure transmission through IP packet encryption, but the voice is still in plaintext in the IP packet, which cannot achieve end-to-end secure transmission. Therefore, if the voice can be encrypted before the voice is transmitted, the leakage of voice information caused by third-party eavesdropping during the subsequent voice processing and transmission can be prevented from the root cause.

However, the data obtained by the digital encryption of the voice signal usually no longer have the characteristics of voice, which will cause difficulties for the voice channel to transmit these data. Because in most voice transmission systems, in order to save the bandwidth of the transmission channel, the voice will be compressed and encoded; the vocoder used in the traditional voice channel mostly uses the mixed parameter coding algorithm to process the voice, and transmits by extracting the model parameters of the voice. Since the encrypted voice data does not have the basic characteristics of the voice signal, the basic parameters of the voice signal will not be extracted when passing through the vocoder, or the data will be treated as noise, which will cause the loss of data information, and the receiving end will not be able to. Decrypt the received encrypted data to obtain the correct voice.

Therefore, the data obtained by digital encryption of the voice signal needs to be processed before being sent, so that it has voice characteristics and can pass through the voice channel smoothly. This is similar to the purpose of modulation. The modulation technique is to transform the digital signal into a certain form to suit the channel transmission. The general modulation method is to change a sinusoidal carrier signal to carry digital data with the typical characteristics of the carrier: amplitude, frequency and phase. However, most conventional data modems cannot be used in systems that use speech codec compression techniques. The main purpose of speech coding is to reduce the number of bits required to represent speech, while still maintaining acceptable speech quality when decoding and recovering, so that the human ear sounds almost the same as the original speech. Therefore, the vocoder defines the optimal frequency band for the speech data it transmits. At the same time, since the human ear's perception of sound frequency is non-linear, and the response of the vocoder channel to different frequency bands is also different, the indiscriminate treatment of the modulation method of the carrier frequency may make the frequency, phase and amplitude characteristics of the data carried by the modem output. Larger distortion occurs due to the difference before and after the sampling waveform or the incompatibility between the carrier frequency and the vocoder channel. Therefore, most modems cannot be directly applied to the channel using the voice codec. It is necessary to redesign a modulation and demodulation scheme, which can convert the digital signal into a voice signal with certain time-frequency characteristics, that is, a voice-like modulation and demodulation scheme. .

SUMMARY OF THE INVENTION

The purpose of the present invention is to provide a voice-like data transmission method based on hybrid modulation, so as to solve the problem that the existing modem cannot be directly applied to the channel using the voice codec, and the digital signal cannot be converted into a voice signal with certain time-frequency characteristics. question.

For solving the above-mentioned technical problems, the concrete technical scheme of the present invention is as follows:

A voice-like data transmission method based on hybrid modulation, comprising the following steps:

Step 1. Perform serial-parallel conversion on the input low-speed data stream to obtain n segments of fixed-length segmented data streams {d ₁ , 1 , d _{1, 2} , ..., d _{1, k} ; d ₂ , 1 , d _{2, 2} ,...,d _2,k ;...;dn _,1 ,dn _,2 ,...,dn _,k }, the amount of data bits contained in each data stream is represented by k;

Step 2. Select the transmission frequency band [f _L , f _H ] of the vocoder channel, where f _L is the lowest frequency allowed to pass by the vocoder, and f _H is the highest frequency allowed to pass by the vocoder. According to different modulation methods, the transmission frequency band [f _L , f _H ] of the vocoder channel is divided into m frequency bands, each frequency band contains N _l frequency points, where l=1, 2,...,m , the frequency interval between adjacent frequency points is Δf, Δf≥1/T, where T is a data symbol period;

Step 3. In each frequency band, map the data into the phase and amplitude parameters of the frequency point according to the modulation mode of the frequency band, and construct the spectrum {X(p)} of the data symbol;

Step 4. Perform inverse fast Fourier transform on the frequency-domain data of the constructed data symbols to obtain the time-domain waveform {x(q)} of the data symbols, and frame n data symbols together with blank symbols and synchronization symbols to obtain Voice-like signal, each group of which completely includes blank symbols, synchronization symbols and data symbols is called a frame of voice-like signal frames;

Step 5: Receive the voice-like signal at the demodulation end, perform frame synchronization and voice-like demodulation operations, and obtain a data stream.

Further, the step used to divide the frequency band is to select the modulation mode in each frequency band, so that the data carried by each frequency band can obtain the same bit error rate after transmission;

The selection of the modulation mode of each frequency band is determined by the maximum number of phases N modulated at each frequency point, which satisfies:

N＜f _s /(M·f(i))

Where f _s is the signal sampling rate, f(i) is the frequency of the frequency point, i=1, 2,..., N _l , and M is the number of sampling point offsets;

When f _s and M are fixed, the transmission frequency band is divided into m frequency bands, and the number of phases of frequency modulation in each frequency band is less than N.

Further, in the step 3, if the phase of each frequency point can carry j-bit data, it can be mapped into 2 ^j optional phase parameters; the mapping of data and amplitude adopts every two frequency points in the frequency band as a group. The relative amplitude modulation method of , the amount of data carried by each frequency band through amplitude modulation is N _l /2 bits, and N _l is the number of frequency points included in each frequency band, where l=1, 2, ..., m.

Further, in the step 4, the frame structure of the signal frame includes n _b blank symbols, a synchronization header symbol and n data symbols, and each data symbol carries k-bit data; when conducting confidential communication, the signal frame also contains The key sequence number of the corresponding key stream.

A kind of speech-like data transmission method based on hybrid modulation of the present invention has the following advantages:

1. The speech-like signal obtained by the present invention based on the hybrid modulation method of sub-band and multi-parameter has the time-frequency characteristics of general speech signals, and the frequency range is within the optimal transmission frequency band range of the vocoder, which can smoothly pass through the vocoder channel and has good performance. The ability to resist non-linear operations of speech compression and decompression;

2. The present invention divides the spectrum space into multiple segments, and selects the best modulation mode in each segment, so that each frequency band can obtain the same minimum bit error rate, while maximizing the bit amount of data symbols carried;

3. The framing and de-framing speed of the present invention is fast, the transmission of low-speed data is not affected, the modulation and demodulation delay is small, and real-time transmission can be realized.

Description of drawings

1 is a schematic flowchart of a method for transmitting voice-like data based on hybrid modulation of the present invention;

Fig. 2 is the amplitude spectrum schematic diagram of the single data symbol of the present invention;

Fig. 3 is the phase spectrum schematic diagram of the single data symbol of the present invention;

4 is a schematic diagram of a frame structure format of a signal frame of the present invention;

FIG. 5 is a schematic diagram of the autocorrelation characteristic of the synchronization symbol of the present invention.

Detailed ways

In order to better understand the purpose, structure and function of the present invention, a hybrid modulation-based speech-like data transmission method of the present invention will be described in further detail below with reference to the accompanying drawings.

As shown in FIG. 1, it is a schematic overall flow diagram of a hybrid modulation-based voice data transmission method proposed by the present invention, and the method specifically includes the following steps:

Step 1. Serial-to-parallel conversion: perform serial-to-parallel conversion on the input low-speed data stream to obtain n segments of fixed-length segmented data streams {d ₁ , 1 , d _{1, 2} , ..., d _{1, k} ; d _{2 , 1} , d ₂ , 2 , ..., d _{2, k} ; ...; d _{n, 1} , d _{n, 2} , ..., d _{n, k} }, each data stream contains The amount of bits is represented by k, and k=30 bits;

Step 2, frequency band division: select the optimal transmission frequency band [f _L , f _H ] of the vocoder channel, where f _L is the lowest frequency allowed by the vocoder, and f _H is the highest frequency allowed by the vocoder. The transmission frequency band [f _L , f _H ] of the vocoder channel is divided into m frequency bands according to the different modulation modes that can be selected, and each frequency band contains N _l frequency points, where l=1, 2, .. ., m, the frequency interval between adjacent frequency points is Δf, which is not less than the reciprocal of a data symbol period T, that is: Δf≥1/T;

Specifically, the best transmission frequency band of a common vocoder channel is [300Hz, 3400Hz], select multiple effective frequency points within this frequency band, set the number of effective frequency points to be N _f =30, and the initial frequency to be f _L =300Hz , the highest frequency is f _H = 3400Hz, and the frequency interval Δf = 100Hz, then the frequency of each frequency point is f(i):

f(i)=f ₀ +i*Δf, i=0, 1, 2, ..., 30-1

Where f ₀ is the starting frequency, generally equal to f _L ; i is the frequency point position, and Δf is the frequency interval between adjacent frequency points.

Each data frame contains the sum of subcarriers corresponding to each frequency point after speech-like modulation.

Further, let the number of sampling points per cycle of the subcarrier corresponding to a single frequency point be n _T =f _s /f(i), f _s represents the sampling rate, and the phase offset caused by the offset of a single sampling point is

Assuming that the sampling point offset of M points occurs when the receiving end performs frame synchronization, and the number of modulated phases of the subcarrier corresponding to the frequency point f(i) is N, the conditions for demodulation error-free are:

Substituting into the above formula, it can be obtained that the number of phases N that can be modulated at each frequency point satisfies:

N＜f _s /(M·f(i))

That is, the number of modulated phases of the subcarrier corresponding to each frequency point is related to the frequency of the frequency point and the accuracy of frame synchronization. When the sampling rate f _s and the number of sampling point offsets M are constant, the frequency of The spectrum is divided into multiple frequency bands, and each frequency band uses a different multi-ary phase shift keying (N-Phase Shift Keying, N-PSK) modulation method according to the number of adjustable phases. When the adjustable phase number N is less than 2, phase modulation is no longer used, that is, other modulation methods such as amplitude shift keying are considered in the high frequency band.

When M=2, the transmission frequency band can be divided into four frequency bands in theory. In order to reduce the complexity of the algorithm and have the transmission effect at the same time, in practical application, three frequency bands are divided: [300Hz, 1000Hz], [1100Hz, 1600Hz], [1700Hz] , 3200Hz]. Fig. 2 and Fig. 3 are schematic diagrams of amplitude frequency and phase frequency of a certain data symbol in this case. The number of frequency points included in frequency band 1 is N ₁ =8 frequency points, using Quadrature Phase Shift Keying (Quadrature Phase Shift Keying, QPSK) modulation mode, the data amount S ₁ (f) = 2*N ₁ =16 bits in the first frequency band; the number of frequency points N ₂ =6 frequency points included in the second frequency band, and the mixed use of packet relative amplitude modulation and binary phase shift Keying (Binary Phase Shift Keying, BPSK) modulation mode, the data amount carried in the second frequency band S ₂ (f)=N ₂ =6 bits; the number of frequency points included in the third frequency band N ₃ =16 frequency points, only grouping relative In the amplitude modulation mode, the phase of each frequency point is kept as 0, and the data amount S ₃ (f)=N ₃ /2=8 bits carried in the third frequency band. The sum of the amount of data contained in the three frequency bands shall satisfy:

S(f)=S ₁ (f)+S ₂ (f)+S ₃ (f)=2*N ₁ +N ₂ +N ₃ /2=30 (bits)

So far, the data amount of k=30 bits contained in each data stream is allocated to the spectrum of its corresponding data symbol.

Step 3, data-spectrum parameter mapping: in each frequency band, according to the modulation mode of the frequency band, the data is mapped into the phase and amplitude parameters of the frequency point, and the spectrum {X(p)} of the data symbol is constructed;

Step 4. Voice-like generation: Inverse fast Fourier transform is performed on the frequency domain data of the constructed data symbols to obtain the time domain waveform {x(q)} of the data symbols, and n data symbols are combined with blank symbols and synchronization symbols. Group frames together to obtain speech-like signals, and each group that completely includes blank symbols, synchronization symbols and data symbols is called a frame of speech-like signal frames;

Specifically, referring to the schematic diagram of FIG. 4 , it is the frame structure format of each speech-like signal, which consists of n _b =6 blank symbols, 1 synchronization header symbol and n = 32 data symbols, and each signal frame contains a total of n _b +n+1=39 symbols, except for the sync header symbol, the time of each symbol is T=10 ms, and the duration of the sync header is 20 ms, so the time of each signal frame is T _all =T*(n _b +n)+2*T=400 milliseconds.

Further, the amount of data transmitted by the data frame is S(t)=n*k=960 bits, so the total data transmission rate v(t) is:

Further, the synchronization symbol should be a fixed time sequence with strong autocorrelation, such as chirp signal, whose time length is 2*T=20 milliseconds, and the instantaneous frequency changes with time in the range of [300Hz, 3400Hz]. and change. Referring to the schematic diagram of FIG. 5, that is, the schematic diagram of the autocorrelation characteristic of the synchronization symbol, only at the non-lag and non-advanced time "0", the synchronization symbol shows strong correlation, and the correlation at other times is very weak; according to this characteristic, the receiving end The same synchronization symbol sample data can be used to perform sliding correlation, and the position of the data symbol in the received speech-like signal can be precisely located according to the set correlation threshold.

Step 5. Voice-like demodulation: receive a voice-like signal at the demodulation end, perform frame synchronization and voice-like demodulation operations, and obtain a low-speed data stream.

Specifically, the bit error rate of data transmission in each frequency band measured at the demodulation end is: frequency band 1 5‰, frequency band 2 5‰, frequency band 3 5‰, the three frequency bands reach the same minimum bit error rate, so that The overall bit error rate is the smallest at 5‰.

The voice-like data transmission method based on hybrid modulation proposed by the present invention can be used to transmit encrypted voice data, so that when intelligible voice is changed into incomprehensible voice, the voice characteristics are also retained. The data can be transmitted smoothly through the voice compression coding channel.

It can be understood that the present invention is described by some embodiments, and those skilled in the art know that various changes or equivalent substitutions can be made to these features and embodiments without departing from the spirit and scope of the present invention. In addition, in the teachings of this invention, these features and embodiments may be modified to adapt a particular situation and material without departing from the spirit and scope of the invention. Therefore, the present invention is not limited by the specific embodiments disclosed herein, and all embodiments falling within the scope of the claims of the present application fall within the protection scope of the present invention.

Claims (4)

1. a kind of speech data transmission method based on hybrid modulation, is characterized in that, comprises the following steps: Step 1. Perform serial-parallel conversion on the input low-speed data stream to obtain n segments of fixed-length segmented data streams {d ₁ , 1 , d _{1, 2} , ..., d _{1, k} ; d ₂ , 1 , d _2,2 ,...,d _2,k ;...;dn _,1 ,dn _,2 ,...,dn _,k }, the amount of data bits contained in each data stream is k means; Step 2. Select the transmission frequency band [f _L , f _H ] of the vocoder channel, where f _L is the lowest frequency allowed by the vocoder, and f _H is the highest frequency allowed by the vocoder; The transmission frequency band [f _L , f _H ] of the vocoder channel is divided into m frequency bands, and each frequency band contains N _l frequency points, where l=1, 2, ..., m, the number of adjacent frequency points is The frequency interval between is Δf, Δf ≥ 1/T, where T is a data symbol period; Step 3. In each frequency band, map the data into the phase and amplitude parameters of the frequency point according to the modulation mode of the frequency band, and construct the spectrum {X(p)} of the data symbol; Step 4. Perform inverse fast Fourier transform on the frequency-domain data of the constructed data symbols to obtain the time-domain waveform {x(q)} of the data symbols; framing n data symbols, blank symbols and synchronization symbols together to obtain Speech-like signal; each group completely contains a combination of blank symbols, synchronization symbols and data symbols, which is called a frame of speech-like signal frames; Step 5: Receive the voice-like signal at the demodulation end, perform frame synchronization and voice-like demodulation operations, and obtain a data stream. 2. the class voice data transmission method based on hybrid modulation according to claim 1, is characterized in that, in described step 2, the step that is used to divide frequency band is to let each frequency band select modulation mode, so that each frequency band carries data, get the same bit error rate after transmission; The selection of the modulation mode of each frequency band is determined by the maximum number of phases N modulated at each frequency point, which satisfies: N＜f _s /(M·f(i)) Where f _s is the signal sampling rate, f(i) is the frequency of the frequency point, i=1, 2,..., N _l , and M is the number of sampling point offsets; When f _s and M are fixed, the transmission frequency band is divided into m frequency bands, and the number of phases of frequency modulation in each frequency band is less than N. 3. the speech-like data transmission method based on hybrid modulation according to claim 1, is characterized in that, in described step 3, it is assumed that the phase of each frequency point can carry j-bit data, then can be mapped into 2 ^j can be Select the phase parameter; the mapping of data and amplitude adopts the relative amplitude modulation method in which every two frequency points in the frequency band are a group, and the amount of data carried by each frequency band through amplitude modulation is N _l /2 bits. 4. the speech-like data transmission method based on hybrid modulation according to claim 1, is characterized in that, in described step 4, the frame structure of signal frame comprises n _b blank symbols, a synchronization header symbol and n data symbols , each data symbol carries k-bit data; when performing secure communication, the signal frame also includes the key sequence number of the corresponding key stream.

Images