CN101594123B - Method for establishing equivalent parallel filter, voice transmission method, device and system - Google Patents

Abstract

The invention relates to a method for establishing an equivalent parallel filter, a voice data transmission method, the equivalent parallel filter and a voice data transmission system. The method for establishing the equivalent parallel filter comprises the following steps: establishing a channel model of a cascade vocoder; performing time domain analysis and frequency domain analysis on the channel of the cascade vocoder to obtain accurate amplitude-frequency response and phase-frequency response; making the channel of the cascade vocoder equivalent to parallel connection of a weak noise interference low pass filter and a strong interference bandpass filter according to amplitude-frequency response characteristics and phase-frequency response characteristics; and making the voice coding and decoding process of the channel of the cascade vocoder equivalent to a parallel model for performing low pass filtering and band pass filtering on signals. The method solves the problem that encrypted voice is transmitted in a channel of mobile voice service, provides basis for designing a voice-band quasi-stable modem, and improves the reliability and effectiveness of transmitting encrypted voice data by the system.

Description

Set up method, speech transmissions method, the Apparatus and system of equivalent parallel filter

Technical field

The present invention relates to Digital Signal Processing, particularly relate to a kind of method, speech transmissions method, Apparatus and system of setting up equivalent parallel filter.

Background technology

Along with development of Communication Technique, mobile communication has become the part of people's daily life, and the user more pays close attention to the safety problem of communication at special occasions when paying close attention to communication quality.Because there is defective in the security mechanism of existing GSM at aspects such as cryptographic object, encryption scope, Cipher Strength, encryption power to make decision; Cause wireless interception or the incident of wiretaping to occur repeatedly; Cause certain loss to the user,, encrypt before and proposed to get into GSM at voice signal in order to solve the safety problem of speech; But; Data encrypted presents stochastic behaviour, no longer possesses the characteristic of voice signal, can't be directly in original mobile voice service channel; Thereby proposed the transmission plan of encrypted voice in mobile network, mainly comprised two kinds at present: based on the data traffic channels transmission with based on the voice service Channel Transmission.

Wherein, for based on the data traffic channels transmission plan, feed back retransmission mechanism because the mobile data services channel adopts, end-to-end delay is bigger, is difficult to satisfy the real-time Transmission requirement of voice service.

For transmission plan, be subject to the performance of mobile vocoder and the mode of operation of cascade based on the voice service channel.In this process, move vocoder and adopt the speech parameters coding, only extract and transmit the parameter that characterizes the voice signal key feature, receiving terminal synthesizes speech through these parameters.Although the synthetic speech of receiving terminal sounds closely similar with the input voice signal, maybe be different fully with the original speech waveform on the waveform sampling point.In addition; Mobile voice service adopts cascade vocoder (Tandem) mode of operation; Be that transmit leg is encoded in mobile terminal MS, base station BTS place decoding (compiles separates); The cascade vocoder mode of operation of recipient's decoding in base station BTS place coding, mobile terminal MS (compiles separates) makes voice signal in network, will introduce distortion inevitably through four times (GSM-GSM connection) or twice (GSM-PSTN connection) code conversion.Therefore; Cascade vocoder in the mobile voice service channel becomes the bottleneck of this transmission plan, needs to seek one type of ability high efficient and reliable transmission and can realize the analog signal (being called the steady vocal band signal of standard) with the reliable conversion of random data in the mobile voice service channel.

At present, type of mainly containing speech synthetic method and conventional modulator approach.Class speech synthetic method is regarded enciphered data as the speech coding parameter; Realize the conversion of data and analog signal through existing speech synthesis analysis method; Although the synthetic analog signal of this method is similar voice signal, modulation rate is lower, only can reach about 1200bps; The error rate is higher, is difficult to satisfy the zero error code requirement of receiving terminal DecryptDecryption module.Conventional modulation method adopts modulator approaches such as FSK, QPSK to realize the conversion of enciphered data and analog signal; But the synthetic analog signal of this method is difficult to satisfy the requirement of high efficient and reliable transmission in the mobile voice service channel; The cascade vocoder is bigger to the distortion of such analog signal; Demodulation bit error rate can't reach performance requirement usually greater than 10e-2.

In research and practice process to prior art, inventor of the present invention finds that in the existing implementation, encrypted voice can't satisfy the requirement of high efficient and reliable transmission in the mobile voice service channel in the transmission of mobile voice service channel.

Summary of the invention

The embodiment of the invention provides the transmission system of transmission method, equivalent parallel filter and the voice data of a kind of method of setting up equivalent parallel filter, voice data; To solve the problem that encrypted voice transmits in the mobile voice service channel, the raising system is to the efficiency of transmission of encrypted voice.

For solving the problems of the technologies described above, the present invention is that embodiment provides a kind of method of setting up equivalent parallel filter, and said method comprises:

Set up the channel model of cascade vocoder;

Channel to said cascade vocoder carries out time domain analysis and frequency-domain analysis, obtains amplitude-frequency response and phase-frequency response accurately;

Make an uproar a little less than according to said amplitude-frequency response characteristic and phase versus frequency response charac t being and disturb the parallel connection of low pass filter and strong jamming band pass filter the equivalence of said cascade vocoder channel;

To the speech coding and decoding processing procedure equivalence of said cascade vocoder channel the parallel model that signal is carried out LPF and bandpass filtering.

Preferably, said channel to the cascade vocoder carries out time domain analysis and frequency-domain analysis, and obtaining accurately, amplitude-frequency response and phase-frequency response specifically comprise:

Channel to said cascade vocoder carries out time domain analysis, obtains the description of wave distortion;

According to the description of said wave distortion the process errors equivalence of said cascade vocoder channel is made an uproar a little less than for low-frequency range and to disturb and the strongly disturbing parallelly connected noise source of high band;

Channel to said cascade vocoder carries out frequency-domain analysis, obtains the amplitude-frequency response characteristic and the phase versus frequency response charac t of cascade vocoder channel.

Preferably, the channel model of said cascade vocoder comprises: when global system for mobile communications GSM-GSM connects, and 4 cascade full rate vocoder GFR-4 channel models; And GSM-public switch telephone network PSTN is when connecting, the channel model of 2 cascade full rate vocoder GFR-2.

Preferably, the channel model of said GFR-4 comprises: the portable terminal of transmitting terminal and the vocoder of base station are right, and the vocoder of receiving terminal base station and portable terminal is right, and all links between transmitting terminal and the receiving terminal.

Preferably, the channel model of said GFR-2 comprises: the vocoder of portable terminal and base station is to all links of, Radio Link, core net transmission link, PSTN transmission link.

Preferably, the mode of the description of said wave distortion is: non-similar estimate to describe with similar estimate description.

Preferably, said method also comprises: construct the input signal that the channel model of said cascade vocoder can be discerned, utilize this input signal to encourage the channel model of said cascade vocoder and the output signal of recorder end at transmitting terminal.

The embodiment of the invention also provides a kind of transmission method of voice data, and said method comprises:

Equivalent parallel filter receives the encryption of transmitting terminal transmission and the voice data of modulation;

Equivalent parallel filter carries out low pass and bandpass filtering in voice data respectively, and filtered data are sent to receiving terminal, and to carry out demodulation successively close so that receiving terminal is to receiving said data, obtains former voice data.

Preferably, said method also comprises: set up equivalent parallel filter, said method such as each said method of claim 1 to 7 of setting up equivalent parallel filter.

Accordingly, the embodiment of the invention also provides a kind of device of setting up equivalent parallel filter, comprising:

Set up the unit, be used to set up the channel model of cascade vocoder;

Analytic unit is used for the said channel of setting up the cascade vocoder of setting up the unit is carried out time and frequency domain analysis, obtains amplitude-frequency response and phase-frequency response accurately;

Make an uproar a little less than with phase versus frequency response charac t the being parallel connection of interference low pass filter and strong jamming band pass filter of first equivalent unit, the amplitude-frequency response characteristic that is used for obtaining with the equivalence of said cascade vocoder channel according to said analytic unit;

Second equivalent unit, being used for said speech coding and decoding treatment process equivalence of setting up the cascade vocoder channel of setting up the unit is the parallel model that signal is carried out LPF and bandpass filtering.

Preferably, said analytic unit comprises:

First analytic unit is used for the said channel of setting up the cascade vocoder of setting up the unit is carried out time domain analysis, obtains the description of wave distortion;

C grade is imitated the unit, and the description of the wave distortion that is used for obtaining according to said first analytic unit is made an uproar the process errors equivalence of said cascade vocoder channel a little less than for low-frequency range and disturbed and the strongly disturbing parallelly connected noise source of high band;

Second analytic unit is used for the said channel of setting up the cascade vocoder of setting up the unit is carried out frequency-domain analysis, obtains the amplitude-frequency response characteristic and the phase versus frequency response charac t of cascade vocoder channel.

Preferably, the said channel model of setting up the cascade vocoder of setting up the unit specifically comprises: the channel model of 4 cascade full rate vocoder GFR-4 and 2 cascade full rate vocoder GFR-2.

Preferably, also comprise:

Structural unit is used to construct the input signal that the channel model of said cascade vocoder can be discerned.

Accordingly, the embodiment of the invention also provides a kind of equivalent parallel filter, comprising: receiving element, equivalent low-pass filter, equivalent bandpass filter and transmitting element, wherein

Said receiving element is used for the process encryption of receiving end/sending end transmission and the voice data after the modulation;

Equivalent low-pass filter is used for the voice data that receives is carried out filtering, and filtered voice data is sent to transmitting element;

Equivalent bandpass filter is used for the voice data that receives is carried out filtering, and filtered voice data is sent to transmitting element;

Said transmitting element is used for reception equivalent low-pass filter and the filtered voice data addition of equivalent bandpass filter are redispatched to receiving terminal.

Accordingly, the embodiment of the invention also provides a kind of transmission system of voice data, comprising: receiving terminal, equivalent parallel filter and receiving terminal, wherein,

Said transmitting terminal be used for the voice data that receives is compressed successively, encrypts, encodes and modulates, and the data after will modulating is input to equivalent parallel filter;

Said equivalent parallel filter is used for carrying out filtering to receiving voice data, and filtered voice data is sent to receiving terminal;

Said receiving terminal is used for modulating successively receiving said data, decodes, deciphering and decompress(ion), obtains former voice data.

Preferably, said equivalent parallel filter equivalent parallel filter as claimed in claim 14.

Can know by technique scheme; The embodiment of the invention not only can be held the connected mode of GSM end to GSM; Can also hold the connected mode of PSTN end to GSM; Make up GFR-4 and GFR-2 channel model respectively, and channel model is carried out the identification analysis, obtain amplitude-frequency response and phase-frequency response accurately from time domain and two angles of frequency domain; The equivalence of 0-1550Hz low-frequency range part is one and receives the low pass filter that is easy to utilize of " small noise interference ", the equivalence of 1550-4000Hz high band part is become one and receives the band pass filter that is difficult to utilize of " strong random disturbances ".Related with having set up between cascade vocoder and low pass, the band pass filter thus, can be the accurate steadily design of modulator-demodulator of voice band foundation is provided, thereby solve the problem that encrypted voice transmits in the mobile voice service channel, thereby improve the efficiency of transmission of system.

Description of drawings

A kind of flow chart of setting up the method for equivalent parallel filter of Fig. 1 for providing in the embodiment of the invention;

Fig. 2 is the structural representation of the GFR-4 black box channel model that provides in the embodiment of the invention;

Fig. 3 is the structural representation of the GFR-2 black box channel model that provides in the embodiment of the invention;

Fig. 4 is the GFR-4 channel that provides in the embodiment of the invention output time domain waveform figure to different frequency single frequency sinusoidal signal;

Fig. 5 is the GFR-2 channel that provides in the embodiment of the invention output time domain waveform figure to different frequency single frequency sinusoidal signal;

Fig. 6 is the time domain analysis adopting in the embodiment of the invention Euclidean distance to estimate to estimate with cosine angle figure as a result;

Fig. 7 is 600Hz single-frequency cosine input that provides in the embodiment of the invention and full phase place logarithm amplitude spectrum and the phase place spectrogram of exporting signal;

Fig. 8 is the output phase shift and the normalized amplitude of GFR-4 and GFR-2 channel in the 0-4000Hz band limits that provides in the embodiment of the invention;

Fig. 9 is the GFR-4 that provides in the embodiment of the invention and the equivalent parallel filter structure of models sketch map of GFR-2 black box system;

Figure 10 is the ber curve figure of each data subcarrier of providing in the embodiment of the invention;

Figure 11 is the flow chart of the transmission method of a kind of voice data of providing in the embodiment of the invention;

A kind of structural representation of setting up the device of equivalent parallel filter of Figure 12 for providing in the embodiment of the invention;

The structural representation of a kind of equivalent parallel filter that Figure 13 provides for the embodiment of the invention;

The structural representation of the transmission system of a kind of voice data that Figure 14 provides for the embodiment of the invention.

Embodiment

It is a kind of based on global system for mobile communications (GSM that the embodiment of the invention provides; Global System forMobile communications) voice service channel cascade vocoder is set up the transmission system of transmission method, equivalent parallel filter and voice data of method, the voice data of equivalent parallel filter; This scheme is held the connected mode of GSM end to GSM; Such as; Can make up the GFR-4 black box channel model (following abbreviation GFR-4 channel) that forms by 4 full rate vocoder cascades, hold the connected mode of public switch telephone network (PSTN, Public Switched Telephone Network) end to GSM; Can make up the GFR-2 black box channel model (following abbreviation GFR-2 channel) that forms by 2 full rate vocoder cascades; Then, channel model is carried out the identification analysis, obtain amplitude-frequency response and phase-frequency response accurately from time domain and two angles of frequency domain; The equivalence of 0-1550Hz low-frequency range part is one and receives the low pass filter that is easy to utilize of " small noise interference ", the equivalence of 1550-4000Hz high band part is become one and receives the band pass filter that is difficult to utilize of " strong random disturbances ".Therefore; Related with having set up between cascade vocoder and low pass, the band pass filter; Can be the accurate steadily design of modulator-demodulator of voice band foundation is provided, thereby solve the problem that encrypted voice transmits in the mobile voice service channel, improve the reliability and the validity of system.

In the present embodiment, for the ease of understanding and describing, GFR-4 black box channel model or GFR-2 black box channel model are called equivalent parallel filter.

In order to make those skilled in the art person understand the technical scheme in the embodiment of the invention better; And make the above-mentioned purpose of the embodiment of the invention, feature and advantage can be more obviously understandable, below in conjunction with accompanying drawing technical scheme in the embodiment of the invention done further detailed explanation.

See also Fig. 1, a kind of flow chart of setting up the method for equivalent parallel filter for providing in the embodiment of the invention is applicable to the transmission of GSM voice service, and said method comprises:

Step 101: the channel model of setting up the cascade vocoder;

In the present embodiment, can hold the connected mode of GSM end to GSM, equipment and link equivalence that voice signal is inputed between the output are a black box system that regards transfer function the unknown as, set up GFR-4 black box channel model; This black box system comprises: the portable terminal of transmitting terminal and the vocoder of base station are right to the vocoder of the base station of, receiving terminal and portable terminal, and all links between transmitting terminal and the receiving terminal.The structural representation of its GFR-4 black box channel model sees Fig. 2 for details.

Also can hold the connected mode of PSTN end to GSM, voice signal is inputed to equipment and the equivalent black box system for another transfer function the unknown of link between the output, set up GFR-2 black box channel model.This black box system comprises: the vocoder of portable terminal and base station is to all links of, Radio Link, core net transmission link, PSTN net transmission link.The structural representation of its GFR-2 black box channel model sees Fig. 3 for details.

Step 102: the channel to said cascade vocoder carries out time domain analysis and frequency-domain analysis, obtains amplitude-frequency response and phase-frequency response accurately; In this step, can be directed against GFR-4 black box channel model, also can be directed against GFR-2 black box channel model, specifically comprise:

1) channel to said cascade vocoder carries out time domain analysis, obtains the description of wave distortion;

Wherein, The mode of the description of said wave distortion is: non-similar estimate to describe with similar estimate description; Specifically comprise: Euclidean distance, inner product, cosine angle, Tanimoto estimate, and other available non-phase Sihes are similar estimates describing method, and present embodiment does not limit.

2) according to the description of said wave distortion the process errors equivalence of said cascade vocoder channel is made an uproar a little less than for low-frequency range and disturb and the strongly disturbing parallelly connected noise source of high band; Promptly between the sum of errors noise jamming intensity that cascade vocoder coding and decoding is introduced, set up relevance;

3) channel to said cascade vocoder carries out frequency-domain analysis, obtains the amplitude-frequency response characteristic and the phase versus frequency response charac t of cascade vocoder channel.

Step 103: make an uproar a little less than according to said amplitude-frequency response characteristic and phase versus frequency response charac t being and disturb the parallel connection of low pass filter and strong jamming band pass filter with the equivalence of said cascade vocoder channel;

In this step, can be directed against GFR-4 black box channel model, also can be directed against GFR-2 black box channel model,

Step 104: will be the parallel model that signal is carried out LPF and bandpass filtering to the speech coding and the decoding treatment process equivalence of said cascade vocoder channel.

In this step, can be directed against GFR-4 black box channel model, also can be directed against GFR-2 black box channel model.

Preferably, said method can also comprise: construct the input signal that the channel model of said cascade vocoder can be discerned, utilize this input signal to encourage the channel model of said cascade vocoder and the output signal of recorder end at transmitting terminal.Such as, can utilize this input signal at transmitting terminal excitation GFR-4 black box channel model or GFR-2 black box channel model, and the output signal of difference recorder end.

Through above-mentioned analysis result to time domain and frequency domain; Can verify the consistent description of voice channel, that is to say, owing to above-mentioned result to time domain analysis and frequency-domain analysis can know; Correlation between low-frequency range output signal and the input signal is bigger, and the correlation of high band is less.In other words, GFR-4 black box channel, perhaps GFR-2 black box channel is less to the sinusoidal signal influence of low-frequency range, and the signal waveform distortion is little, and very big to the high frequency band signal influence, wave distortion is big.

Below through the concrete identification instance of GSM full rate vocoder GFR-4 channel model is explained implementation procedure of the present invention.

1, is configured to the input signal of Model Distinguish;

Generally speaking, the voice band frequency range is 300-3400Hz.Be without loss of generality structure cosine function x (t)=A in 0～4000Hz frequency range _xCos (2 π ft+ θ ₀), wherein, signal frequency f presses the 50Hz frequency step, traversal 0～4000Hz frequency range; Get initial phase θ ₀=300, get vocoder valid data peaked half the signal amplitude, i.e. A _x=16376.With sample frequency f _s=8000Hz samples and quantizes, and obtains digital signal x (n)=A _xCos (2 π nf/f _s+ θ ₀).

This step is an optional step.

2, time domain Transmission characteristics;

With input signal x (n)=A _xCos (2 π nf/f _s+ θ ₀) excitation GFR-4 channel; Recorder end output signal y (n); Wherein the GFR-4 channel is to the output time domain waveform figure of different frequency single frequency sinusoidal signal, and as shown in Figure 4, present embodiment is an example with frequency 50Hz, 300Hz, 600Hz, 1000Hz, 1200Hz and 1600Hz.Wherein abscissa is represented sampling point, and ordinate is represented the normalization amplitude.Can know that by this figure the GFR-4 channel is different to the response of different frequency input signal, and system is also different to the time of stable state transition by transient state.For low frequency signal, there is bigger distortion in the 1-2 frame signal, and system is in transient state, and the signal distortion later on of the 3rd frame is very little, and system is tending towards stable state.For higher frequency signals, system's transient state time extends to the 3-4 frame, is tending towards stable state then.

In like manner, Fig. 5 is the output waveform figure of GFR-2 channel to different frequency single frequency sinusoidal signal.Can be known that by this figure the process of itself and GFR-4 channel is similar, promptly the GFR-2 channel is different to the response of different frequency input signal, and system is also different to the time of stable state transition by transient state.For low frequency signal, there is bigger distortion in the 1-2 frame signal, and system is in transient state, and the signal distortion later on of the 3rd frame is very little, and system is tending towards stable state.For higher frequency signals, system's transient state time extends to the 3-4 frame, is tending towards stable state then.

In order to further describe the time domain transmission characteristic of GFR-4 channel to the different frequency cosine signal, present embodiment is that example is assessed the correlation between output signal and the input signal to adopt similarity measure.Present embodiment is that example is explained with Euclidean distance, cosine angle:

At first, establish vector x=(x ₁, x ₂..., x _l), vectorial y=(y ₁, y ₂..., y _l).

(1) Euclidean distance:

$d_2(x,y)=\sqrt{{(x-y)}^T(x-y)}=\sqrt{\underset{i=1}{\overset{l}{\mathrm{\Σ}}}{(x_i-y_i)}^2}---\left(1\right)$

d ₂(x, y)>=0, and vector and the distance between itself equal 0, and correlation is maximum.Along with the increase of value, the correlation between the vector reduces.

(2) cosine is similar estimates:

$S_{\cos \mathrm{ine}}(x,y)=\frac{x^{T}y}{\left|\right|x\left|\right|\·\left|\right|y\left|\right|}---\left(2\right)$

Wherein $\left|\right|x\left|\right|=\sqrt{\underset{i=1}{\overset{\mathrm{<figure-callout\; id="2"\; label="l\; x\; i"\; filenames="H200910139601XE00031.png,H200910139601XE00032.png"\; state="\{\{state\}\}">l</figure-callout>}}{\mathrm{\&Sigma;}}}{x}_i^{}{}_2^{}},$ $\left|\right|y\left|\right|=\sqrt{\underset{i=1}{\overset{l}{\mathrm{\&Sigma;}}}{y}_i^2}.$

S _Cosine(x y) has reflected angle between the vector, and value is big more, and the angle between the vector is more little, when angle is 0, gets maximum 1.

Secondly; Calculate Euclidean distance and the cosine angle of normalization stable state output y (n) and corresponding list entries x (n); Its result is the time domain analysis adopting Euclidean distance to estimate in the embodiment of the invention to estimate with cosine angle figure as a result respectively shown in the black in Fig. 6 (1) and 6 (2).Can be known that by Fig. 6 (1) and 6 (2) correlation between low-frequency range output signal and the input signal is bigger, the correlation of high band is less.In other words, the GFR-4 channel is less to the sinusoidal signal influence of low-frequency range, and the signal waveform distortion is little, and very big to the high frequency band signal influence, wave distortion is big.

In like manner, for the normalization stable state output y (n) of GFR-2 channel and Euclidean distance and the cosine angle of corresponding list entries x (n), its result is also shown in the light/dark balance of Fig. 6 (1) and 6 (2).

At last, get the less frequency range 0-1550Hz of wave distortion.0-1550Hz is defined as low-frequency range, and 1550-4000Hz is a high band.Because the distortion of waveform causes through the GFR-4 Channel Transmission; And the low-frequency range wave distortion is little, the high band wave distortion is big; Therefore, the embodiment of the invention is interference, the strongly disturbing parallelly connected noise source of high band of making an uproar a little less than the low-frequency range with the distortion equivalence of GFR-4 Channel Transmission introducing.In fact this noise source is exactly the coding of GSM FR cascade vocoder and coding and decoding error and the code conversion error that decoding is introduced, thereby has converted the error of cascade vocoder coding and decoding to interference noise source, and has obtained interference strength description qualitatively.

3, frequency domain response specificity analysis;

At first, the frequency response after through system is analyzed to a certain frequency sine wave, is without loss of generality selecting frequency f=600Hz.Input/output signal is carried out spectrum analysis, as shown in Figure 7, be the full phase place logarithm amplitude spectrum and the phase place spectrogram of input of 600Hz single-frequency cosine and output signal.In this Fig. 7, abscissa is a number of samples, and ordinate is the dB or the number of degrees.

From the amplitude spectrogram, can find out to have the harmonic component of many amplitudes for-50db in the amplitude spectrum of input signal x (n), this is cosine signal x (t) quantizes the back introducing through sample circuit a error.The peak value spectral line of input signal x (n) and output signal y (n) does not produce new frequency component all at the k=19 place, explains that the GFR-4 channel is a linear system.Wherein, the signal to noise ratio of x (n) is-100db, and the signal to noise ratio of y (n) is about-40db, and system receives noise jamming, and whole harmonic waves of x (n) are flooded by noise.

Can find out that from the phase place spectrogram the phase spectrum of signal x (t) and x (n) is the horizontal linear characteristic, in peak value spectral line (k=19) position and on every side the phase value of position of spectral line equals initial phase.The phase spectrum of y (n) then receives noise jamming, and only near the phase spectrum the peak value spectral line (k=19) can be thought the initial phase of signal approx.

Hence one can see that, and the GFR-4 channel has the frequency consistency to the 600Hz cosine signal, but the amplitude of signal and phase place all receive the influence of transmitted noise.In fact this transmitted noise is exactly the coding of GSM FR vocoder and coding and decoding error and the code conversion error that decoding is introduced.

Calculate phase deviation and the normalized amplitude that transmitted noise brings below.Select N=512, try to achieve first phase θ through full phase place APFFT spectral analysis method _z=30.00060, phase deviation Δ θ=0.00060, normalized amplitude ${\hat{A}}_z=0.99935.$

For other frequency f, calculate first phase, phase deviation and normalized amplitude respectively, its result shown in Fig. 8 (1) and 8 (2), is the output phase shift and the normalized amplitude of GFR-4 and GFR-2 channel in the 0-4000Hz band limits that provides in the embodiment of the invention respectively.Can be known that by Fig. 8 (1) and 8 (2) the normalization changes in amplitude of 50-1550Hz is less than 0.1, wherein the phase deviation of 500-1550Hz frequency range is less than 0.5 degree, and the phase deviation of 50-450 frequency range is less than 2 degree.

Hence one can see that, and the GFR-4 channel has certain " transparent " transmission characteristic for 0-1550Hz single frequency sinusoidal signal, can Approximate Equivalent be a low pass filter that is easy to utilize that disturbed by small noise; Producing very big " obstruction " for 1550-4000Hz single frequency sinusoidal signal, can Approximate Equivalent be a band pass filter that is difficult to utilize that disturbed by very noisy.Thereby is the parallel model that signal is carried out LPF and bandpass filtering with vocoder to the speech parameters extraction of input signal and the processing procedure equivalence of synthesizing; As shown in Figure 9, be the GFR-4 that provides in the embodiment of the invention and the equivalent parallel filter structure of models sketch map of GFR-2 black box system.

For transmission realistic simulation signal, spectral characteristic is more complicated than single frequency sinusoidal signal, after the vocoder coding and decoding; Actual low pass frequency band range dwindles to some extent, and the transparent transmission characteristic can decrease, and can be similar to think that the noise jamming that receives strengthens; As long as the receiving terminal demodulator is highly resistant to these noise effects; With the analog signal reliable demodulation that certain distortion is arranged that receives, so, still can the low-frequency range equivalence be one and receive the low pass filter that small noise disturbs." obstruction " that high band partly receives can be bigger, and noise jamming is bigger in other words, more is difficult to utilize.

4, simulating, verifying

Because the OFDM sub-carrier frequencies uniformly-spaced increases progressively, be easy to assess the frequency response characteristic in a certain little frequency range in the continuous frequency range, so the present invention adopts OFDM to realize the conversion of random data and analog signal and carries out conclusion and verify.

The ber curve figure of each data subcarrier that receiving terminal receives is shown in 10.Result by checking can know that the planisphere convergence of data subcarrier 1-20 (frequency range 0-1050Hz) is better, and demodulation bit error rate explains that less than 0.1% this frequency range noise jamming is very little; Along with frequency increases, planisphere is dispersed gradually, and the error rate increases, and when subcarrier 25 (frequency 1250Hz), planisphere is fuzzy fully, can't demodulation.Explanation is for the realistic simulation signal, and the low pass frequency band range dwindles, and permeability reduces.That is to say that simulation result has confirmed the correctness of cascade vocoder equivalence filter model.

This shows; Set up related between the low pass, band pass filter of the embodiment of the invention with cascade vocoder and equivalence; Can be the accurate steadily design of modulator-demodulator of voice band foundation is provided; Thereby solve the problem that encrypted voice transmits in the mobile voice service channel, improve the reliability and the validity of system.

Accordingly, the embodiment of the invention also provides a kind of transmission method of voice data, and its flow chart is shown in figure 11, and said method comprises:

Step 201: equivalent parallel filter receives the encryption of transmitting terminal transmission and the voice data of modulation;

Step 203: equivalent parallel filter sends to receiving terminal with filtered data after the voice data are carried out filtering, and to carry out demodulation successively close so that receiving terminal is to receiving said data, obtains former voice data.

Preferably, said method can also comprise: set up equivalent parallel filter, the said implementation procedure of setting up the method for equivalent parallel filter sees Fig. 1 and preferred embodiment thereof for details, repeats no more at this.

Also see also Figure 12; Be a kind of structural representation of setting up the device of equivalent parallel filter that provides in the embodiment of the invention; This device comprises: set up unit 121, analytic unit 122, first equivalent unit 123 and second equivalent unit 124; Wherein, the said unit 121 of setting up is used to set up the channel model of cascade vocoder; Said analytic unit 122 is used for the said channel of setting up the cascade vocoder of setting up unit 121 is carried out the time and frequency domain analysis analysis, obtains amplitude-frequency response and phase-frequency response accurately; Make an uproar a little less than with phase versus frequency response charac t the being parallel connection of interference low pass filter and strong jamming band pass filter of said first equivalent unit 123, the amplitude-frequency response characteristic that is used for obtaining with the equivalence of said cascade vocoder channel according to said analytic unit 122; Second equivalent unit 124, being used for said speech coding and decode procedure equivalence of setting up the cascade vocoder channel of setting up unit 121 is the parallel model that signal is carried out LPF and bandpass filtering.

Preferably; Said analytic unit can comprise: first analytic unit, C grade are imitated the unit and second analytic unit, wherein, and said first analytic unit; Be used for the said channel of setting up the cascade vocoder of setting up the unit is carried out time domain analysis, obtain the description of wave distortion; Said C grade is imitated the unit, and the description of the wave distortion that is used for obtaining according to said first analytic unit is made an uproar the process errors equivalence of said cascade vocoder channel a little less than for low-frequency range and disturbed and the strongly disturbing parallelly connected noise source of high band; Said second analytic unit is used for the said channel of setting up the cascade vocoder of setting up the unit is carried out frequency-domain analysis, obtains the amplitude-frequency response characteristic and the phase versus frequency response charac t of cascade vocoder channel.

Preferably, the said channel model of setting up the cascade vocoder of setting up the unit specifically can comprise: the channel model of 4 cascade full rate vocoder GFR-4 and 2 cascade full rate vocoder GFR-2, but be not limited to this, also can adaptability set up other models.

Preferably, said device can also comprise: structural unit is used to construct the input signal that the channel model of said cascade vocoder can be discerned.

The function of each unit and the implementation procedure that acts on the correspondence that sees said method for details repeat no more at this in the said device.

Accordingly; The embodiment of the invention also provides a kind of equivalent parallel filter; Its structural representation sees Figure 13 for details, and said equivalent parallel filter can comprise: receiving element 131, equivalent low-pass filter 132, equivalent bandpass filter 133 and transmitting element 134, wherein

Said receiving element 131 is used for the process encryption of receiving end/sending end transmission and the voice data after the modulation; Said equivalent low-pass filter 132 is used for the voice data that receives is carried out filtering, and filtered voice data is sent to transmitting element 134; Said equivalent bandpass filter 133 is used for the voice data that receives is carried out filtering, and filtered voice data is sent to transmitting element 134; Said transmitting element 134 is used for sending to receiving terminal after reception equivalent low-pass filter 132 and the equivalent bandpass filter 133 filtered voice data additions.

The function of each unit and the implementation procedure that acts on the correspondence that sees said method for details repeat no more at this in the said equivalent parallel filter.

Accordingly, the embodiment of the invention also provides a kind of transmission system of voice data, and its structural representation sees Figure 14 for details, and said system can comprise: transmitting terminal 141, equivalent parallel filter 142 and receiving terminal 143, wherein,

Said transmitting terminal 141 be used for the voice data that receives is compressed successively, encrypts, encodes and modulates, and the data after will modulating is input to equivalent parallel filter; Said equivalent parallel filter 142 is used for the said data that receive are carried out filtering respectively in low-frequency range and high band, and filtered data is sent to receiving terminal; Said receiving terminal 143 is used for modulating successively receiving said data, decodes, deciphering and decompress(ion), obtains former voice data.

Wherein, the described equivalent parallel filter of said equivalent parallel filter such as Figure 14.

The function of each unit and the implementation procedure that acts on the correspondence that sees said method for details repeat no more at this in the said system.

Can know by the foregoing description; The transmission system of transmission method, equivalent parallel filter and the voice data of a kind of method of setting up equivalent parallel filter that the embodiment of the invention provides, voice data; No matter be directed against the connected mode that GSM holds the GSM end; Still hold the connected mode of PSTN end to GSM; Can make up by GFR-4 and GFR-2 channel model, and channel model carried out the identification analysis, obtain amplitude-frequency response and phase-frequency response accurately from time domain and two angles of frequency domain; The equivalence of 0-1550Hz low-frequency range part is one and receives the low pass filter that is easy to utilize of " small noise interference ", the equivalence of 1550-4000Hz high band part is become one and receives the band pass filter that is difficult to utilize of " strong random disturbances ".Related with having set up between cascade vocoder and low pass, the band pass filter thus, can be the accurate steadily design of modulator-demodulator of voice band foundation is provided, thereby solve the problem that encrypted voice transmits in the mobile voice service channel, thereby improve the efficiency of transmission of system.

Through the description of above execution mode, those skilled in the art can be well understood to the present invention and can realize by the mode that software adds essential general hardware platform, can certainly pass through hardware, but the former is better execution mode under a lot of situation.Based on such understanding; The part that technical scheme of the present invention contributes to prior art in essence in other words can be come out with the embodied of software product; This computer software product can be stored in the storage medium, like ROM/RAM, magnetic disc, CD etc., comprises that some instructions are with so that a computer equipment (can be a personal computer; Server, the perhaps network equipment etc.) carry out the described method of some part of each embodiment of the present invention or embodiment.

The above only is a preferred implementation of the present invention; Should be pointed out that for those skilled in the art, under the prerequisite that does not break away from the principle of the invention; Can also make some improvement and retouching, these improvement and retouching also should be regarded as protection scope of the present invention.

Claims (13)

1. a method of setting up equivalent parallel filter is characterized in that, comprising:

Set up the channel model of cascade vocoder;

Channel to said cascade vocoder carries out time domain analysis and frequency-domain analysis, obtains amplitude-frequency response and phase-frequency response accurately, specifically comprises: the channel to said cascade vocoder carries out time domain analysis, obtains the description of wave distortion; According to the description of said wave distortion the process errors equivalence of said cascade vocoder channel is made an uproar a little less than for low-frequency range and to disturb and the strongly disturbing parallelly connected noise source of high band; Channel to said cascade vocoder carries out frequency-domain analysis, obtains the amplitude-frequency response characteristic and the phase versus frequency response charac t of cascade vocoder channel;

Make an uproar a little less than according to said amplitude-frequency response characteristic and phase versus frequency response charac t being and disturb the parallel connection of low pass filter and strong jamming band pass filter the equivalence of said cascade vocoder channel;

To the speech coding and decoding processing procedure equivalence of said cascade vocoder channel the parallel model that signal is carried out LPF and bandpass filtering.

2. method according to claim 1 is characterized in that, the channel model of said cascade vocoder comprises: when global system for mobile communications GSM-GSM connects, and 4 cascade full rate vocoder GFR-4 channel models; And GSM-public switch telephone network PSTN is when connecting, the channel model of 2 cascade full rate vocoder GFR-2.

3. method according to claim 2; It is characterized in that; The channel model of said GFR-4 comprises: the portable terminal of transmitting terminal and the vocoder of base station are right, and the vocoder of receiving terminal base station and portable terminal is right, and all links between transmitting terminal and the receiving terminal.

4. method according to claim 2 is characterized in that, the channel model of said GFR-2 comprises: the vocoder of portable terminal and base station is to all links of, Radio Link, core net transmission link, PSTN transmission link.

5. according to each described method of claim 1 to 4, it is characterized in that the mode of the description of said wave distortion is: non-similar estimate to describe with similar estimate description.

6. according to each described method of claim 1 to 4; It is characterized in that; Said method also comprises: construct the input signal that the channel model of said cascade vocoder can be discerned; Utilize this input signal to encourage the channel model of said cascade vocoder and the output signal of recorder end at transmitting terminal.

7. the transmission method of a voice data is characterized in that, comprising:

Set up equivalent parallel filter; The said process of setting up equivalent parallel filter comprises: the channel model of setting up the cascade vocoder; Channel to said cascade vocoder carries out time domain analysis and frequency-domain analysis, obtains amplitude-frequency response and phase-frequency response accurately, specifically comprises: the channel to said cascade vocoder carries out time domain analysis, obtains the description of wave distortion; According to the description of said wave distortion the process errors equivalence of said cascade vocoder channel is made an uproar a little less than for low-frequency range and to disturb and the strongly disturbing parallelly connected noise source of high band; Channel to said cascade vocoder carries out frequency-domain analysis, obtains the amplitude-frequency response characteristic and the phase versus frequency response charac t of cascade vocoder channel; Make an uproar a little less than according to said amplitude-frequency response characteristic and phase versus frequency response charac t being and disturb the parallel connection of low pass filter and strong jamming band pass filter the equivalence of said cascade vocoder channel; To the speech coding and decoding processing procedure equivalence of said cascade vocoder channel the parallel model that signal is carried out LPF and bandpass filtering;

Equivalent parallel filter receives the encryption of transmitting terminal transmission and the voice data of modulation;

Equivalent parallel filter carries out low pass and bandpass filtering in voice data respectively, and filtered data are sent to receiving terminal, and to carry out demodulation successively close so that receiving terminal is to receiving said data, obtains former voice data.

8. method according to claim 7 is characterized in that, said method also comprises:

Set up equivalent parallel filter, said method such as each said method of claim 2 to 7 of setting up equivalent parallel filter.

9. a device of setting up equivalent parallel filter is characterized in that, comprising:

Set up the unit, be used to set up the channel model of cascade vocoder;

Analytic unit; Be used for the said channel of setting up the cascade vocoder of setting up the unit is carried out time and frequency domain analysis; Obtain amplitude-frequency response and phase-frequency response accurately; Specifically comprise: first analytic unit, be used for the said channel of setting up the cascade vocoder of setting up the unit is carried out time domain analysis, obtain the description of wave distortion; C grade is imitated the unit, and the description of the wave distortion that is used for obtaining according to said first analytic unit is made an uproar the process errors equivalence of said cascade vocoder channel a little less than for low-frequency range and disturbed and the strongly disturbing parallelly connected noise source of high band; Second analytic unit is used for the said channel of setting up the cascade vocoder of setting up the unit is carried out frequency-domain analysis, obtains the amplitude-frequency response characteristic and the phase versus frequency response charac t of cascade vocoder channel;

Make an uproar a little less than with phase versus frequency response charac t the being parallel connection of interference low pass filter and strong jamming band pass filter of first equivalent unit, the amplitude-frequency response characteristic that is used for obtaining with the equivalence of said cascade vocoder channel according to said analytic unit;

Second equivalent unit, being used for said speech coding and decoding treatment process equivalence of setting up the cascade vocoder channel of setting up the unit is the parallel model that signal is carried out LPF and bandpass filtering.

10. device according to claim 9 is characterized in that, the said channel model of setting up the cascade vocoder of setting up the unit specifically comprises: the channel model of 4 cascade full rate vocoder GFR-4 and 2 cascade full rate vocoder GFR-2.

11. device according to claim 9 is characterized in that, also comprises:

Structural unit is used to construct the input signal that the channel model of said cascade vocoder can be discerned.

12. an equivalent parallel filter is characterized in that, comprising: receiving element, equivalent low-pass filter, equivalent bandpass filter and transmitting element, wherein

Said receiving element is used for the process encryption of receiving end/sending end transmission and the voice data after the modulation;

Equivalent low-pass filter is used for the voice data that receives is carried out filtering, and filtered voice data is sent to transmitting element;

Equivalent bandpass filter is used for the voice data that receives is carried out filtering, and filtered voice data is sent to transmitting element;

Said transmitting element is used for reception equivalent low-pass filter and the filtered voice data addition of equivalent bandpass filter are redispatched to receiving terminal.

13. the transmission system of a voice data is characterized in that, comprising: transmitting terminal, equivalent parallel filter and receiving terminal, wherein,

Said transmitting terminal be used for the voice data that receives is compressed successively, encrypts, encodes and modulates, and the data after will modulating is input to equivalent parallel filter;

Said equivalent parallel filter is used for carrying out filtering to receiving voice data, and filtered voice data is sent to receiving terminal; Specifically comprise: receiving element, equivalent low-pass filter, equivalent bandpass filter and transmitting element, wherein said receiving element is used for the process encryption of receiving end/sending end transmission and the voice data after the modulation; Equivalent low-pass filter is used for the voice data that receives is carried out filtering, and filtered voice data is sent to transmitting element; Equivalent bandpass filter is used for the voice data that receives is carried out filtering, and filtered voice data is sent to transmitting element; Said transmitting element is used for reception equivalent low-pass filter and the filtered voice data addition of equivalent bandpass filter are redispatched to receiving terminal;

Said receiving terminal is used for modulating successively receiving said data, decodes, deciphering and decompress(ion), obtains former voice data.

Images