CN106161785A - A kind of smart mobile phone real-time communication method based on hidden acoustical signal - Google Patents

Abstract

The invention discloses a kind of smart mobile phone real-time communication method based on hidden acoustical signal, by utilizing orthogonal frequency division multiplexi modelled signal, determine subcarrier modulation modes, analyze original audio signal and determine the position of applicable embedding data bag, embedding data signal adaptively, packet header is detected, channel is estimated, extract eight step such as signal and data correction, utilize the audio signal that the masking effect of human auditory system is easy to get with life camber, guaranteeing not affect in main channel while human ear audio frequency audio experience, achieve real-time non-dominant Speaker Microphone data hiding communication.

Description

A kind of smart mobile phone real-time communication method based on hidden acoustical signal

Technical field

The invention belongs to acoustic communication technical field, be specifically related to a kind of smart mobile phone based on hidden acoustical signal real-time Communication means.

Background technology

Popular along with smart machine, people are widely used these equipment and obtain and propagate information.Meanwhile, electronics is wide Accuse plate, notebook computer etc. and be equipped with screen and the equipment of speaker, have become as conveniently information source.Broadband service is public The Global Internet phenomenon report of department's Sandvine semiannual period is pointed out: occupy at north America region, video and voice data stream Peak period the broadband network flow more than 70%.Under this trend, it is desirable to watch (listening to) video (sound user Frequently) while, it is possible to the smart machine held to user transmits other useful information.Such as, football fan is at the viewing live ratio of NBA Games-time, sportsman and the background introduction of team and racing season performance statistics data can be received on smart machine.Existing technology is led to Often directly display additional information；Or first encode information onto and be then displayed on screen, then for visual image (such as Quick Response Code) Image is caught by the smart machine held by user, and is decoded.The former will affect the viewing experience of user；Right The skew of the shake of camera, angle on the latter, smart machine, the distance between user and screen, barrier, relative movement, all Acquisition of information can be affected, cause the motility of communication system to be substantially reduced.Due to audio signal can by diffraction and reflection around Cross barrier to propagate, and propagate to four sides, therefore, the problems referred to above can be solved by audio signal as the carrier of additional information.

This invention address that the audio signal that the masking effect utilizing human auditory system is easy to get with life camber, study and visit A kind of communication system transmitting hidden information between speaker and mobile microphone of rope.

Human ear is relevant to the perception of volume and the frequency of acoustical signal, and its most sensitive frequency range is 2～4KHz.Frequency Time higher, the sound of the most more louder volume could be discovered by human ear.Meanwhile, the hearing system of people has frequency domain and shelters and cover with time domain Cover effect.The former refers to that, in two signals that frequency is close, what volume was bigger will shelter less one.The latter refers to when two When signal occurs the most simultaneously, the signal given great volume will hinder people to volume compared with the perception of small-signal.Therefore, in order to ensure user Audio experience to original audio signal, needs to select its signal vehicle frequency.

The communication channel of audio signal comprises speaker and mike, realizes the signal of telecommunication respectively to audio signal, audio frequency letter Number to the conversion of the signal of telecommunication.The response frequency scope of most of speakers is 50～20000Hz, but different loudspeaker frequency choosings The character of selecting property decline is different, and can decline to a great extent in high band performance.Mike also has the character of frequency selective fading, many The acoustical signal that the mike of number smart mobile phone is higher than 17KHz to frequency cannot respond.And the sound that human ear can be heard Frequency range is 20～18000Hz, and this scope has exceeded the frequency range that mike can respond.Therefore, audio frequency is utilized to believe Number carrying human ear cannot the additional information of perception, need signal is carried out fine design.In actual application, outside noise is done Disturb, the propagation loss of audio signal, code capacity, the power between the robustness of the not sentience of additional signal and information transmission The factors such as weighing apparatus, are required for being considered.

Summary of the invention

In order to solve above-mentioned technical problem, the invention provides a kind of smart mobile phone based on hidden acoustical signal and lead in real time The method of letter.

The technical solution adopted in the present invention is: a kind of smart mobile phone real-time communication method based on hidden acoustical signal, It is characterized in that, comprise the following steps:

Step 1: utilize orthogonal frequency division multiplexi modelled signal；

Step 2: determine subcarrier modulation modes；

Step 3: analyze original audio signal, determine the position of applicable embedding data bag；

Step 4: embedding data signal adaptively；

Step 5: packet header is detected；

Step 6: channel is estimated；

Step 7: extract signal；

Step 8: data correction.

As preferably, implementing of step 1 includes following sub-step:

Step 1.1: selecting the frequency 64 subcarriers between 8～14.3KHz to be used for carrying data message, frequency exists 7 subcarriers between 14.4～15KHz carry control signal；All subcarrier widths are chosen for 100Hz；

Step 1.2: determine the mode of modulated signal according to the Energy distribution situation of original audio；When original audio is more than The spectral energy density of 8KHz is less than 3*10^-2During J/Hz, the specific form on-off keying (OOK) of ASK is utilized to modulate letter Number, and before embedding data, delete the original audio part more than 8KHz；Otherwise, use EDK to carry out modulated signal, and embedding number According to the front deletion original audio part more than 14.3KHz；

Step 1.3: be some packets by the data flow point of embedding, select in the first half of persistent period frequency from f_minIt is increased to f_max, later half is reduced to f in the time_minLinear FM signal as packet header (f_min=17KHz, f_max= 20KHz), and fill silence period of one section of 50ms；Data signal being distributed in each packet, use frequency is again The pilot signal of 14.5KHz is estimated for time selectivity, Doppler frequency estimation；Afterbody at marking signal chooses one section of spy The signal of measured length, as Cyclic Prefix, is used for resisting intersymbol interference and inter-carrier interference.、

As preferably, the process that implements of step 2 is: the audio signal of corresponding informance bag, its low-frequency range audio signal ESD byBe calculated, the ESD of high band audio signal byIt is calculated；Its In, l is sampling number in each bag；F (KHz) representation signal frequency；F (f) represents the amplitude of the signal that frequency is f；Δf(i, J) frequency range is represented, i.e. f ∈ [f_i,f_j]；

Acquiescence modulator approach is ASK, when meeting E_ph＞ E_high, andTime select EDK be modulator approach (E_highAnd R_hl Representing that two are weighed threshold value, the present invention arranges E_high=3*10^-2J/Hz,)。

As preferably, the process that implements of step 3 is: for the audio fragment that each symbol is corresponding, calculate whole Energy spectral density E of individual frequency band audio signal_st, energy spectral density E of audio signal_sl, energy spectral density E of high band audio signal_sh, Computational methods are identical with information bag；

Definition threshold value E_min, represent the minimum audio frequency energy spectral density that implant data symbol needs；Work as E_st＞ E_minShi Caixiang sound Frequently embedding data in fragment, does not the most carry out data embedding.

As preferably, step 4 implements process and is:

The subcarrier that ASK mode is modulated, its energy intensity computing formula embedding signal is as follows:

$E_{am}=\left\{\begin{array}{cc}64*{\mathrm{\&beta;}}^2{E}_{sl}& E_{sl}<E_{max}\\ 64*{\mathrm{\&beta;}}^2{E}_{max}& E_{sl}\&GreaterEqual;E_{max}\end{array}\right.$

The subcarrier that EDK mode is modulated, its strength calculation formula embedding signal is as follows:

$E_{en}=\left\{\begin{array}{cc}64*{\mathrm{\&beta;}}^2{E}_{sl}{B}_{si}& E_{sl}<E_{max}\\ 64*{\mathrm{\&beta;}}^2{E}_{max}{B}_{si}& E_{sl}\&GreaterEqual;E_{max}\end{array}\right.$

Wherein, β represents embedment strength coefficient, B_siIt is the Bandwidth adjustment value in EDK, E_maxIt is that measurement embedding signal power spectrum is close The threshold value of degree maximum, E_slRepresent the energy spectral density of signal.

As preferably, implementing of step 5 includes following sub-step:

Step 5.1: utilize envelope detection to detect packet header signal；

Step 5.2: obtain subcarrier modulation modes.

As preferably, implementing of step 6 includes following sub-step:

Step 6.1: use the least square estimation method, utilizes in each information bag on the subcarrier that first symbol is corresponding Pilot tone, carry out frequency selectivity estimation；

Definition w (n) is environment noise, and h (n) is channel impulse response, and N is the sample number in each mark, and x (n) is for connecing The signal sequence received；After removing Cyclic Prefix, the signal expression that in first symbol, receiving end receives is:

$y\left(n\right)=x\left(n\right)\&CircleTimes;h\left(n\right)+w\left(n\right),n=0,1,\mathrm{...},N-1;$

By Fourier transformation, being transformed into frequency domain from time domain by y (n), expression formula is:

Y (k)=X (k) * H (k)+W (k) k=0,1 ..., N 1；

The then estimated value of channel impulse response, computing formula is:

$H_e\left(k\right)=\frac{Y_p\left(k\right)}{X_p\left(k\right)}=\frac{X_p\left(k\right)H_p\left(k\right)+W_p\left(k\right)}{X_p\left(k\right)}=H_p\left(k\right)+\frac{W_p\left(k\right)}{X_p\left(k\right)};$

Wherein, X_pK () represents pilot frequency sequence, W_pK () represents the environment noise at pilot tone, H_pK () represents that at pilot tone, channel rings Should,Represent estimation difference；

Step 6.2: use the least square estimation method, the pilot tone utilized in each symbol on the subcarrier of 14.5KHz is believed Number carry out time selectivity estimation and Doppler frequency estimation；

Frequency f because of transmitting terminal pilot signal_sIt is known that by frequency f of pilot signal in the signal of detection receiving terminal_rJust Can calculate the factor of determination of Doppler frequency deviation effect:And then pass through formula:Calculate Frequency deviation to all subcarriers；Here v represents the aerial spread speed of sound, v₀Representing the translational speed of mobile phone, θ represents Angle between speaker and mobile phone moving direction.

As preferably, step 7 implements process and is: a length identical with subcarrier bandwidth " data window ", is responsible for Intercept mid frequency and the identical data of first sub-carrier frequencies；According to corresponding subcarrier-modulated method, signal is carried out Demodulation；" data window " moves forward the length of a subcarrier bandwidth afterwards, until all subcarriers in embed data all by Extract.

As preferably, in step 7, have with the average energy of the audio fragment of corresponding symbol owing to embedding the energy of signal Close, thus be accordingly used in the threshold size extracting each symbol, will be adjusted according to the average energy of audio signal.

As preferably, implementing of step 8 includes following sub-step:

Step 8.1: error correction in symbol；In a symbol, utilize RS code to carry out error detection and correction noise and draw The mistake risen；Before data encoding is RS code, last element of original information is set to other all elements XOR value, recipient's data decoded to RS code carry out same xor operation, for verifying the correctness of data；

Step 8.2: intersymbol erasure correction；Symbol in information bag cell (i) represents, wherein i ∈ [1,30], Cell (i) (j) represents the bit on jth subcarrier, wherein j ∈ [1,64]；Choose last m the symbol in information bag Number as parity check symbol, orderTo each j ∈ [1,60], orderWhen only one symbol generation multi-bit error in s relevant symbol, this The symbol of mistake is just recovered by other s-1 symbol.

Relative to prior art, the invention has the beneficial effects as follows: communication distance and angle are broader, for non-line of sight communication Environment and equipment mobile robustness are higher.

Accompanying drawing explanation

Fig. 1 is the schematic diagram of the embodiment of the present invention.

Detailed description of the invention

Understand and implement the present invention for the ease of those of ordinary skill in the art, below in conjunction with the accompanying drawings and embodiment is to this Bright it is described in further detail, it will be appreciated that enforcement example described herein is merely to illustrate and explains the present invention, not For limiting the present invention.

Present invention is primarily based on spatial-acoustic communication, it is considered to human ear and smart machine acoustic characteristic, it is proposed that a kind of based on The experimental technique of the smart mobile phone real-time Communication for Power of hidden acoustical signal and system.What this method made full use of human auditory system shelters effect The audio signal should being easy to get with life camber, is guaranteeing not affect in main channel while human ear audio frequency audio experience, it is achieved Real-time non-dominant loudspeaker-microphone data hiding communication.The present invention can be as existing hidden screen-photographing unit Supplementing of communication system, improves rate of data signalling and credibility.

The method that the present invention provides can realize flow process with computer software technology.See Fig. 1, the one that the present invention provides Smart mobile phone real-time communication method based on hidden acoustical signal, comprises the following steps:

Step 1, utilizes orthogonal frequency division multiplexi modelled signal, and the process of implementing is:

Step 1.1, selects the frequency 64 subcarriers between 8～14.3KHz to be used for carry data signals, and frequency exists 7 subcarriers between 14.4～15KHz carry control signal, and all subcarrier widths are chosen for 100Hz.

Step 1.2, according to the Energy distribution situation of original audio signal, selects to use EDK or ASK technology to adjust often Signal on individual subcarrier；Being packet by this data stream, each packet contains 31 symbols (comprising data bit) and 1 The individual packet header signal for synchronizing, and embedding cycle prefix before each symbol.

The implementation process that embodiment is concrete is described as follows:

First select the frequency 64 subcarriers between 8～14.3KHz to be used for carrying data message, frequency 14.4～ 7 subcarriers between 15KHz carry control signal.

Secondly, it is more than the spectral energy density of 8KHz less than 3*10 when original audio^-2During J/Hz, utilize the special form of ASK Formula break-make key carrys out modulated signal, and deletes the original audio part more than 8KHz before embedding data；Otherwise, EDK is used Carry out modulated signal, and before embedding data, delete the original audio part more than 14.3KHz；

Finally, the data flow point of embedding is some packets by the present invention.Select frequency in the first half of persistent period From being increased to, the linear FM signal being reduced in the later half time is as packet header, and fills the silence period of one section of 50ms.Lead Frequency is for ofdm signal transmitting terminal and the synchronization of receiving terminal.This example is chosen, the whole pilot signal persistent period is 100ms.Again Data signal is distributed in each packet, use frequency be 14.5KHz pilot signal for time selectivity estimate, how general Le frequency displacement is estimated.Choose the Data Labels persistent period of a length of 100ms in this example.Afterbody at marking signal chooses one section The signal of length-specific is 10ms as the persistent period in Cyclic Prefix, this example, is used for resisting intersymbol interference and inter-carrier interference.

Step 2, determines subcarrier modulation modes, the audio fragment corresponding with each packet is carried out energy and divides Analysis, selects subcarrier modulation modes (ASK or EDK) for this packet；The control of a 14.6KHz is embedded in the packet header of packet Signal processed, represents the modulation system that corresponding subcarrier uses.

The process of implementing is:

Definition l is sampling number in each bag；F (KHz) representation signal frequency；F (f) represents the width of the signal that frequency is f Degree；(i, j) represents frequency range to Δ f, i.e. f ∈ [f_i,f_j]。

The then audio signal of corresponding informance bag, its average energy spectrum density (ESD) can be by being calculated.Low-frequency range audio frequency is believed Number ESD can be byBe calculated, the ESD of high band audio signal byCalculate Arrive.

Acquiescence modulator approach is ASK, when meeting E_ph＞ E_high, andTime select EDK be modulator approach.This enforcement Example selects E_high=3*10^-2J/Hz,And the signal embedding 14.6KHz in each pilot tone shows to select Modulator approach.

Step 3, analyzes original audio signal, determines the position of applicable embedding data bag；To with packet in each symbol Number corresponding audio fragment carries out energy spectrometer when meeting condition set in advance, adaptive to this fragment in subsequent step Answer ground embedding data, the most do not carry out data embedding.The control letter of 14.8KHz is embedded in each such audio fragment Number, as the labelling of its whether embedding data.

The process of implementing is:

For the audio fragment that each symbol is corresponding, calculate energy spectral density E of whole frequency band audio signal_st, audio frequency Energy spectral density E of signal_sl, energy spectral density E of high band audio signal_sh, computational methods and the E in step 2_pt、E_pl、E_phPhase With.Here calculating is directed to the audio fragment corresponding with each mark in information bag.

Definition threshold value E_min, represent the minimum audio frequency energy spectral density that implant data symbol needs.Only work as E_st>E_minShi Caixiang Embedding data in audio fragment, does not the most carry out data embedding.This example is chosen E_min=6*10^-3J/Hz.Plant at 14.8KHz Enter a control signal to indicate that this section audio is with or without embedding signal.

Step 4, adaptively embedding data signal；Original audio signal energy is the highest, and the intensity embedding signal is the biggest, instead The lowest.

The process of implementing is:

The subcarrier that ASK mode is modulated, its energy intensity computing formula embedding signal is as follows:

$E_{am}=\left\{\begin{array}{cc}64*{\mathrm{\&beta;}}^2{E}_{sl}& E_{sl}<E_{max}\\ 64*{\mathrm{\&beta;}}^2{E}_{max}& E_{sl}\&GreaterEqual;E_{max}\end{array}\right.$

The subcarrier that EDK mode is modulated, its strength calculation formula embedding signal is as follows:

$E_{en}=\left\{\begin{array}{cc}64*{\mathrm{\&beta;}}^2{E}_{sl}{B}_{si}& E_{sl}<E_{max}\\ 64*{\mathrm{\&beta;}}^2{E}_{max}{B}_{si}& E_{sl}\&GreaterEqual;E_{max}\end{array}\right.$

Wherein, β represents embedment strength coefficient, B_siIt is the Bandwidth adjustment value in EDK, E_maxIt is that measurement embedding signal power spectrum is close The threshold value of degree maximum, E_slRepresent the energy spectral density of signal；

In the present embodiment, B is chosen_si=20Hz.E_maxIt is the threshold value weighed and embed signal energy spectral density maximum, this example In take E_max=5*10^-2J/Hz.When the energy of original audio signal strengthens, owing to embedding signal, there is enough robusts Property, therefore its intensity keeps constant, otherwise by probably due to noise is excessive and affect user's audition body to original audio signal Test.

Estimating for the ease of channel, the energy of transmitting terminal pilot signal must be known to recipient, therefore leads in this example Frequently the energy size of signal takes fixed value.

After embedding data signal, through inverse Fourier transform, just can be sent to receiving terminal by speaker.

Step 5, packet header is detected, and utilizes packet header medium frequency to carry out the tune of determination flag subcarrier in the control signal of 14.6KHz Adjusting method.

The process of implementing is:

Step 5.1: utilize envelope detection to detect packet header signal.In theory, what envelope maximum was corresponding is the position in packet header Put.But in actual application, due to signal ring time and the existence of rise time, in the signal of receiving terminal, near position, packet header Envelope size closely.This will cause the synchronous error of at most 5 sampled points, causes unpredicted phase offset.? In native system, corresponding 4410 sampled points of each symbol, therefore it is less than the timing error sub-carrier signal of 5 sampled points Amplitude and Energy distribution have little to no effect.

Step 5.2: obtain subcarrier modulation modes.Behind packet header, location, believed by the control that its medium frequency is 14.6KHz Number, receiving terminal can know the modulation system of subcarrier.

Step 6, carries out channel estimation；Use pectination-bulk hybrid pilot, first pass through first symbol pair in packet The pilot tone on subcarrier answered, carries out frequency selectivity estimation (FSE)；Recycle in each symbol on 14.5KHz subcarrier Pilot tone carries out time selectivity estimation (TSE) and Doppler frequency estimation (DFOE).

The process of implementing is:

Step 6.1: use least-squares estimation (LSE) method, utilize the son that in each information bag, first symbol is corresponding Pilot tone on carrier wave, carries out frequency selectivity estimation (FSE).

Definition w (n) is environment noise, and h (n) is channel impulse response, and N is the sample number in each mark, and x (n) is for connecing The signal sequence received；After removing Cyclic Prefix, the signal expression that in first symbol, receiving end receives is:

By Fourier transformation, being transformed into frequency domain from time domain by y (n), expression formula is:

Y (k)=X (k) * H (k)+W (k) k=0,1 ..., N 1；

The then estimated value of channel impulse response, computing formula is:

$H_e\left(k\right)=\frac{Y_p\left(k\right)}{X_p\left(k\right)}=\frac{X_p\left(k\right)H_p\left(k\right)+W_p\left(k\right)}{X_p\left(k\right)}=H_p\left(k\right)+\frac{W_p\left(k\right)}{X_p\left(k\right)};$

Wherein, X_pK () represents pilot frequency sequence, W_pK () represents the environment noise at pilot tone, H_pK () represents that at pilot tone, channel rings Should,Represent estimation difference；

In the present embodiment, owing to only encoding the frequency signal more than 8KHz, outside noise has little to no effect, Therefore estimation difference is the least, and the frequency selectivity decline of symbol approximates therewith.

Step 6.2: use least-squares estimation (LSE) method, utilize leading on the subcarrier of 14.5KHz in each symbol Frequently signal carries out time selectivity estimation (TSE) and Doppler frequency estimation (DFOE).

When receiving terminal is when mobile, due to the existence of Doppler effect, the amplitude of channel response and phase place in a symbol Will change.In order to compensate estimation difference, need to take the movement of receiving terminal into account.Due to transmitting terminal pilot signal Frequency f_sKnown (14.5KHz), by frequency f of pilot signal in the signal of detection receiving terminal_rJust can calculate Doppler frequency deviation The factor of determination of effect:And then pass through formula:All subcarriers can be calculated Frequency deviation.Here v represents the aerial spread speed of sound, v₀Representing the translational speed of mobile phone, θ represents speaker and mobile phone Angle between moving direction.

After eliminating the impact of Doppler frequency deviation, all of data signal can be accurately positioned.

Step 7, extracts signal, and the process of implementing is:

After DFOE, each subcarrier carrying data be refined location, hence with channel estimation the most reducible go out Primary signal.

The present invention defines a length " data window " identical with subcarrier bandwidth, responsible intercepting mid frequency and first The data that sub-carrier frequencies is identical.According to corresponding subcarrier-modulated method, signal is demodulated." data window " forward afterwards The length of a mobile subcarrier bandwidth, until the data embedded in all subcarriers are all extracted.Owing to embedding signal Energy relevant with the average energy of the audio fragment of corresponding symbol, thus be accordingly used in the threshold size extracting each symbol, root It is adjusted according to the average energy of audio signal.

Step 8, data correction；Error in data distribution situation in test packet, major part the most repeatedly The mistake of symbol is less than 3 bits, it may be possible to owing to the noise in external environment causes；Small part symbol error is more than 10 ratios Special, it may be possible to owing to multipath effect causes, need in symbol, use more odd encoder, it is ensured that the reliability of data.

Design orthogonal error-correcting system, be divided into error correction, intersymbol erasure correction two parts in symbol.The former uses The error in data caused due to noise in symbol is detected and corrects by RS code；For the latter, in symbol, error correction completes After, it is known which symbol is insecure, therefore last m the symbol in availability data bag is as even-odd check Mark carrys out error correction.

The process of implementing is:

Step 8.1: error correction in symbol.In a symbol, emphasis processes the mistake that noise causes.In this example, RS code is selected to carry out error detection and correction.In the finite field comprising 15 elements, it is possible to correct at mostIndividual mistake unit Element, and detect the combination in any of at most n-k Error Elements.In order to improve error detection capability, it is being RS by data encoding Before Ma, last element of original information is set to the XOR value of other all elements.After RS code is decoded by recipient Data carry out same xor operation, for verifying the correctness of data.

Step 8.2: intersymbol erasure correction.The purpose of this step is to correct a large amount of mistakes in a few symbols, this A little mistakes cannot be corrected by RS code.Symbol in information bag cell (i) represents, wherein i ∈ [1,30], Cell (i) (j) represents the bit on jth subcarrier, wherein j ∈ [1,64].Owing in symbol, error correction is complete Cheng Hou, the present invention which symbol known is insecure, and other reliable symbols in information bag therefore can be utilized Reacquire the information in these symbols.In this example, last m the symbol in information bag is chosen as very Even parity check symbol.OrderTo each j ∈ [1,64], orderWhen only one symbol generation multi-bit error in s relevant symbol, this The symbol of mistake just can be recovered by other s-1 symbol.

It should be appreciated that the part that this specification does not elaborates belongs to prior art.

It should be appreciated that the above-mentioned description for preferred embodiment is more detailed, can not therefore be considered this The restriction of invention patent protection scope, those of ordinary skill in the art, under the enlightenment of the present invention, is weighing without departing from the present invention Profit requires under the ambit protected, it is also possible to make replacement or deformation, within each falling within protection scope of the present invention, this The bright scope that is claimed should be as the criterion with claims.

Claims (10)

1. a smart mobile phone real-time communication method based on hidden acoustical signal, it is characterised in that comprise the following steps:

Step 1: utilize orthogonal frequency division multiplexi modelled signal；

Step 2: determine subcarrier modulation modes；

Step 3: analyze original audio signal, determine the position of applicable embedding data bag；

Step 4: embedding data signal adaptively；

Step 5: packet header is detected；

Step 6: channel is estimated；

Step 7: extract signal；

Step 8: data correction.

Smart mobile phone real-time communication method based on hidden acoustical signal the most according to claim 1, it is characterised in that step Rapid 1 implement includes following sub-step:

Step 1.1: select the frequency 64 subcarriers between 8～14.3KHz to be used for carrying data message, frequency 14.4～ 7 subcarriers between 15KHz carry control signal；All subcarrier widths are chosen for 100Hz；

Step 1.2: determine the mode of modulated signal according to the Energy distribution situation of original audio；When original audio is more than 8KHz's Spectral energy density is less than 3*10^-2During J/Hz, utilize the specific form on-off keying of ASK to carry out modulated signal, and embedding The original audio part more than 8KHz is deleted before data；Otherwise, use EDK to carry out modulated signal, and delete former before embedding data The beginning audio frequency part more than 14.3KHz；

Step 1.3: be some packets by the data flow point of embedding, selects in the first half of persistent period frequency from f_minIncrease Greatly f_max, later half is reduced to f in the time_minLinear FM signal as packet header, and fill the silence period of one section of 50ms； Again data signal is distributed in each packet, use frequency be 14.5KHz pilot signal for time selectivity estimate, many General Le frequency displacement is estimated；Afterbody at marking signal chooses the signal of one section of length-specific as Cyclic Prefix, is used for resisting intersymbol Interference and inter-carrier interference.

Smart mobile phone real-time communication method based on hidden acoustical signal the most according to claim 1, it is characterised in that step The process that implements of rapid 2 is: the audio signal of corresponding informance bag, the ESD of its low-frequency range audio signal by Be calculated, the ESD of high band audio signal byIt is calculated；Wherein, sampled point in l is each bag Number；F (KHz) representation signal frequency；F (f) represents the amplitude of the signal that frequency is f；(i, j) represents frequency range to Δ f, i.e. f ∈ [f_i,f_j]；

Acquiescence modulator approach is ASK, when meeting E_ph>E_high, andTime select EDK be modulator approach；Wherein, E_highAnd R_hl Represent that two are weighed threshold value.

Smart mobile phone real-time communication method based on hidden acoustical signal the most according to claim 1, it is characterised in that step The process that implements of rapid 3 is: for the audio fragment that each symbol is corresponding, calculate the power spectrum of whole frequency band audio signal Density E_st, energy spectral density E of audio signal_sl, energy spectral density E of high band audio signal_sh, computational methods are identical with information bag；

Definition threshold value E_min, represent the minimum audio frequency energy spectral density that implant data symbol needs；Work as E_st>E_minShi Caixiang audio fragment Middle embedding data, does not the most carry out data embedding.

Smart mobile phone real-time communication method based on hidden acoustical signal the most according to claim 1, it is characterised in that step Rapid 4 implement process is:

The subcarrier that ASK mode is modulated, its energy intensity computing formula embedding signal is as follows:

$E_{am}=\left\{\begin{array}{cc}64*{\mathrm{\&beta;}}^2{E}_{sl}& E_{sl}<E_{max}\\ 64*{\mathrm{\&beta;}}^2{E}_{max}& E_{sl}\&GreaterEqual;E_{max}\end{array}\right.$

The subcarrier that EDK mode is modulated, its strength calculation formula embedding signal is as follows:

$E_{en}=\left\{\begin{array}{cc}64*{\mathrm{\&beta;}}^2{E}_{sl}{B}_{si}& E_{sl}<E_{max}\\ 64*{\mathrm{\&beta;}}^2{E}_{max}{B}_{si}& E_{sl}\&GreaterEqual;E_{max}\end{array}\right.$

Wherein, β represents embedment strength coefficient, B_siIt is the Bandwidth adjustment value in EDK, E_maxIt is to weigh to embed signal energy spectral density The threshold value of big value, E_slRepresent the energy spectral density of signal.

Smart mobile phone real-time communication method based on hidden acoustical signal the most according to claim 1, it is characterised in that step Rapid 5 implement includes following sub-step:

Step 5.1: utilize envelope detection to detect packet header signal；

Step 5.2: obtain subcarrier modulation modes.

Smart mobile phone real-time communication method based on hidden acoustical signal the most according to claim 1, it is characterised in that step Rapid 6 implement includes following sub-step:

Step 6.1: use the least square estimation method, utilizes in each information bag leading on the subcarrier that first symbol is corresponding Frequently, frequency selectivity estimation is carried out；

Definition w (n) is environment noise, and h (n) is channel impulse response, and N is the sample number in each mark, and x (n) is for receiving Signal sequence；After removing Cyclic Prefix, the signal expression that in first symbol, receiving end receives is:

$y\left(n\right)=x\left(n\right)\&CircleTimes;h\left(n\right)+w\left(n\right),n=0,1,...,N-1;$

By Fourier transformation, being transformed into frequency domain from time domain by y (n), expression formula is:

Y (k)=X (k) * H (k)+W (k) k=0,1 ..., N 1；

The then estimated value of channel impulse response, computing formula is:

$H_e\left(k\right)=\frac{Y_p\left(k\right)}{X_p\left(k\right)}=\frac{X_p\left(k\right)H_p\left(k\right)+W_p\left(k\right)}{X_p\left(k\right)}=H_p\left(k\right)+\frac{W_p\left(k\right)}{X_p\left(k\right)};$

Wherein, X_pK () represents pilot frequency sequence, W_pK () represents the environment noise at pilot tone, H_pK () represents channel response at pilot tone,Represent estimation difference；

Step 6.2: using the least square estimation method, the pilot signal utilized in each symbol on the subcarrier of 14.5KHz is entered Row time selectivity is estimated and Doppler frequency estimation；

Frequency f because of transmitting terminal pilot signal_sIt is known that by frequency f of pilot signal in the signal of detection receiving terminal_rJust can count The factor of determination of calculation Doppler frequency deviation effect:And then pass through formula:It is calculated institute There is the frequency deviation of subcarrier；Here v represents the aerial spread speed of sound, v₀Representing the translational speed of mobile phone, θ represents and raises one's voice Angle between device and mobile phone moving direction.

Smart mobile phone real-time communication method based on hidden acoustical signal the most according to claim 1, it is characterised in that step Rapid 7 implement process is: a length identical with subcarrier bandwidth " data window ", responsible intercepting mid frequency and first The data that sub-carrier frequencies is identical；According to corresponding subcarrier-modulated method, signal is demodulated；" data window " forward afterwards The length of a mobile subcarrier bandwidth, until the data embedded in all subcarriers are all extracted.

Smart mobile phone real-time communication method based on hidden acoustical signal the most according to claim 8, it is characterised in that: step In rapid 8, relevant with the average energy of the audio fragment of corresponding symbol owing to embedding the energy of signal, thus be accordingly used in each symbol of extraction Number threshold size, to be adjusted according to the average energy of audio signal.

Smart mobile phone real-time communication method based on hidden acoustical signal the most according to claim 1, it is characterised in that Implementing of step 8 includes following sub-step:

Step 8.1: error correction in symbol；In a symbol, RS code is utilized to carry out what error detection and correction noise caused Mistake；Before data encoding is RS code, last element of original information is set to the XOR of other all elements Value, recipient's data decoded to RS code carry out same xor operation, for verifying the correctness of data；

Step 8.2: intersymbol erasure correction；Symbol in information bag cell (i) represents, wherein i ∈ [1,30], Cell (i) (j) represents the bit on jth subcarrier, wherein j ∈ [1,64]；Choose the last m in information bag Symbol is as parity check symbol, orderTo each j ∈ [1,60], orderWhen only one symbol generation multi-bit error in s relevant symbol, this The symbol of mistake is just recovered by other s-1 symbol.