CN104065777A - Mobile communication device - Google Patents

Abstract

The invention relates to a mobile communication device which comprises a machine shell, a device body, a telephone receiver and a microphone. The device body is arranged in the machine shell in a built-in mode. The telephone receiver comprises a vibrating membrane plate and a piezoelectric ceramic piece fixed to the vibrating membrane plate. The vibrating membrane plate extends outwards to form the machine shell. According to the mobile communication device, the vibrating membrane plate is extended to form the machine shell of the mobile communication device, a user can answer the phone by attaching the ear to any position of the machine shell, and use is convenient; no telephone receiver hole needs to be formed in the machine shell, and thus water and dust can be prevented effectively.

Description

Mobile communication equipment

Technical field

The present invention relates to portable electric appts, relate in particular to a kind of mobile communication equipment.

Background technology

Mobile communication equipment comprises receiver and microphone, i.e. receiver and microphone.Microphone is used for gathering sound, and sound wave is converted into the signal of telecommunication.Receiver, for the signal of telecommunication is converted into sound wave, is converted into sound wave by the signal of telecommunication.In prior art, can only singlehanded adept machine during driving, may cause mobile phone to turn upside down in a hurry or reverse side is answered, now the singlehanded direction that is difficult to change mobile phone, microphone just can not effectively be collected user's sound, and user is also difficult to not hear the sound that receiver sends.

Summary of the invention

The technical problem to be solved in the present invention is, the defect that exists receiver to receive calls at ad-hoc location for mobile communication equipment in prior art provides a kind of mobile communication equipment, can realize that receiver is comprehensive to receive calls.

The technical solution adopted for the present invention to solve the technical problems is: construct a kind of mobile communication equipment, comprise casing, equipment body, receiver and microphone, described equipment body is built in described casing, described receiver involving vibrations lamina membranacea and be fixed on the piezoelectric ceramic piece in described diaphragm panel, described diaphragm panel stretches out and forms described casing.

The invention provides a kind of mobile communication equipment, particularly, described microphone comprises acoustical-electrical transducer and impedance transformer, and described acoustical-electrical transducer comprises microphone array, and described microphone array comprises described diaphragm panel.

The invention provides a kind of mobile communication equipment, further, described acoustical-electrical transducer also comprises sound source separation module, and described sound source separation module is for separating of the different sound source of at least two groups.

The invention provides a kind of mobile communication equipment, particularly, described sound source separation module comprises: A/D signal conversion unit, and for converting analog signal to digital signal, described analog signal is from least two described microphone arrays; Dividing frequency band unit, for described digital signal is carried out to dividing frequency band, to be converted to frequency domain input; Error minimal solution computing unit, for each frequency band, described error minimal solution computing unit comprises at least two sound source vectors, according to the default steering vector of described sound source vector, calculate estimated signal, the disaggregation of described error minimal solution computing unit output has minimal error between described estimated signal and the input of described frequency domain; Optimal models computing unit, each the frequency band selection Frequency Domain Solution that is used to described solution to concentrate, described Frequency Domain Solution is minimum l _pthe weighted sum of norm and described error; And signal synthesis unit, for described Frequency Domain Solution is transformed into time domain.

The invention provides a kind of mobile communication equipment, further, described sound source separation module also comprises auditory localization parts, and described guiding vector is by described auditory localization component retrieval.

The invention provides a kind of mobile communication equipment, further, described sound source separation module also comprises direction power calculation parts and direction search parts.

The invention provides a kind of mobile communication equipment, further, described error minimal solution computing unit calculates the minimal error solution of sound source vector described in each, and the null value sound source quantity of described sound source vector equates, and zero valued elements quantity equates.

The invention provides a kind of mobile communication equipment, further, described optimal models computing unit is concentrated and is selected to separate from the described error minimal solution of output, the sliding average that described solution is described error and l _pthe weighted sum of the described sliding average of norm.

The invention provides a kind of mobile communication equipment, further, described sound source separation module also comprises CPU and memory cell.

The invention provides a kind of mobile communication equipment, preferably, described piezoelectric ceramic piece is bonded in the inner side of described casing.

The invention provides a kind of mobile communication equipment, further, described receiver also comprises speech processing module.

The present invention can reach following beneficial effect: by diaphragm panel is extended to the casing that mobile communication equipment is made in expansion, the optional position that user presses close to this casing by ear all can receive calls, easy to use; Simultaneously on casing without offering receiver hole, effective water proof and dust proof.

Accompanying drawing explanation

Below in conjunction with drawings and Examples, the invention will be further described, in accompanying drawing:

Fig. 1 is the structural representation of mobile communication equipment provided by the invention;

Fig. 2 is the hardware schematic diagram of the sound source separation module of mobile communication equipment provided by the invention;

Fig. 3 is the principle schematic of the sound source separation module of mobile communication equipment provided by the invention;

Drawing reference numeral explanation:

1, casing 2, equipment body

3, receiver 4, microphone

31, piezoelectric ceramic piece 41, impedance transformer

42, microphone array 43, A/D signal conversion unit

44, CPU 45, memory cell

46, memory cell 47, auditory localization parts

48, direction power calculation parts 49, dividing frequency band unit

50, direction search parts 51, error minimal solution computing unit

52, l _pnorm calculation unit 53, optimal models computing unit

54, signal synthesis unit 55, sound source separation module

Embodiment

For technical characterictic of the present invention, object and effect being had more clearly, understand, now contrast accompanying drawing and describe the specific embodiment of the present invention in detail.

As shown in Figure 1, the invention provides a kind of mobile communication equipment, comprise casing 1, equipment body 2, receiver 3 and microphone 4, equipment body 2 is built in casing 1.Receiver 3 involving vibrations lamina membranaceas and be fixed on the piezoelectric ceramic piece 31 in diaphragm panel, diaphragm panel stretches out and forms casing 1.

Herein, the material of piezoelectric ceramic piece is a kind ofly can, by mechanical energy and the electric energy ceramic material of conversion mutually, belong to Inorganic Non-metallic Materials.Piezoelectric ceramic utilizes its material under mechanical stress effect, causes the center relative displacement of inner positive and negative charge and polarizes, and the bound charge that causes material two end surfaces to occur that symbol is contrary is piezoelectric effect and making, and has responsive characteristic.Piezoelectric ceramic has extremely sensitive characteristic, can convert extremely faint mechanical oscillation to the signal of telecommunication.

On the basis of such scheme, further, as shown in Figure 1, microphone 4 comprises acoustical-electrical transducer and impedance transformer 41, acoustical-electrical transducer involving vibrations lamina membranacea.As preferably, acoustical-electrical transducer is the casing that diaphragm panel stretches out and forms.That is: microphone 4 shares diaphragm panel with receiver 3.When user talks facing to arbitrary position of casing, all can be by accurately gathering sound, easy to use.

As shown in Figure 2, acoustical-electrical transducer comprises microphone array 42, and the quantity of microphone array 42 is two groups or more.Diaphragm panel is parts of microphone array 42, and as preferably, microphone array 42 is electret microphone.

In order to realize two groups and different sound sources separation more than two, make different sound sources all can be able to separation, thereby the user voice of microphone 4 ends can clearly be transmitted, as shown in Figure 2, acoustical-electrical transducer also comprises sound source separation module 55.When sound source quantity surpasses number of microphone, for example, microphone is 1 group, and sound source quantity is 3 groups; Or number of microphone is 2 groups, sound source quantity is 4 groups, can isolate the sound of each sound source by sound source separation module 55 is set, and is specially adapted under hands-free situation the other side and is easy to hear that sound talks.Meanwhile, if microphone 4 ends also exist background noise, echo or reverberation, this sound source separation module 55 also can be isolated the sound of each sound source.

On the basis of technique scheme, as shown in Figure 2, particularly, sound source separation module 55 comprises: A/D signal conversion unit 43, and for converting analog signal to digital signal, analog signal is from least two microphone arrays 42; Dividing frequency band unit 49, for digital signal is carried out to dividing frequency band, to be converted to frequency domain input; Error minimal solution computing unit 51, for each frequency band, error minimal solution computing unit 51 comprises at least two sound source vectors, the steering vector default according to sound source vector calculates estimated signal, and the disaggregation of error minimal solution computing unit 51 outputs has minimal error between estimated signal and frequency domain input; Optimal models computing unit 53, is used to and separates each frequency band selection Frequency Domain Solution of concentrating, and Frequency Domain Solution is minimum l _pthe weighted sum of norm and error; Sound source separation module 55 also comprises signal synthesis unit 54, for Frequency Domain Solution is transformed into time domain.

As Fig. 2 and as shown in Figure 3, sound source separation module 55 also comprises auditory localization parts 47, and guiding vector obtains by auditory localization parts 47.Further, sound source separation module 55 also comprises direction power calculation parts 48 and direction search parts 50.

Particularly, as shown in Figure 3, error minimal solution computing unit 51 calculates the minimal error solution of each sound source vector, and the null value sound source quantity of sound source vector equates, and zero valued elements quantity equates.Optimal models computing unit 53 is concentrated and is selected to separate from the error minimal solution of output, separates sliding average and l into error _pthe weighted sum of the sliding average of norm.

On the basis of technique scheme, for convenient, connect, as shown in Figure 1, as preferably, piezoelectric ceramic piece 31 is bonded in the inner side of casing 1.

As preferably, receiver 3 also comprises speech processing module.

Below in conjunction with 2 and Fig. 3, elaborate the principle of the sound source separation module of mobile communication equipment provided by the invention:

As shown in Figure 2, be the hardware configuration of sound source separation module 55, sound source separation module 55 comprises CPU 44 and storage device 45, storage device 46.The processing procedure of sound source separation is to carry out in CPU 44, and storage device 45 is preferably RAM memory, and the data of using in sound source processing procedure and program are kept in the storage device 46 of ROM formation.Microphone array 42 comprises at least microphone element of two or more, each microphone element measure analog sound pressure level, and the quantity of supposing microphone is M.A/D signal conversion unit 43 converts analog signal to digital signal, i.e. sampling, and can synchronously to the signal of a plurality of passages, sample.The simulation sound pressure level of microphone is sent to A/D signal conversion unit 43.Pre-set and need separated sound quantity, and be kept in storage device 46.Needing separated sound quantitaes is N, and the processing ability value of CPU is set according to the size of N.

In Fig. 3,11 norms of being used as cost function by 11 Norm minimum methods in the time of except separated sound, the power that comprises noise component(s) in the sound of separating also takes in as cost value.The solution of the optimal models computing unit 53 noise signal power of output and the weighted sum minimum of 11 norms.

A/D signal conversion unit 43 converts simulation sound pressure level to digital signal for each passage, and the digital signal after conversion is carried out according to the sampling time sequence setting in advance, and the numerical data converting is X (t, j), and wherein t is the time, and j is microphone numbering.When A/D signal conversion unit 43 is when t=0 starts A/D conversion constantly, often once sample, t adds 1.For example: the 50th sampled data of the 1st microphone is expressed as to X (50,1).

Dividing frequency band unit 49 is t=τ * frame to be moved to the data that (frame_shift) move (frame_shift)+frame length (frame_size) to t=τ * frame carry out Fourier transform or wavelet analysis, be transformed into dividing frequency band signal, convert each microphone of j=1 to j=M to dividing frequency band signal.Dividing frequency band signal after changing with expression formula (1) description is below as the sound source vector with corresponding microphone.

X(f,τ) (1)

Wherein: f represents the small tenon of dividing frequency band number.

This sound of voice and music seldom has larger range value, and they are sparse signals of a lot of null values, and therefore, voice signal can the high laplacian distribution of enough probability value of zero.When voice signal is approximately to laplacian distribution, log-likelihood can be regarded as between positive and negative the symbol of 11 norms is reversed.Can have echo by mixing, reverberation and ambient noise signal be approximately Gaussian Profile.Therefore, the log-likelihood of the noise signal comprising in input signal can be regarded as to the reversion of the square error symbol between input signal and voice signal.The angle of estimating from MAP is seen the most probable solution that will find, it is maximum likelihood solution, because the log-likelihood sum of the log-likelihood of noise signal and voice signal is got to maximum solution as maximum likelihood solution, therefore can be by the signal of the weighted sum minimum of input signal and 11 norm mean errors as maximum likelihood solution.But, owing to being difficult to find such exact solution, so be necessary by some approximate solutions.For example: in 11 Norm minimum methods, the signal of input does not have error, finding out the signal of the weighted sum minimum of 11 norms separates the most.

In input signal, exist under the hypothesis of error, the weighted sum minimum value of the mean error of input signal and 11 norms is similar to.As mentioned above, this sound of voice and music is seldom to have to be significantly worth sparse signal, therefore, regards them the signal often with approximate zero amplitude/null value as.Therefore,, for each moment and frequency, suppose to only have the source of students fewer than number of microphone to there is the range value of non-zero.11 norms obtain number of microphone increase and diminish along with having null value, along with having null value, obtain number of microphone minimizing and become large.Therefore, it can be regarded as to degree of rarefication tolerance.When thering is null value and obtain the quantity of sound source and equal number of microphone, 11 norms are approximately to fixed value.If apply this when sound source quantity is N (have null value and obtain N dimension complex vector), be similar to, can provide the solution with respect to input signal with minimal error.

Error minimal solution computing unit calculates according to expression formula (2).

${\hat{S}}_L(f,\mathrm{\τ})=\underset{s(f,\mathrm{\τ})\∈L-\mathrm{dimensional\; sp\; arse\; set}}{\arg \min }{|X(f,\mathrm{\τ})-A\left(f\right)S(f,\mathrm{\τ})|}^2---\left(2\right)$

For each of L dimension sparse set (L-dimensional sparse set) calculates error minimal solution.L dimension sparse set is a N dimension complex vector with L zero valued elements.The solution with minimal error calculating is the maximum likelihood solution of each sound-source signal among L dimension sparse set.The solution with minimal error is a N dimension complex vector.Corresponding element is the estimated value of the source signal of corresponding sound source.A (f) is M * N complex matrix, has the steering vector from corresponding sound source position to microphone in its row.For example, the first row of A (f) is the guiding vector from first sound source to microphone array 42.A (f) is by direction search component computes the output of Fig. 3.Error minimal solution computing unit 51 in Fig. 3 is L from each L error of calculation minimal solution of 1 to M.When L=M, calculate a plurality of error minimal solutions, the error minimal solution output that whole a plurality of solutions are all L=M in this case.In this example, for number of elements, equal to have each in the N dimension complex vector of sound source quantity of null value, found error minimal solution.Owing to not being tied to the quantity of sound source, for number of elements, equal to have each in the N n dimensional vector n of quantity of element of null value, can find a solution.Even if be not equal to the quantity of the element with zero degree, if equal sound source quantity, owing to 11 norms can being approximately to fixed value, there is the quantity that null value obtains sound source so, be also enough to find error minimal solution.

Also can apply expression formula (3), and need not above-mentioned expression formula (2).

${\hat{S}}_{L,j}(f,\mathrm{\τ})=\underset{s(f,\mathrm{\τ})\∈{\mathrm{\Ω}}_{L,j}}{\arg \min }{|X(f,\mathrm{\τ})-A\left(f\right)S(f,\mathrm{\τ})|}^2$

error _L，j(f，τ)＝||X(f，τ)-A(f)S(f，τ|| ²

$j_{\min }=\underset{j}{\arg \min }\underset{m=-k}{\overset{k}{\mathrm{\Σ}}}\mathrm{\γ}\left(m\right){\mathrm{error}}_{L,j}(f,\mathrm{\τ}+m)$

${\hat{S}}_L(f,\mathrm{\τ})={\hat{S}}_{L,j_{\min }}(f,\mathrm{\τ})---\left(3\right)$

Ω L, j is that among L dimension sparse set, the value of identical element is a zero N Wei Fushiliangji.Phonetic speech power has positive correlation on time orientation.Therefore, at given τ, there is the sound source of large value, even also can there is large value in τ ± k.This means and can regard less sliding average in error term τ direction as more approach true solution solution.In other words, for each model Ω L, j, by using the sliding average of error term as new error term, can find the solution that more approaches true solution.γ (m) is sliding average, by this structure, easily selects the solution relevant with time orientation.While using moving average to find error minimal solution, each the N dimension complex vector equating for number of elements except null value sound source quantity, must calculate error minimal solution.Even if this is to equate because of sound source quantity, if number of elements is different, also owing to thering is positive correlation on time orientation, can not be similar to.

L in Fig. 3 _pnorm calculation unit 52, according to the error minimal solution calculating by each L dimension sparse set, utilizes expression formula below to calculate l _pnorm:

$l_{p,L}(f,\mathrm{\τ})={\left(\underset{i=1}{\overset{N}{\mathrm{\Σ}}}{\left|{\hat{S}}_{L,i}(f,\mathrm{\τ})\right|}^p\right)}^{\frac{1}{p}}---\left(4\right)$

${\hat{S}}_{L,i}(f,\mathrm{\τ})---\left(5\right)$

${\hat{S}}_L(f,\mathrm{\τ})---\left(6\right)$

Expression formula (5) is i element of expression formula (6).

Variable p is the parameter between 0 to 1 setting in advance, l _pnorm is the tolerance of the sparse degree of expression formula (6), and in expression formula (6), has while approaching zero compared with multielement less.Because voice are sparse, therefore when the value of expression formula (4) hour, can think that expression formula (6) more approaches true solution.Can be by expression formula (4) as choice criteria while in brief, selecting true solution.

The l of expression formula (4) _pthe calculated value of norm can be replaced by sliding average, just as the calculating of error minimal solution:

$\mathrm{avg}-l_{p,L}(f,\mathrm{\τ})=\underset{m=-k}{\overset{k}{\mathrm{\Σ}}}\mathrm{\γ}\left(m\right){\left(\underset{i=1}{\overset{N}{\mathrm{\Σ}}}{\left|{\hat{S}}_{L,j\min i}(f,\mathrm{\τ}+m)\right|}^p\right)}^{\frac{1}{p}}---\left(7\right)$

Because phonetic speech power has positive correlation on time orientation, therefore by replacing it by moving average, can find the solution that approaches true solution.Phonetic speech power only has slightly and changes on time orientation.Therefore, a certain frame can be had significantly to the sound source of being worth regards as in the frame adjacent with this frame and also has significantly value.Each that optimal models computing unit 53 in Fig. 3 is corresponding L dimension sparse set is found out the optimal solution of found error minimal solution;

$L_{\min }=\underset{L}{\arg \min }\mathrm{\α}{\left|\right|X(f,\mathrm{\τ})-A\left(f\right)S(f,\mathrm{\τ})\left|\right|}^2+l_{p,L}(f,\mathrm{\τ})---\left(8\right)$

$\hat{S}(f,\mathrm{\τ})={\hat{S}}_{L_{\min }}(f,\mathrm{\τ})---\left(9\right)$

Expression formula (8) and expression formula (9) output are separated, and make error term and l _pthe weighted average of norm item is minimum.This solution is posterior probability maximal solution.In order to find optimal solution, the same with 11 Norm minimum solutions with error minimal solution, expression formula (8) and expression formula (9) can replace with sliding average:

$L_{\min }=\underset{L}{\arg \min }\mathrm{\α}{\mathrm{error}}_L(f,\mathrm{\τ})+\mathrm{avg}-l_{p,L}(f,\mathrm{\τ})$

$\hat{S}(f,\mathrm{\τ})={\hat{S}}_{L_{\min }}(f,\mathrm{\τ})---\left(10\right)$

According to conventional method, in the processing procedure corresponding to optimal models computing unit 53, do not select from L=2 ..., the solution of M, and L=1 is optimal solution.There is the problem that produces noise in this method.In the solution of L=1, for each f and τ, except a sound source, all values is zero.Sometimes, except a sound source, may exist all values all to approach zero solution.While meeting this condition, the solution of L=1 becomes optimal solution, but is not to satisfy condition.If always suppose L=1, when the sound source of two or more has maximum, just can not find Xie Binghui and produce music noise so.In order to find optimal solution the error minimal solution from finding for each L dimension sparse set, it is optimum for L from 1 to M which sparse set that this optimal models alternative pack is determined, even and the zero solution that greatly also can find of the value of two or more sound sources ratio, thereby the appearance of inhibition music noise.

Signal synthesis unit 54 in Fig. 3 carries out the calculating of optimal solution for each frequency band

$\hat{S}(f,\mathrm{\τ})---\left(11\right)$

By inverse Fourier transform or inverse wavelet transform, turn back to time-domain signal expression formula (12).

$\hat{S}(f,\mathrm{\τ})---\left(12\right)$

By doing like this, the time-domain signal that can obtain each sound source is estimated.Auditory localization parts in Fig. 3 calculate the direction of sound source according to expression formula (13).

$\mathrm{dir}(f,\mathrm{\τ})=\underset{\mathrm{\θ}\∈\mathrm{\Ω}}{\arg \max }{\left|{a_{\mathrm{\θ}}}^{*}(f,\mathrm{\τ})X(f,\mathrm{\τ})\right|}^2---\left(13\right)$

Ω is that the search norm of sound source is enclosed, and sets in advance in ROM3.

a _θ(f，τ) (14)

Expression formula (14) is the steering vector from Sounnd source direction θ to microphone array 42, and its size is to normalize to 1.When source signal is s (f, τ), in microphone array 42, observe the sound from Sounnd source direction θ, by expression formula (15), represent:

X _θ(f，τ)＝s(f，τ)a _θ(f，τ)

(15)

The Ω that comprises institute's sound source in expression formula (13) is kept in storage device 46 in advance, the direction power calculation in Fig. 3 for parts expression formula (16) calculate the sound power of a source in each direction.

$P\left(\mathrm{\θ}\right)=\underset{f}{\mathrm{\Σ}}\underset{\mathrm{\τ}=0}{\overset{K}{\mathrm{\Σ}}}\mathrm{\δ}(\mathrm{\θ}=\mathrm{dir}(f,\mathrm{\τ}))\log {\left|{a_{\mathrm{\θ}}}^{*}(f,\mathrm{\τ})X(f,\mathrm{\τ})\right|}^2---\left(16\right)$

δ is such a function, and only having when the equation of variable is set up is just 1, is zero while being false.Direction search parts search peak P (θ) in Fig. 3 calculate Sounnd source direction, export M * N steering vector matrix A (f), have the steering vector of Sounnd source direction in this matrix column.Peak value searching, by descending P (θ), can calculate N high-order source of students direction, or when P (θ) exceeds fore-and-aft direction (when it becomes maximum), calculate N high-order Sounnd source direction.Error minimal solution computing unit is used as A (f) by this information in expression formula (2), finds error minimal solution.Direction search parts search A (f) carry out automatic estimation voice direction, even if audio direction is unknown, thereby can make sound source separated.

Fig. 3 shows the handling process of this embodiment, and the voice of input are with sound pressure level, to receive in each microphone element.Convert the sound pressure level of each microphone element to digital signal, it is to carry out when the data that each frame is moved are offset that the dividing frequency band of frame length is processed.In the dividing frequency band signal obtaining, only have τ=1 ..., k is used to estimate Sounnd source direction, and calculates steering vector matrix A (f).

By A (f) for searching for τ=1 ... the true solution of dividing frequency band signal.Gained optimal solution is synthesized, to obtain the Signal estimation of each sound source.The Signal estimation of each synthetic sound source is output signal.This output signal is for each sound source, to isolate the signal of sound, and produces the sound of the content of speaking of easily understanding each sound source.

By reference to the accompanying drawings embodiments of the invention are described above; but the present invention is not limited to above-mentioned embodiment; above-mentioned embodiment is only schematic; rather than restrictive; those of ordinary skill in the art is under enlightenment of the present invention; enclose in situation not departing from the norm that aim of the present invention and claim protect, also can make a lot of forms, within these all belong to protection of the present invention.

Claims (10)

1. a mobile communication equipment, comprise casing (1), equipment body (2), receiver (3) and microphone (4), described equipment body (2) is built in described casing (1), it is characterized in that: described receiver (3) involving vibrations lamina membranacea and be fixed on the piezoelectric ceramic piece (31) in described diaphragm panel, described diaphragm panel stretches out and forms described casing (1).

2. mobile communication equipment according to claim 1, it is characterized in that, described microphone (4) comprises acoustical-electrical transducer and impedance transformer (41), described acoustical-electrical transducer comprises microphone array (42), and described microphone array (42) comprises described diaphragm panel.

3. mobile communication equipment according to claim 2, is characterized in that, described acoustical-electrical transducer also comprises sound source separation module (55), and described sound source separation module (55) is for separating of the different sound source of at least two groups.

4. mobile communication equipment according to claim 3, is characterized in that, described sound source separation module (55) comprising:

A/D signal conversion unit (43), for converting analog signal to digital signal, described analog signal is from least two described microphone arrays (42);

Dividing frequency band unit (49), for described digital signal is carried out to dividing frequency band, to be converted to frequency domain input;

Error minimal solution computing unit (51), for each frequency band, described error minimal solution computing unit (51) comprises at least two sound source vectors, according to the default steering vector of described sound source vector, calculate estimated signal, the disaggregation of described error minimal solution computing unit (51) output has minimal error between described estimated signal and the input of described frequency domain;

Optimal models computing unit (53), each the frequency band selection Frequency Domain Solution that is used to described solution to concentrate, described Frequency Domain Solution is minimum l _pthe weighted sum of norm and described error;

And signal synthesis unit (54), for described Frequency Domain Solution is transformed into time domain.

5. mobile communication equipment according to claim 4, is characterized in that, described sound source separation module (55) also comprises auditory localization parts (47), and described guiding vector obtains by described auditory localization parts (47).

6. mobile communication equipment according to claim 5, is characterized in that, described sound source separation module (55) also comprises direction power calculation parts (48) and direction search parts (50).

7. mobile communication equipment according to claim 6, it is characterized in that, described error minimal solution computing unit (51) calculates the minimal error solution of sound source vector described in each, and the null value sound source quantity of described sound source vector equates, and zero valued elements quantity equates.

8. mobile communication equipment according to claim 7, is characterized in that, described optimal models computing unit (53) is concentrated and selected to separate from the described error minimal solution of output, the sliding average that described solution is described error and l _pthe weighted sum of the described sliding average of norm.

9. mobile communication equipment according to claim 8, is characterized in that, described sound source separation module (55) also comprises CPU (44) and memory cell (45,46).

10. according to the mobile communication equipment described in claim 1-9 any one, it is characterized in that, described piezoelectric ceramic piece (31) is bonded in the inner side of described casing (1).