CN103620992B - Balancer and equalization methods - Google Patents

Abstract

Purpose is to provide a kind of can receive the technology of signal by accurate compensation under circumstances.2nd wave filter (303) is according to the 1st filtering signal (s(t)) and the 2nd filter factor (w(t)) generate the 2nd filtering signal (c(t)), compensator (304) is according to the 2nd filtering signal (c(t)) docking collect mail number (r(t)) distortion components compensate.Error transfer factor device (306) is according to distorted signal (p(t)) and the 1st filtering signal (s(t)) at least any one and the 2nd filtering signal (c(t)), judge for calculate the 2nd filter factor error signal (e(t)) accuracy, and according to this accuracy correction error signal (e(t)).Coefficient calculator (305) is according to the 1st filtering signal (s(t)) and by the error signal (e(t) after error transfer factor device (306) correction), calculate the 2nd filter factor (w(t) used in the 2nd wave filter (303)).

Description

Balancer and equalization methods

Technical field

The present invention relates to balancer and the equalization methods that the transmission path distortion receiving signal after using OFDM modulation mode to be modulated is compensated.

Background technology

It is said that in general, in the environment of receiving wireless signal while movement, there are the following problems: the multipath fading caused due to the interference of electric wave and the drastically variation of the transmission path circumstances with mobile generation, the signal quality to up to receptor easily deteriorates.Even if it is therefore desirable to for the new mobile reception technique being also accurately received signal under severe transmission path circumstances.

As one of this technology for carrying out signal reception accurately, there is transmission path equalization technology.In this transmission path equalization technology, it is possible to by the signal arriving receptor is multiplied by suitable filter factor, compensate the variation (transmission path distortion) of the transmission path circumstances that (equilibrium) produces because of noise etc..That is, according to this technology, it is possible to reduce multipath fading, the distorted signals of adjoint high-speed mobile and the effect of noise of receptor institute superposition.

Fig. 1 illustrates the balancer with the present invention associates, the example of application transport path estimation technology balancer.It addition, in the following description, the balancer shown in Fig. 1 is referred to as " association balancer ".Prime at this association balancer is provided with the broadcast singal the receiving OFDM modulation mode receptor as reception signal, and this reception signal is imported into association balancer.

In the association balancer shown in Fig. 1, first reception signal is input to the 1st reference signal test section 1.In the 1st reference signal test section 1, extract the reference signal that this reception signal is comprised.Herein, reference signal is the signal using specific frequency and amplitude to send, such as in the reception signal with the configuration of the signal shown in Fig. 2, with the discrete reference signal 11(sand shape shade being inserted into data signal 12 in frequency direction and time orientation at regular intervals) corresponding.The reference signal extracted is exported distortion detection portion 2 by the 1st reference signal test section 1.

Distortion detection portion 2, in the case of the input with this reference signal, calculates the distortion components of this reference signal.The amplitude component of the reference signal extracted by the 1st reference signal test section 1 is detected in such as distortion detection portion 2, and with this amplitude component divided by known ideal amplitude value, i.e. send time the amplitude of reference signal.Further, distortion detection portion 2 using utilize this division arithmetic to obtain value as the distortion components of reference signal, would indicate that the distorted signal of this distortion components exports time interpolation filtering portion 3.According to this method, it is possible to the distortion level of reference signal 11 is quantized.

Temporal interpolation filtering part 3 is by carrying out signal processing to the distorted signal from distortion detection portion 2, generate in addition to the distortion components of the reference signal represented by distorted signal, be also represented by the 1st filtering signal of the distortion components of a part (the most sometimes referred to as " part data signal ") for data signal.It addition, part data signal refers to such as data signal between adjacent reference signal on time orientation, its distortion components is estimated according to reference signal and filter factor.

Interpolation across frequency filtering part 4 is by carrying out signal processing to the 1st filtering signal from temporal interpolation filtering part 3, generate in addition to the distortion components of the reference signal represented by the 1st filtering signal and part data signal, be also represented by the 2nd filtering signal of the distortion components of the remainder (the most sometimes referred to as " remaining data signal ") of data signal.It addition, remaining data signal refers to data signal between adjacent reference signal and part data signal the most in a frequency direction, its distortion components is estimated according to the 1st filtering signal and filter factor.

As it has been described above, Interpolation across frequency filtering part 4 generates the 2nd filtering signal of the distortion components of the distortion components of expression reference signal and whole (part data signal and the remaining data signals) of data signal, and the 2nd filtering signal is exported equilibrium portion 5.Equilibrium portion 5 will receive signal divided by the 2nd filtering signal.Thus, the docking collection of letters number transmission path distortion comprised compensates.

It addition, in above such transmission path equalization technology (transmission path distortion method of estimation), it is known to following learning algorithm: the action parameter that suitable coefficient of determination computational algorithm is used when transmission path estimation environment, thus improves distortion compensation capability.If using such as learning algorithm suitably to determine temporal interpolation filtering part 3 and/or the filter factor of Interpolation across frequency filtering part 4, then can improve temporal interpolation filtering part 3 and/or the estimated accuracy of Interpolation across frequency filtering part 4, it is possible to increase distortion compensation capability.Especially, in being present in the rear side of system and estimating the Interpolation across frequency filtering part 4 of distortion of more data signal, it is desirable to higher estimated accuracy, therefore it is said that in general, utilize learning algorithm suitably to determine the filter factor of Interpolation across frequency filtering part 4.

Then, use this Fig. 1 that this learning algorithm is described.The 2nd filtering signal generated in Interpolation across frequency filtering part 4 is imported into equilibrium portion 5, and is also input to the 2nd reference signal test section 6.2nd reference signal test section 6 extracts in the distortion components represented by the 2nd filtering signal, the distortion components of reference signal, and this distortion components is exported Error Calculation portion 7.

Error Calculation portion 7 is by distortion components based on the reference signal from distortion detection portion 2 with from error signal (such as representing the error signal of difference of two distortion components) output of comparative result of distortion components of reference signal of the 2nd reference signal test section 6 to coefficient calculations portion 8.

Coefficient calculations portion 8 have references to the 1st filtering signal from temporal interpolation filtering part 3 and the learning algorithm of the error signal from Error Calculation portion 7 by use, calculates the filter factor of Interpolation across frequency filtering part 4.

According to the association balancer taking above method, it is possible to suitably transmission path estimation distortion.But, in the environment of signal power to noise power ratio (CNR:Carrier-to-Noise power Ratio) when receiving is relatively low, exists and can not carry out the problem that the situation of coefficient calculations accurately is more.During additionally, use the method under the weak electric field environment that received electric field strength is the most weak, there is coefficient calculations speed and reduce and the problem of transmission path follow performance degradation.Further, the learning algorithm of in association balancer, to calculate Interpolation across frequency filtering part 4 coefficient only carries out action on the opportunity of input reference signal, therefore there is the calculating speed of this algorithm and the problem that convergence precision is limited.

Therefore, in order to solve this problem, propose following method: the judgement signal (such as representing the signal of hard decision (hard determination) result) obtained by determinant after utilizing transmission path distortion to compensate, that improves coefficient calculations algorithm follows speed and precision.

Such as, Patent Document 1 discloses following method: the docking collection of letters number value and the transmission path estimation value that obtain divided by hard decision result compare, and carry out the coefficient calculations of Interpolation across frequency filtering part 4 according to the error signal representing this comparative result.According to the method, not only to reference signal and also therefore total data signal application factor computational algorithm can be expected the raising of transmission path estimation precision.

Additionally, Patent Document 2 discloses following method: definition represents the error signal of the signal differential before and after hard decision, by using this error signal to come the CNR of the signal after transmission path estimation distortion compensation, the thus action of control coefrficient computational algorithm.According to the method, in the case of the transmission path distortion composition remained after transmission path distortion compensates and noise contribution are relatively big, it is possible to make the having stable behavior of coefficient calculations algorithm.

Citation

Patent documentation

Patent documentation 1: Japanese Unexamined Patent Application Publication 2008-543186 publication

Patent documentation 2: No. 4459507 publications of Japanese Patent Publication No.

Summary of the invention

Invent problem to be solved

But, in the method disclosed in patent documentation 1, do not consider the accuracy of transmission path estimation value, therefore in the environment of the easy step-down of accuracy of transmission path estimation value, in the environment of i.e. received signal power and noise power are contended with or under the most weak weak electric field environment of received electric field strength, possibly filter factor optimization cannot be made.

Additionally, in method as patent documentation 2, in the environment of the CNR of the signal after transmission path distortion compensates is relatively low, desired action may be cannot be carried out.Specifically, in the environment of the noise power that comprised of signal after transmission path distortion compensates and desired signal power are contended with, the probability of the hard decision carrying out mistake uprises, and estimates CNR according to the hard decision result of mistake sometimes.In these circumstances it is possible to cannot accurately compensate for receiving signal.

The present invention completes the most in view of the above problems, its object is to provide a kind of technology that can accurately compensate under circumstances receiving signal.

For solving the means of problem

The transmission path distortion received in signal of the balancer of the present invention OFDM modulation mode to comprising reference signal and data signal compensates, in this balancer, there is the distortion detector of the distorted signal of the distortion components generating the described reference signal that the described reception signal of expression is comprised.Described balancer has: the 1st wave filter, and it generates the distortion components of the described reference signal of expression and the 1st filtering signal of the distortion components of the described data signal of part according to described distorted signal；2nd wave filter, it generates the distortion components of the described reference signal of expression and the 2nd filtering signal of the distortion components of whole described data signal according to described 1st filtering signal and filter factor；And compensator, the distortion components of described reception signal is compensated by it according to described 2nd filtering signal.Described balancer has: error transfer factor device, it is according at least any one and described 2nd filtering signal in described distorted signal and described 1st filtering signal, judge the accuracy being used for calculating the error signal of described filter factor, and correct described error signal according to this accuracy；And coefficient calculator, it, according to described 1st filtering signal with by the described error signal after the correction of described error transfer factor device, calculates the described filter factor that described 2nd wave filter uses.

Invention effect

According to the present invention, based at least any one and the 2nd filtering signal in distorted signal and the 1st filtering signal, it is determined that for calculating the accuracy of the error signal of filter factor, and correct error signal according to this accuracy.Therefore, it is possible to the action of control coefrficient computational algorithm adaptively, therefore, it is possible to accurately compensate under circumstances receiving signal.

Accompanying drawing explanation

Fig. 1 is the block diagram of the structure example illustrating association balancer.

Fig. 2 is the figure of the example illustrating the signal configuration received in signal.

Fig. 3 is the block diagram of the structure of the balancer illustrating embodiment 1.

Fig. 4 is the block diagram of the structure of the distortion detector illustrating embodiment 1.

Fig. 5 is the block diagram of the structure example of the 1st wave filter illustrating embodiment 1.

Fig. 6 is the block diagram of the structure example of the 2nd wave filter illustrating embodiment 1.

Fig. 7 is the block diagram of the structure of the compensator illustrating embodiment 1.

Fig. 8 is the block diagram of the structure of the error transfer factor device illustrating embodiment 1.

Fig. 9 is the figure illustrating the relation between received electric field strength, electric field grade and the coefficient of the 2nd correction signal.

Figure 10 is the figure of the relation between the coefficient illustrating variance yields and the 3rd correction signal.

Figure 11 is the figure of the action of the balancer for embodiment 1 is described.

Figure 12 is the figure of the action of the balancer for embodiment 1 is described.

Figure 13 is the figure of the action of the balancer for embodiment 1 is described.

Figure 14 is the figure of the action of the balancer for embodiment 1 is described.

Figure 15 is the block diagram of the structure of the balancer illustrating embodiment 2.

Figure 16 is the block diagram of the structure of the error transfer factor device illustrating embodiment 2.

Figure 17 is the block diagram of the structure of the balancer illustrating embodiment 3.

Figure 18 is the block diagram of the structure of the error transfer factor device illustrating embodiment 3.

Figure 19 is the block diagram of the structure of the signal intensity determinant illustrating embodiment 3.

Figure 20 is the block diagram of the structure of the noise intensity determinant illustrating embodiment 3.

Figure 21 is the block diagram of the structure of the error transfer factor device illustrating embodiment 4.

Figure 22 is the block diagram of the structure of the signal intensity determinant illustrating embodiment 4.

Figure 23 is the block diagram of the structure of the noise intensity determinant illustrating embodiment 4.

Detailed description of the invention

＜ embodiment 1 ＞

Fig. 3 is the block diagram of the structure of the balancer illustrating embodiment of the present invention 1.As shown in Figure 3, balancer 300 has distortion detector 301, the 1st wave filter 302 using the 1st filter factor to be filtered input signal, uses the 2nd wave filter 303, compensator 304, coefficient calculator 305 and the error transfer factor device 306 being filtered input signal as the 2nd filter factor of variable filter factor, and this balancer 300 is transfused to receive signal r(t) and error signal e (t).It addition, the mark in article represents the moment carrying out signal processing to (t) of each signal, but for the sake of simplicity each signal in figure is eliminated (t).

This balancer 300 uses the signal obtained through distortion detector the 301, the 1st wave filter the 302 and the 2nd wave filter 303, docking collections of letters r(t) transmission path distortion compensate (equilibrium), and by thus obtained equalizing signal q(t) export arrive determinant 307.

Additionally, the balancer 300 of present embodiment exporting and the output of the 2nd wave filter 303 according to the 1st wave filter 302, it is determined that for calculating the accuracy of the error signal e (t) of the 2nd filter factor.Further, balancer 300 according to this accuracy correction error signal e (t), and according to thus obtained correction error signal g(t) calculate the 2nd filter factor of the 2nd wave filter 303.Balancer 300 according to such present embodiment, can adaptively control to carry out the action of the coefficient calculations algorithm of above-mentioned compensation (equilibrium), even if therefore receiving signal r(t) noise power that comprised and desired signal power contend with in the environment of or the most weak weak electric field environment of received electric field strength inferior, it is also possible to stably accurate compensation receives signal.Hereinafter, the structure of the balancer 300 of such present embodiment is illustrated.

First, the reception signal r(t of balancer 300 it is input to) it is imported into distortion detector 301.Herein, the reception signal r(t of balancer 300 it is input to) it is the signal of the OFDM modulation mode comprising reference signal and data signal.Additionally, as receiving signal r(t), it is desirable to it is the output signal from the time-to-frequency-domain converter as representative with discrete Fourier transform (DFT) (DCT:Discrete Fourier Transform) or fast fourier transform (FFT:Fast Fourier Transform).

In addition, receiving signal r(t) reference signal that comprised is contemplated to be the signal using specific frequency and amplitude to send, specifically, as in figure 2 it is shown, be contemplated to be the discrete reference signal 11 being inserted at regular intervals in frequency direction and time orientation in data signal 12.It addition, in the present embodiment, illustrating to receive signal r(t) reference signal that comprised is discrete reference signal 11, but this reference signal is not limited to discrete reference signal 11.It addition, in the following description, the most discrete reference signal 11 is referred to as " reference signal 11 ".

Distortion detector 301 from receive signal r(t) detection reference signal 11 distortion components, and generate expression this distortion components distorted signal p(t).Further, distortion detector 301 is by this distorted signal p(t) export to the 1st wave filter 302.

Fig. 4 is the block diagram of the structure example illustrating distortion detector 301.As shown in Figure 4, distortion detector 301 has reference signal extractor 401, distortion detection divider 402 and ideal signal generator 403.

Reference signal extractor 401 from receive signal r(t) extraction reference signal 11.Additionally, reference signal 11 is inserted into data signal 12 at predetermined intervals according to time/frequency, the most such as, can be become the time switch of conducting state by the time every regulation and selected and extract the selector of the signal being modulated to regulation carrier wave to constitute reference signal extractor 401.The discrete reference signal 11 extracted is exported distortion detection divider 402 by reference signal extractor 401.

Distortion detection divider 402 by the reference signal from reference signal extractor 401 divided by the known ideal signal generated in preferable signal generator 403, and export its result as distorted signal p(t).Herein, this ideal signal expectation be expected in the case of the most not by transmission path distortion and influence of noise from reference signal extractor 401 export preferable reference signal identical.According to the distortion detector 301 taking this method, it is possible to generate the distorted signal p(t quantized by the distortion level of reference signal 11).

Return Fig. 3,1st wave filter 302 is according to the distorted signal p(t from distortion detector 301), estimate the distortion components of a part (part data signal 12) for data signal 12, and generate the distortion components of this part data signal 12 of expression and the 1st filtering signal s(t of the distortion components of reference signal 11).Further, the 1st wave filter 302 is by the 1st filtering signal s(t) export to the 2nd wave filter 303, coefficient calculator 305 and error transfer factor device 306.

It addition, as it has been described above, the reception signal r(t of present embodiment) comprise reference signal 11 as shown in Figure 2.In this case, it is desirable to by distorted signal p(t) wave filter being filtered about its time orientation is applied to the 1st wave filter 302.I.e. it is desired to be transmitted the interpolation (interpolation) of the time orientation of path distortion components according to the distortion components of reference signal 11 adjacent on time orientation.According to this wave filter, the 1st filtering signal s(t can be generated), 1st filtering signal s(t) not only represent distorted signal p(t) represented by the distortion components of reference signal 11, the distortion components of the part data signal 12 being also represented by time orientation between adjacent reference signal 11.

Fig. 5 is the block diagram illustrating the concrete structure example when finite impulse response (FIR) mode filter utilizing Digital Signal Processing is applied to 1 wave filter 302.The 1st wave filter 302 shown in this Fig. 5 has the 1st retarder group 501, m the 1st multiplier the 502, the 1st adder 503 and the storage being made up of the 1st filtering the 1st delayer～the 1st filtering m delayer and has the 1st filtering the 1st coefficient～the 1st filter factor memorizer 504 of the 1st filtering m coefficient being equivalent to the 1st filter factor.

First, the distorted signal p(t of the 1st wave filter 302 it is input to) it is imported into the 1st retarder group 501 according to its time sequencing.Input signal is postponed the time T of regulation by each delayer output of the 1st retarder group 501₁After signal.And, these delayers are attached, the 1st filtering is made to become signal to become distorted signal p(t with the input signal of the 1st delayer), and the input of the 1st filtering kth delayer (2≤k≤m-1) and output respectively becomes the 1st and filters and filter with the input of (k+1) delayer with the output of (k-1) delayer and the 1st.

1st filtering jth coefficient (1≤j≤m) of the 1st filter factor memorizer 504 is multiplied with the output of the 1st filtering jth delayer in the 1st corresponding multiplier 502, and its multiplied result is output to the 1st adder 503.

Finally, m the multiplied result inputted is added by the 1st adder 503, and exports its result as the 1st filtering signal s(t).According to the 1st wave filter 302 taking this method, it is possible to distortion components based on reference signal 11 generates above-mentioned 1st filtering signal s(t).

Additionally, herein, illustrate that the 1st wave filter 302 is the situation of the finite impulse response (FIR) mode filter utilizing Digital Signal Processing, but be not limited to this, can also be unlimited impulse type wave filter based on Digital Signal Processing, or can also be wave filter based on analog signal processing.

Return Fig. 3, input, to the 2nd wave filter 303, the 1st filtering signal s(t generated by the 1st wave filter 302) and the 2nd filter factor w(t that calculated by coefficient calculator 305) i.e. the individual 2nd filtering jth coefficient w of n_j(t) (1≤j≤n).It addition, coefficient calculator 305 will be described in detail afterwards.

2nd wave filter 303 is according to the 1st filtering signal s(t from the 1st wave filter 302) and from the 2nd filter factor w(t of coefficient calculator 305), estimate the distortion components of the remainder (remaining data signal 12) of data signal 12.Further, the 2nd wave filter 303 generates expressed portion divided data signal 12 and whole distortion components of the distortion components of remaining data signal 12, i.e. data signal 12, and the 2nd filtering signal c(t of the distortion components of reference signal 11).2nd wave filter 303 is by the 2nd filtering signal c(t) export compensator 304 and error transfer factor device 306.

It addition, as it has been described above, the reception signal r(t of present embodiment) comprise reference signal 11 as shown in Figure 2.In this case, it is desirable to by the 1st filtering signal s(t) wave filter being filtered about its frequency direction is applied to the 2nd wave filter 303.I.e. it is desired to be transmitted the interpolation (interpolation) of the frequency direction of path distortion components according to the distortion components of the most adjacent reference signal 11 and part data signal 12.According to this wave filter, the 2nd filtering signal c(t can be generated), 2nd filtering signal c(t) not only represent the 1st filtering signal s(t) represented by reference signal 11 and the distortion components of part data signal 12, be also represented by the distortion components of remaining data signal 12 between adjacent reference signal 11 and part data signal 12 in a frequency direction.

Fig. 6 is the block diagram illustrating concrete structure example when finite impulse response (FIR) mode filter based on Digital Signal Processing is applied to 2 wave filter 303.The 2nd wave filter 303 shown in this Fig. 6 has the 2nd retarder group 601 being made up of the 2nd filtering the 1st delayer～the 2nd filtering the n-th delayer, multiple 2nd multiplier the 602 and the 2nd adder 603.

First, it is input to the 1st filtering signal s(t of the 2nd wave filter 303) every certain symbolic unit, frequency order from high to low or order from low to high according to subcarrier (subcarrier) are imported into the 2nd retarder group 601.Input signal is postponed the time T of regulation by each delayer output of the 2nd retarder group 601₂After signal.And, these delayers are attached, the 2nd filtering is made to become the 1st filtering signal s(t with the input signal of the 1st delayer), and the input of the 2nd filtering kth delayer (2≤k≤n-1) and output respectively becomes the 2nd and filters and filter with the input of (k+1) delayer with the output of (k-1) delayer and the 2nd.

The 2nd filter factor w(t from coefficient calculator 305) the i.e. the 2nd filtering jth coefficient w_jT () (1≤j≤n) is multiplied with the output of the 2nd filtering jth delayer in the 2nd corresponding multiplier 602, its multiplied result is output to the 2nd adder 603.

Finally, n the multiplied result inputted is added by the 2nd adder 603, and exports its result as the 2nd filtering signal c(t).According to the 2nd wave filter 303 taking this method, it is possible to distortion components based on reference signal 11 and part data signal 12 generates above-mentioned 2nd filtering signal c(t).

Additionally, herein, illustrate that the 2nd wave filter 303 is the situation of finite impulse response (FIR) mode filter based on Digital Signal Processing, but be not limited to this, can also be unlimited impulse type wave filter based on Digital Signal Processing, or can also be wave filter based on analog signal processing.Additionally, T time delay of the 1st wave filter 302₁, and T time delay of the 2nd wave filter 303₂Identical value, number m of the delayer of the 1st retarder group 501 and number n of the 2nd retarder group 601 can be set to and can also be set to identical value.Further, the 1st multiplier the 502 and the 2nd multiplier 602 can be set to identical structure, and the 1st adder the 503 and the 2nd adder 603 can also be set to identical structure.

Return Fig. 3, compensator 304 is according to the 2nd filtering signal c(t from the 2nd wave filter 303) docking collections of letters r(t) distortion components compensate (equilibrium), and by thus obtained equalizing signal q(t) export arrive determinant 307.

Fig. 7 is the block diagram of the concrete structure example of the compensator 304 illustrating and carrying out this action.Compensator 304 shown in this Fig. 7 have be transfused to receive signal r(t) and the 2nd filtering signal c(t) equilibrium divider 701.This equilibrium divider 701 reception signal r(t comprising transmission path distortion) divided by representing that the i.e. transmission path distortion of distortion components estimated by above-mentioned steps estimates the 2nd filtering signal c(t of composition).Therefore, it is possible to utilize compensator 304 to obtain the equalizing signal q(t that transmission path distortion composition is compensated).

Return Fig. 3, it is determined that device 307 receives the equalizing signal q(t from compensator 304) as the output of balancer 300.And, it is determined that device 307 is to this equalizing signal q(t) it is decoded, and export thus obtained decoding signal d(t).It addition, this determinant 307 such as can be made up of the arithmetical unit carrying out hard decision, it is also possible to be made up of the arithmetical unit carrying out soft decision.

Error transfer factor device 306 is according to the 1st filtering signal s(t from the 1st wave filter 302) and from the 2nd filtering signal c(t of the 2nd wave filter 303) correction error signal e (t), and generate correction error signal g(t).It addition, hereinafter this error transfer factor device 306 will be described in detail.

Coefficient calculator 305 is according to the 1st filtering signal s(t) and error signal e (t) after being corrected by error transfer factor device 306, calculate the 2nd filter factor w(t used in the 2nd wave filter 303).In the present embodiment, coefficient calculator 305 is according to the 1st filtering signal s(t) and correction error signal g(t), calculate n the 2nd filtering jth coefficient w_j(t) (1≤j≤n).More specifically, coefficient calculator 305 is by the 2nd filter factor w(t-τ that will obtain before current time in error transfer factor device 306), n correction error signal g(t-τ) ..., g(t-n τ) and the individual 1st filtering signal s(t-τ of n) ..., s(t-n τ) be updated to following formula (1), calculate the 2nd current filter factor w(t) (the 2nd filtering jth coefficient w_j(t) (1≤j≤n)).Wherein, τ is the time quantum of regulation.

[formula 1]

w_j(t)=w_j(t-τ)+g(t-jτ)s(t-jτ)…(1)

Then, error transfer factor device 306 is described in detail.In the present embodiment, input the 1st filtering signal s(t to error transfer factor device 306), the 2nd filtering signal c(t) and error signal e (t).Herein, error signal e (t) refer to for calculate the 2nd filter factor w(t) signal, such as can use the 1st filtering signal s(t) and the 2nd filtering signal c(t) define as following formula (2).

[formula 2]

E (t)=c (t)-s (t) ... (2)

Additionally, error signal e (t) such as can use the equalizing signal q(t from compensator 304) and from the decoding signal d(t of determinant 307) define as following formula (3).If defined as following formula (3), then it is contemplated that transmission path distortion composition and noise contribution.

[formula 3]

E (t)=d (t)-q (t) ... (3)

Additionally, error signal e (t) such as can also consider that the accuracy corresponding with the degree of transmission path distortion defines as following formula (4).

[formula 4]

E (t)={ d (t)-q (t) }/{ c (t) }²…(4)

Additionally, error signal e (t) such as can also consider that the degree with transmission path distortion and the accuracy receiving signal amplitude corresponding define as following formula (5).

[formula 5]

e(t)=r(t)×{d(t)-q(t)}/{c(t)}²…(5)

It is explained above the example of error signal e (t), but is not limited to example above.

It addition, error transfer factor device 306 is according to the 1st filtering signal s(t) and the 2nd filtering signal c(t) decision errors signal e(t) accuracy.Further, error transfer factor device 306 is according to this accuracy correction error signal e (t), and generates correction error signal g(t used in coefficient calculator 305).

Fig. 8 is the block diagram of the concrete structure example of the error transfer factor device 306 illustrating and carrying out this action.Error transfer factor device 306 shown in this Fig. 8 has signal converter the 801, the 1st signal processing arithmetical unit 811, signal intensity determinant 821, noise intensity determinant 831 and error corrector 841.Further, the 1st signal processing arithmetical unit 811 was made up of equalization arithmetical unit 812 and normalization arithmetical unit 813, and error corrector 841 is made up of the 1st corrector the 842, the 2nd corrector the 843 and the 3rd corrector 844.

It is input to the 1st filtering signal s(t of error transfer factor device 306) it is imported into equalization arithmetical unit 812 and normalization arithmetical unit 813, it is input to the 2nd filtering signal c(t of error transfer factor device 306) it is imported into signal converter 801 and signal intensity determinant 821, the error signal e (t) being input to error transfer factor device 306 is imported into the 1st corrector 842.Then, each structural element of error transfer factor device 306 is illustrated.

Signal converter 801 is according to the 2nd filtering signal c(t) size generate the 1st correction signal x₁(t), and by the 1st correction signal x₁T () exports to the 1st corrector 842.Such as, signal converter 801 carries out as following formula (6)～(8) signal conversion and generates the 1st correction signal x₁(t).Wherein, α is constant.It addition, the signal conversion of signal converter 801 is not limited to this.

[formula 6]

x₁(t)=α × c (t) ... (6)

[formula 7]

x₁(t)=α × { c (t) }²…(7)

[formula 8]

X₁(t)=α×{c(t)}^1/2…(8)

1st corrector 842 is according to the 1st correction signal x₁T () correction error signal e (t) generates the 1st intermediate error signal e₁(t), and it is output to the 2nd corrector 843.Such as, the 1st corrector 842 will be to the 1st correction signal x₁The signal that t () and error signal e (t) are accumulative and obtain is as the 1st intermediate error signal e₁(t).

1st signal processing arithmetical unit 811 is to the 1st filtering signal s(t) carry out statistical disposition, generate the average 1st filtering signal s as the 1st statistical signal_ave(t) and normalization the 1st filtering signal s as the 2nd statistical signal_nrm(t).

In the present embodiment, the 1st signal processing 811(arithmetical unit equalization arithmetical unit 812) by by the unit of time of regulation to self having the signal of pending statistical disposition, the i.e. the 1st filtering signal s(t) average, thus generate average 1st filtering signal s_ave(t).Additionally, 1st signal processing 811(arithmetical unit equalization arithmetical unit 812) it is configured to carry out performing the average treatment in batches of handling averagely when signal input number of times has reached the number of times of regulation, it is also possible to and the rolling average being configured to whenever input signal carry out performing handling averagely processes.

Additionally, in the present embodiment, the 1st signal processing 811(arithmetical unit normalization arithmetical unit 813) by with self having the signal of pending statistical disposition, the i.e. the 1st filtering signal s(t) divided by average 1st filtering signal s_aveT () (being normalized) generates normalization the 1st filtering signal s_nrm(t).As above the average 1st filtering signal s generated_aveT () is imported into signal intensity determinant 821, normalization the 1st filtering signal s_nrmT () is imported into noise intensity determinant 831.

Signal intensity determinant 821 is according to average 1st filtering signal s_ave(t) and the 2nd filtering signal c(t) judge to receive signal r(t) signal intensity.It addition, as described later, receive signal r(t) signal intensity corresponding with the accuracy of error signal e (t), the judgement therefore implementing this signal intensity is corresponding with the judgement of this accuracy of enforcement.Signal intensity determinant 821 generates the 2nd correction signal x according to this signal intensity₂(t), and by the 2nd correction signal x₂T () exports to the 2nd corrector 843.It addition, the 2nd correction signal x₂T () is from the 1st intermediate error signal e of the 1st corrector 842 output for correction₁The signal of (t).Hereinafter, to the 2nd correction signal x in signal intensity determinant 821₂T the generation of () is specifically described.

The signal intensity determinant 821 of present embodiment according in the judgement of signal intensity use signal in except the 2nd filtering signal c(t) in addition to signal, average 1st filtering signal s_aveThe size of (t), it is determined that received electric field strength (i.e. receives electric field environment).Herein, signal intensity determinant 821 has for average 1st filtering signal s_aveT the size of () carries out two threshold values classified, and judge average 1st filtering signal s_aveT which in 3 the differentiation scopes being divided out by these threshold values () belong to.Further, signal intensity determinant 821, from " highfield ", " middle electric field " and " weak electric field " such 3 grades of received electric field strengths, selects and average 1st filtering signal s_aveOne grade corresponding to t differentiation scope that () is fallen into is as the result of determination of received electric field strength.It addition, in the above description, signal intensity determinant 821 is according to average 1st filtering signal s_aveT the size of () judges received electric field strength (reception electric field environment), but be not limited to this, it is also possible to judge to receive power level (reception power rating).

Additionally, signal intensity determinant 821 is according to the 2nd filtering signal c(t) size judge the 2nd filtering signal c(t) electric field grade.Such as, signal intensity determinant 821 has for the 2nd filtering signal c(t) two threshold values classifying of size, and judge the 2nd filtering signal c(t) belong in 3 the differentiation scopes being divided out by these threshold values which.Further, signal intensity determinant 821, from " high-grade ", " middle grade " and " inferior grade " such 3 grades of electric field grades, selects and the 2nd filtering signal c(t) grade corresponding to the differentiation scope that fallen into be as the result of determination of electric field grade.It addition, in the above description, signal intensity determinant 821 is according to the 2nd filtering signal c(t) size judge the 2nd filtering signal c(t) electric field grade, but be not limited to this, it is also possible to judge the 2nd filtering signal c(t) power grade.

Additionally, as it is shown in figure 9, signal intensity determinant 821 has the 2nd correction signal x₂The coefficient of (t) respectively with combination for average 1st filtering signal s_ave3 grades of received electric field strengths of (t) and for the 2nd filtering signal c(t) the table that is mapped of 9 groups of 3 grades of electric field grades.Signal intensity determinant 821, with reference to this table, the received electric field strength determined according to self and an electric field grade, selects a 2nd correction signal x₂The coefficient of (t).

Herein, in the present embodiment, by the 2nd correction signal x₂The coefficient of (t), error signal e (t) accuracy and receive signal r(t) signal intensity interrelated, the 2nd correction signal x₂T the coefficient of () is the biggest, the accuracy of error signal e (t) the highest and receive signal r(t) signal intensity the strongest.Therefore, received electric field strength that the signal intensity determinant 821 of present embodiment determines according to self and electric field grade, judge the reception signal r(t corresponding with the accuracy of error signal e (t)) signal intensity, and according to this signal intensity generate the 2nd correction signal x₂(t).

Additionally, in the present embodiment, as it is shown in figure 9, in the case of received electric field strength is middle electric field, as need not correct the 1st intermediate error signal e₁T the situation of (), from the 2nd correction signal x that signal intensity determinant 821 output factor is " 1 "₂(t).Additionally, in the case of received electric field strength is weak electric field, output has value for the factor beta more than " 1 "_HMThe 2nd correction signal x of (M=1,2,3)₂T (), in the case of received electric field strength is highfield, it is " 1 " factor beta below that output has value_LMThe 2nd correction signal x of (M=1,2,3)₂(t).That is, the signal intensity determinant 821 of present embodiment is judged to that received electric field strength is the most weak, and the signal intensity of the accuracy the highest (receiving signal r(t) of error signal e (t) is the strongest), more increase the 2nd correction signal x₂The coefficient (level of signal) of (t).

Additionally, in the weak electric field that received electric field strength is identical, the coefficient (β of the electric field grade (high-grade) bigger with the difference of received electric field strength (weak electric field)_H1) than the coefficient (β of the electric field grade (inferior grade) less with the difference of received electric field strength (weak electric field)_H3) little.Equally, in the highfield that received electric field strength is identical, the coefficient (β of the electric field grade (inferior grade) bigger with the difference of received electric field strength (highfield)_L3) than the coefficient (β of the electric field grade (high-grade) less with the difference of received electric field strength (highfield)_L1) little.That is, the signal intensity determinant 821 of present embodiment is judged to that received electric field strength is the biggest with the difference absolute value of electric field grade, and the signal intensity of the accuracy the lowest (receiving signal r(t) of error signal e (t) is the most weak), more reduce the 2nd correction signal x₂The coefficient (level of signal) of (t).

It addition, be explained above received electric field strength and the situation of electric field grade judgement accuracy that signal intensity determinant 821 determines according to self, but it is not limited to this, it is also possible to the reception power level determined according to self and power grade judge accuracy.Additionally, received electric field strength and the respective rank of electric field grade are not limited to 3, it is also possible to be 2 grades or more than 4 grades.Additionally, the 2nd correction signal x₂Coefficient (the β of (t)_HM、1、β_LM) combination be not limited to the magnitude relationship shown in Fig. 9.

The 2nd correction signal x of signal strength determinant 821 is carried out to the 2nd corrector 843 input₂(t).2nd corrector 843 is according to the 2nd correction signal x₂(t) correction the 1st intermediate error signal e₁T () generates the 2nd intermediate error signal e₂(t), and it is output to the 3rd corrector 844.Such as, the 2nd corrector 843 will be to the 2nd correction signal x₂(t) and the 1st intermediate error signal e₁T signal that () carries out adding up and obtain is as the 2nd intermediate error signal e₂(t).

Normalization the 1st filtering signal s from the 1st signal processing arithmetical unit 811 is inputted to noise intensity determinant 831_nrm(t).Noise intensity determinant 831 is according to normalization the 1st filtering signal s_nrm(t) judge the 1st filtering signal s(t) noise intensity.It addition, as described later, the 1st filtering signal s(t) noise intensity corresponding with the accuracy of error signal e (t), therefore implement this noise intensity judge corresponding with the judgement of this accuracy of enforcement.Noise intensity determinant 831 generates the 3rd correction signal x according to this noise intensity₃(t), and by the 3rd correction signal x₃T () exports to the 3rd corrector 844.It addition, the 3rd correction signal x₃T () is from the 2nd intermediate error signal e of the 2nd corrector 843 output for correction₂The signal of (t).Hereinafter, to the 3rd correction signal x in noise intensity determinant 831₃T the generation of () is specifically described.

The noise intensity determinant 831 of present embodiment is according to the signal judged for noise intensity, i.e. normalization the 1st filtering signal s_nrmT () calculates variance yields s_var(t).It addition, herein, normalization the 1st filtering signal s is inputted by the unit of time of regulation to noise intensity determinant 831_nrmT (), noise intensity determinant 831 calculates this normalization the 1st filtering signal s about the frequency of subcarrier_nrmThe variance yields of (t).

Noise intensity determinant 831 has for the variance yields s calculated_varT the size of () carries out two threshold values classified, and judge variance yields s_varT which in 3 the differentiation scopes being divided out by these threshold values () belong to.Further, noise intensity determinant 831, from " very noisy ", " middle noise " and " small noise " such 3 grades of noise intensities, selects and variance yields s_varOne grade corresponding to t differentiation scope that () falls into is as the result of determination of noise intensity.

Additionally, as shown in Figure 10, noise intensity determinant 831 has the 3rd correction signal x₃The coefficient of (t) respectively with for variance yields s_varT table that 3 grades of noise intensities of () are mapped.Noise intensity determinant 831, with reference to this table, the noise intensity determined according to self, selects a 3rd correction signal x₃The coefficient of (t).

Herein, in the present embodiment, the 3rd correction signal x₃The coefficient of (t) not only with the 1st filtering signal s(t) noise intensity corresponding, also corresponding with the accuracy of error signal e (t), the 3rd correction signal x₃T the coefficient of () is the biggest, the accuracy of error signal e (t) is the highest.Therefore, the noise intensity determinant 831 of present embodiment is according to variance yields s_varT () judges the 1st filtering signal s(t corresponding with the accuracy of error signal e (t)) noise intensity, and according to this noise intensity generation the 3rd correction signal x₃(t).

Additionally, in the present embodiment, in the case of noise intensity is small noise, output has the coefficient gamma that value is bigger_L1The 3rd correction signal x₃T (), in the case of noise intensity is very noisy, output has the coefficient gamma that value is less_L3The 3rd correction signal x₃(t).That is, noise intensity is the most weak, and the noise intensity determinant 831 of present embodiment more increases the 3rd correction signal x₃The coefficient (level of signal) of (t).It addition, correct the 2nd intermediate error signal e at needs₂In the case of (t), it is being set to the 3rd correction signal x that output factor is " 1 "₃In the case of (t), it is desirable to γ_L1≤1.Additionally, the grade of noise intensity is not limited to 3, it is also possible to be 2 grades or more than 4 grades.

3rd corrector 844 is enter to the 3rd correction signal x of self noise intensity determinant 831₃(t).3rd corrector 844 is according to the 3rd correction signal x₃(t) correction the 2nd intermediate error signal e₂T () generates above-mentioned correction error signal g(t), and it is output to coefficient calculator 305.Such as, the 3rd corrector 844 will be to the 3rd correction signal x₃(t) and the 2nd intermediate error signal e₂T signal that () carries out adding up and obtain is as correction error signal g(t).

In error corrector 841, the 1st corrector the 842～the 3rd corrector 844 carries out above-mentioned action.Therefore, error corrector 841 is according to the 1st～the 3rd correction signal x₁(t)～x₃(t) correction error signal e (t).

According to the balancer of the present embodiment being made up of above structure, for the state shown in Figure 11～Figure 14, it is possible to according to receiving signal r(t) power and CNR, carry out having considered the control of the coefficient calculations algorithm of transmission path status.Hereinafter, use Figure 11～Figure 14 that concrete action example and the effect thereof of the balancer of present embodiment are illustrated.

Figure 11 and Figure 12 is to illustrate average 1st filtering signal s sometime_ave(t) and the 2nd filtering signal s(t) the figure of an example.As shown in figure 11, the 1st filtering signal s higher in average signal level, average_aveT () is relatively big in the case of, signal intensity determinant 821 carries out received electric field strength being set to the judgement of " highfield ", being i.e. set to the judgement of strong electric field environment.On the other hand, as shown in figure 12, the 1st filtering signal s relatively low in average signal level, average_aveT () is less in the case of, signal intensity determinant 821 carries out received electric field strength being set to the judgement of " weak electric field ", being i.e. set to the judgement of weak electric field environment.

Then, under any environment shown in Figure 11 and Figure 12, subcarrier frequency f_mUnder the 2nd filtering signal c(t) electric field grade be and the H of the average signal level same degree under weak electric field environment_e.Herein, at the 2nd filtering signal c(t) electric field grade be H_eIn the case of Zuo You, signal intensity determinant 821 is by the 2nd filtering signal c(t) electric field grade be judged to " inferior grade ".

Then, in the environment of shown in Figure 11, received electric field strength and electric field grade are judged as " highfield " and " inferior grade ", and received electric field strength becomes big with the difference absolute value of electric field grade.In this case, signal intensity determinant 821 be judged to receive signal r(t) signal intensity relatively low (accuracy of error signal e (t) is relatively low), output there is the factor beta that value is less_L3The 2nd correction signal x₂(t).

On the other hand, in the environment of shown in Figure 12, received electric field strength and electric field grade are judged as " weak electric field " and " inferior grade ", and received electric field strength diminishes with the difference absolute value of electric field grade.In this case, signal intensity determinant 821 be judged to receive signal r(t) signal intensity higher (accuracy of error signal e (t) is higher), output there is the factor beta that value is bigger_H3The 2nd correction signal x₂(t).

Then, Figure 13 and Figure 14 is to illustrate the 1st filtering signal s(t sometime) the figure of an example.As shown in figure 13, at above-mentioned variance yields s_varT () is relatively big in the case of, it is desirable to signal power contends with noise power.In this case, noise intensity determinant 831 is judged to that noise intensity has, compared with strong (accuracy of error signal e (t) is relatively low), output, the coefficient gamma that value is less_L3The 3rd correction signal x₃(t).On the other hand, as shown in figure 14, at above-mentioned variance yields s_varT () is less in the case of, it is desirable to signal power becomes sufficiently large relative to noise power.In this case, noise intensity determinant 831 is judged to that noise intensity more weak (accuracy of error signal e (t) is higher), output have the coefficient gamma that value is bigger_L1The 3rd correction signal x₃(t).

Balancer according to above present embodiment and equalization methods, it is determined that for calculating the accuracy of the error signal e (t) of the 2nd filter factor, and according to this accuracy correction error signal e (t).Therefore, it is possible to the action of adaptively control coefrficient computational algorithm, therefore, it is possible to accurately compensate under circumstances receiving signal r(t).

In addition, in the present embodiment, according to the 1st filtering signal s(t) and the 2nd filtering signal c(t), it is determined that the signal intensity corresponding with the accuracy of error signal e (t) and noise intensity, and correct error signal e (t) according to these signal intensitys and noise intensity.Therefore, it is possible to according to receiving signal r(t) signal intensity (such as received electric field strength, receive power level) and CNR consider transmission path status, it is possible to control coefrficient computational algorithm exactly.Therefore, even if receiving signal r(t) noise power that comprised and desired signal power contend with in the environment of or the most weak weak electric field environment of received electric field strength under, it is also possible to stably compensate and receive signal r(t).

Additionally, in the present embodiment, according to based on average 1st filtering signal s_aveThe received electric field strength of (t)/reception power level and the 2nd filtering signal c(t) electric field grade/power grade, generate correction error signal e (t) the 2nd correction signal x₂(t).Therefore, it is possible to instantaneous value and meansigma methods both sides according to received signal strength suitably judge to transmit path circumstances, therefore, it is possible to correct error signal e (t) accurately.

Additionally, in the present embodiment, according to normalization the 1st filtering signal s_nrmThe variance yields s of (t)_varT (), generates the 3rd correction signal x being corrected error signal e (t)₃(t).Therefore, it is possible to suitably judge noise circumstance, therefore, it is possible to correct error signal e (t) accurately according to the statistic (herein for variance yields) of received signal strength.

＜ embodiment 2 ＞

Figure 15 is the block diagram of the structure of the balancer illustrating embodiment of the present invention 2.It addition, below, in the explanation about the balancer of present embodiment, to the structural element mark identical label similar with the structural element of explanation in embodiment 1, and the description thereof will be omitted.

The present embodiment difference from above-mentioned embodiment 1 is: the signal in the error transfer factor device 306 of the balancer being input to the embodiment 1 shown in pie graph 3 is the 1st filtering signal s(t), the 2nd filtering signal c(t) and error signal e (t), on the other hand, the signal being input to constitute in the error transfer factor device 316 of the balancer of present embodiment is distorted signal p(t), the 2nd filtering signal c(t) and error signal e (t).Being accompanied by this, the error transfer factor device 316 of present embodiment is according to distorted signal p(t) and the 2nd filtering signal c(t) decision errors signal e(t) accuracy, and according to this accuracy correction error signal e (t).

Figure 16 is the block diagram of the structure example of the error transfer factor device 316 illustrating the present embodiment carrying out this action.According to this Figure 16, error transfer factor device 316 does not have the 1st signal processing arithmetical unit 811, signal intensity determinant 821 and a noise intensity determinant 831 of embodiment 1, and has the 1st signal processing arithmetical unit 851, signal intensity determinant 861 and noise intensity determinant 871.Further, it is made up of equalization arithmetical unit 852 and normalization arithmetical unit 853 for the 1st signal processing arithmetical unit 851.

1st signal processing arithmetical unit 851 of present embodiment is to distorted signal p(t) carry out the 1st signal processing arithmetical unit 811 with embodiment 1 to the 1st filtering signal s(t) the identical process of the statistical disposition that carries out.That is, the 1st signal processing arithmetical unit 811 is to distorted signal p(t) carry out statistical disposition, generate averaged distorted signal p as the 1st statistical signal_ave(t) and the normalization distorted signal p as the 2nd statistical signal_nrm(t).

More specifically, the 1st signal processing 851(arithmetical unit equalization arithmetical unit 852) by by the unit of time of regulation to self having the signal of pending statistical disposition, i.e. distorted signal p(t) average, thus generate averaged distorted signal p_ave(t).Additionally, 1st signal processing 851(arithmetical unit equalization arithmetical unit 852) it is configured to carry out performing the average treatment in batches of handling averagely when signal input number of times has reached the number of times of regulation, it is also possible to and the rolling average being configured to whenever input signal carry out performing handling averagely processes.

Additionally, the 1st signal processing 851(arithmetical unit normalization arithmetical unit 853) by with self having the signal of pending statistical disposition, i.e. distorted signal p(t) divided by averaged distorted signal p_aveT () (being normalized) generates normalization distorted signal p_nrm(t).As above averaged distorted signal p generated_aveT () is imported into signal intensity determinant 861, normalization distorted signal p_nrmT () is imported into noise intensity determinant 871.

The signal intensity determinant 861 of present embodiment carries out the signal intensity determinant 821 with embodiment 1 and the roughly the same action of noise intensity determinant 831 with noise intensity determinant 871.That is, signal intensity determinant 861 is according to averaged distorted signal p_ave(t) and the 2nd filtering signal c(t) judge the reception signal r(t corresponding with the accuracy of error signal e (t)) signal intensity, and according to this signal intensity generation the 2nd correction signal x₂(t).Additionally, noise intensity determinant 871 is according to normalization distorted signal p_nrmT () judges the distorted signal p(t corresponding with the accuracy of error signal e (t)) noise intensity, and according to this noise intensity generation the 3rd correction signal x₃(t).

Balancer according to above present embodiment and equalization methods, based on distorted signal p(t) and the 2nd filtering signal c(t), judge the signal intensity corresponding with the accuracy of error signal e (t) and noise intensity, and correct error signal e (t) according to these signal intensitys and noise intensity.Therefore, even if in the case of the output of the 1st and the 2nd wave filter 302,303 has been absorbed in labile state, also be able to according to receiving signal r(t) signal intensity (such as received electric field strength, receive power level) and CNR consider transmission path status, it is possible to control coefrficient computational algorithm exactly.This effect described further below.

First, in balancer as embodiment 1, such as some reason and in the case of making the operating state of the 1st wave filter 302 become instability, the 1st filtering signal s(t) statistical property may be different from statistical property originally.More specifically, such as in the case of the invariant system at the 1st wave filter 302 has been absorbed in divergent trend, 1st filtering signal s(t) it is superimposed the noise than the higher grade of noise that originally should detect, the probability that during therefore the noise intensity at noise intensity determinant 831 judges, generation mistake judges uprises.Its result, may decision errors signal e(t exactly) accuracy.

But, even if the operating state of the 1st wave filter 302 becomes unstable, the noise produced therefrom overlaps onto distorted signal p(t) probability relatively low.Herein, according to the balancer of present embodiment, use distorted signal p(t) judge noise intensity, therefore, it is possible to independently accurately judge noise intensity in noise intensity determinant 831 with the operating state of the 1st wave filter the 302 and the 2nd wave filter 303.

Therefore, balancer according to present embodiment and equalization methods, even if in the case of the output of the 1st and the 2nd wave filter 302,303 has been absorbed in labile state, also be able to according to receiving signal r(t) signal intensity (such as received electric field strength, receive power level) and CNR consider transmission path status, it is possible to control coefrficient computational algorithm exactly.

＜ embodiment 3 ＞

Figure 17 is the block diagram of the structure of the balancer illustrating embodiment of the present invention 3.It addition, below, in the explanation about the balancer of present embodiment, to the structural element mark identical label similar with the structural element of explanation in embodiment 1, and the description thereof will be omitted.

The present embodiment difference from above-mentioned embodiment 1 is: the signal in the error transfer factor device 306 of the balancer being input to the embodiment 1 shown in pie graph 3 is the 1st filtering signal s(t), the 2nd filtering signal c(t) and error signal e (t), on the other hand, the signal being input to constitute in the error transfer factor device 326 of the balancer of present embodiment is distorted signal p(t), the 1st filtering signal s(t), the 2nd filtering signal c(t) and error signal e (t).It is accompanied by this, the error transfer factor device 326 of present embodiment is according to distorted signal p(t) and the 1st filtering signal s(t) at least any one and the 2nd filtering signal c(t) decision errors signal e(t) and accuracy, and according to this accuracy correction error signal e (t).

Figure 18 is the block diagram of the structure example of the error transfer factor device 326 illustrating the present embodiment carrying out this action.According to this Figure 18, error transfer factor device 326 has the 1st signal processing arithmetical unit 811, error corrector the 841, the 2nd signal processing arithmetical unit 881, signal intensity determinant 891 and noise intensity determinant 901.

1st signal processing arithmetical unit 811 as embodiment 1, to the 1st filtering signal s(t) carry out statistical disposition, generate the average 1st filtering signal s as the 1st statistical signal_ave(t) and normalization the 1st filtering signal s as the 2nd statistical signal_nrm(t).Here the average 1st filtering signal s generated_aveT () is imported into signal intensity determinant 891, normalization the 1st filtering signal s_nrmT () is imported into noise intensity determinant 901.

There is for 2nd signal processing arithmetical unit 881 equalization arithmetical unit 882 and normalization arithmetical unit 883 of the equalization arithmetical unit 852 and normalization same action arithmetical unit 853 that carry out with embodiment 2, carry out the process identical with the 1st signal processing arithmetical unit 851 of embodiment 2.That is, the 2nd signal processing arithmetical unit 881 is to distorted signal p(t) carry out statistical disposition, generate averaged distorted signal p as the 3rd statistical signal_ave(t) and the normalization distorted signal p as the 4th statistical signal_nrm(t).More specifically, the 2nd signal processing 881(arithmetical unit equalization arithmetical unit 882) by by the unit of time of regulation to self having the signal of pending statistical disposition, i.e. distorted signal p(t) average, thus generate averaged distorted signal p_ave(t).Additionally, the 2nd signal processing 881(arithmetical unit normalization arithmetical unit 883) by with self having the signal of pending statistical disposition, i.e. distorted signal p(t) divided by averaged distorted signal p_aveT () (being normalized) generates normalization distorted signal p_nrm(t).Here averaged distorted signal p generated_aveT () is imported into signal intensity determinant 891, normalization distorted signal p_nrmT () is imported into noise intensity determinant 901.

The signal intensity determinant 891 of present embodiment is according to average 1st filtering signal s_ave(t) and averaged distorted signal p_aveAt least any one and the 2nd filtering signal c(t in (t)) judge the reception signal r(t corresponding with the accuracy of error signal e (t)) signal intensity, and according to this signal intensity generation the 2nd correction signal x₂(t).

Figure 19 is the block diagram of the concrete structure example of the signal intensity determinant 891 illustrating and carrying out this action.Signal intensity determinant 891 shown in this Figure 19 has and is transfused to average 1st filtering signal s_ave(t) and averaged distorted signal p_ave1st comparator 892 of (t) and be transfused to output and the 2nd filtering signal c(t of the 1st comparator 892) the 1st determinant 893.

1st comparator 892 is according to average 1st filtering signal s_ave(t) and averaged distorted signal p_aveAt least any one acquirement signal in (t), and export this acquired signal as the 1st comparison signal ps_ave(t).Such as, the 1st comparator 892 is to average 1st filtering signal s_ave(t) and averaged distorted signal p_aveT () compares, and according to its comparative result, export any one in these signals as the 1st comparison signal ps_ave(t).Or, the 1st comparator 892 can export and be averaged the 1st filtering signal s_ave(t) and averaged distorted signal p_aveThe signal that t addition of () is average and obtain is as the 1st comparison signal ps_ave(t).

1st determinant 893 carries out the action roughly the same with the signal intensity determinant 821 of embodiment 1.That is, the 1st determinant 893 is according to the 1st comparison signal ps_ave(t) and the 2nd filtering signal c(t) judge the reception signal r(t corresponding with the accuracy of error signal e (t)) signal intensity, and according to this signal intensity generation the 2nd correction signal x₂(t).Hereinafter, to the 2nd correction signal x in the 1st determinant 893₂T the generation of () is specifically described.

1st determinant 893 receives signal r(t according to being used for) the signal that judges of signal intensity in except the 2nd filtering signal c(t) in addition to signal, the i.e. the 1st comparison signal ps_aveThe size of (t), it is determined that received electric field strength (i.e. receives electric field environment).Herein, the 1st determinant 893 has for the 1st comparison signal ps_aveT the size of () carries out two threshold values classified, and judge the 1st comparison signal ps_aveT which in 3 the differentiation scopes being divided out by these threshold values () belong to.Further, the 1st determinant 893, from " highfield ", " middle electric field " and " weak electric field " such 3 grades of received electric field strengths, selects and the 1st comparison signal ps_aveOne grade corresponding to t differentiation scope that () is fallen into is as the result of determination of received electric field strength.It addition, in the above description, the 1st determinant 893 is according to the 1st comparison signal ps_aveT the size of () judges received electric field strength (reception electric field environment), but be not limited to this, it is also possible to judge to receive power level (reception power rating).

Additionally, the 1st determinant 893 is according to the 2nd filtering signal c(t) size judge the 2nd filtering signal c(t) electric field grade.Such as, the 1st determinant 893 has for the 2nd filtering signal c(t) two threshold values classifying of size, and judge the 2nd filtering signal c(t) belong in 3 the differentiation scopes being divided out by these threshold values which.Further, the 1st determinant 893, from " high-grade ", " middle grade " and " inferior grade " such 3 grades of electric field grades, selects and the 2nd filtering signal c(t) grade corresponding to the differentiation scope that fallen into be as the result of determination of electric field grade.It addition, in the above description, the 1st determinant 893 is according to the 2nd filtering signal c(t) size judge the 2nd filtering signal c(t) electric field grade, but be not limited to this, it is also possible to judge the 2nd filtering signal c(t) power grade.

Additionally, the 1st determinant 893 has the 2nd correction signal x₂The coefficient of (t) respectively be combined with for the 1st comparison signal ps_ave3 grades of received electric field strengths of (t) and for the 2nd filtering signal c(t) the table that is mapped of 9 groups of 3 grades of electric field grades, i.e. there is the table identical with the table shown in Fig. 9.1st determinant 893, with reference to this table, the received electric field strength determined according to self and an electric field grade, selects a 2nd correction signal x₂The coefficient of (t).

It addition, be described herein received electric field strength and the situation of electric field grade judgement accuracy that the 1st determinant 893 determines according to self, but it is not limited to this, it is also possible to the reception power level determined according to self and power grade judge accuracy.Additionally, received electric field strength and the respective rank of electric field grade are not limited to 3, it is also possible to be 2 grades or more than 4 grades.

Returning Figure 18, the noise intensity determinant 901 of present embodiment is according to normalization the 1st filtering signal s_nrm(t) and normalization distorted signal p_nrmIn (t) at least any one, it is determined that the noise intensity of this any one signal corresponding with the accuracy of error signal e (t), and according to this noise intensity generate the 3rd correction signal x₃(t).

Figure 20 is the block diagram of the concrete structure example of the noise intensity determinant 901 illustrating and carrying out this action.Noise intensity determinant 901 shown in this Figure 20 has and is transfused to normalization the 1st filtering signal s_nrm(t) and normalization distorted signal p_nrm2nd comparator 902 of (t) and be transfused to the 2nd determinant 903 of output of the 2nd comparator 902.

2nd comparator 902 is according to normalization the 1st filtering signal s_nrm(t) and normalization distorted signal p_nrmIn (t) at least any one and obtain signal, and using this acquired signal as the 2nd comparison signal ps_nrmT () exports.Such as, the 2nd comparator 902 is to normalization the 1st filtering signal s_nrm(t) and normalization distorted signal p_nrmT () compares, and according to its comparative result, using any one in these signals as the 2nd comparison signal ps_nrmT () exports.More specifically, the 2nd comparator 902 is only at normalization the 1st filtering signal s_nrmT the variance yields size of () is less than normalization distorted signal p_nrmIn the case of the variance yields size of (t), by normalization the 1st filtering signal s_nrmT () is as the 2nd comparison signal ps_nrmT () exports.Further, the 2nd comparator 902 is in the case of in addition, by the 2nd comparison signal ps when the operating state of the 1st wave filter 302 grade sometime before is stablized_nrm(t) (hereinafter referred to as " the 2nd comparison signal ps before_nrm(t) ") as the 2nd current comparison signal ps_nrmT () exports.It addition, the 2nd comparison signal ps before Gai_nrmT () is stored in the memorizer (not shown) of the such as the 2nd comparator 902.

Or, the 2nd comparator 902 can export and take normalization the 1st filtering signal s_nrm(t) and normalization distorted signal p_nrmThe signal that t addition of () is average and obtain is as the 2nd comparison signal ps_nrm(t).Further, the 2nd comparator 902 is such as set to have predetermined threshold, at normalization the 1st filtering signal s_nrm(t) and normalization distorted signal p_nrmT any one in two signals or carry out two signals less than in the case of this predetermined threshold, can be added the average and signal that obtains as the 2nd comparison signal ps by the variance yields both sides of ()_nrmT () exports.

2nd determinant 903 carries out the action roughly the same with the noise intensity determinant 831 of embodiment 1.That is, the 2nd determinant 903 is according to the 2nd comparison signal ps_nrmT () judges the above-mentioned noise intensity corresponding with the accuracy of error signal e (t), and generate the 3rd correction signal x according to this noise intensity₃(t).Hereinafter, to the 3rd correction signal x in the 2nd determinant 903₃T the generation of () is specifically described.

2nd determinant 903 is according to the 2nd comparison signal ps_nrmT () calculates variance yields ps_var(t).Herein, the 2nd determinant 903 has for the variance yields ps calculated_varT the size of () carries out two threshold values classified, and judge variance yields ps_varT which in 3 the differentiation scopes being divided out by these threshold values () belong to.Further, the 2nd determinant 903, from " very noisy ", " middle noise " and " small noise " such 3 grades of noise intensities (noise circumstance), selects and variance yields ps_varOne rank corresponding to t differentiation scope that () is fallen into is as the result of determination of noise intensity (noise circumstance).

Additionally, the 2nd determinant 903 has the 3rd correction signal x3(t) coefficient respectively with for variance yields ps_varT table that 3 grades of noise intensities of () are mapped, i.e. identical with the table shown in Figure 10 table.2nd determinant 903, with reference to this table, the noise intensity determined according to self, selects a 3rd correction signal x₃The coefficient of (t).It addition, the grade of noise intensity is not limited to 3 grades, it is also possible to be 2 grades or more than 4 grades.

Balancer according to above such present embodiment and equalization methods, according to the 1st filtering signal s(t), distorted signal p(t) and the 2nd filtering signal c(t), judge the signal intensity corresponding with the accuracy of error signal e (t) and noise intensity, and correct error signal e (t) according to these signal intensitys and noise intensity.Therefore, even if in the case of the output of the 1st wave filter 302 has been absorbed in labile state, also be able to the signal that will obtain according to the comparative result of output of the output of distortion detector 301 and the 1st wave filter 302, as above-mentioned for obtain the signal of accuracy and with continued reference to.

Below in order to more specifically describe this effect in detail, it is set to the 2nd comparator 902 the most only at normalization the 1st filtering signal s_nrmT the variance yields of () is less than normalization distorted signal p_nrmIn the case of the variance yields of (t), export normalization the 1st filtering signal s_nrmT () is as the 2nd comparison signal ps_nrm(t).Further, the 2nd comparator 902 export in the case of in addition above-mentioned before the 2nd comparison signal ps_nrmT () is as the 2nd comparison signal ps_nrm(t).

Now, when the operating state of the 1st wave filter 302 becomes instability, normalization the 1st filtering signal s_nrmT the variance yields of () compares the normalization distorted signal p corresponding with the input of the 1st wave filter 302_nrmT the variance yields of () increased.In the case of Gai, not by normalization the 1st filtering signal s_nrmT () is as the 2nd comparison signal ps_nrmT () exports, and by the 2nd comparison signal ps before_nrmT () is as the 2nd current comparison signal ps_nrmT () exports.

Therefore, balancer according to above such present embodiment and equalization methods, even if in the case of the output of the 1st wave filter 302 has been absorbed in labile state, also be able to according to receiving signal r(t) signal intensity (such as received electric field strength, receive power level) and CNR, consider transmission path status, thus control coefrficient computational algorithm exactly.Therefore, even if receiving signal r(t) noise power that comprised and desired signal power contend with in the environment of or the most weak weak electric field environment of received electric field strength under with the mobile high speed variation producing transmission path time, it is also possible to stably compensate and receive signal r(t).

＜ embodiment 4 ＞

The block diagram of the structure of the balancer of embodiment of the present invention 4 and the balancer illustrating embodiment 3 are shown The block diagram (Figure 19) of structure is identical.It addition, below, in the explanation about the balancer of present embodiment, to the structural element mark identical label similar with the structural element of explanation in embodiment 3, and the description thereof will be omitted.

Constitute the error transfer factor device 336 of balancer of present embodiment as the error transfer factor device 326 of embodiment 3, according to distorted signal p(t) and the 1st filtering signal s(t) at least any one and the 2nd filtering signal c(t) carry out decision errors signal e(t) and accuracy, and according to this accuracy correction error signal e (t).

Figure 21 is the block diagram of the structure example of the error transfer factor device 336 illustrating the present embodiment carrying out this action.According to this Figure 21, error transfer factor device 336 has the 1st signal processing arithmetical unit 811, error corrector the 841, the 2nd signal processing signal processing arithmetical unit the 881, the 3rd arithmetical unit 911, signal intensity determinant 921 and noise intensity determinant 931.There is for 3rd signal processing arithmetical unit 911 equalization arithmetical unit 912 and normalization arithmetical unit 913.

1st with the structure of the 2nd signal processing arithmetical unit 811,881 and action and explanation identical in embodiment 3.But following aspect is different from embodiment 3: average 1st filtering signal s_ave(t) and averaged distorted signal p_aveT () is imported into signal intensity determinant 921；And normalization the 1st filtering signal s_nrm(t) and normalization distorted signal p_nrmT () is imported into noise intensity determinant 931.

3rd signal processing arithmetical unit 911 is to the 2nd filtering signal c(t) carry out the 1st signal processing arithmetical unit 811 with embodiment 1 to the 1st filtering signal s(t) the identical process of the statistical disposition that carries out.That is, the 3rd signal processing arithmetical unit 911 is to the 2nd filtering signal c(t) carry out statistical disposition, generate the average 2nd filtering signal c as the 5th statistical signal_ave(t) and normalization the 2nd filtering signal c as the 6th statistical signal_nrm(t).

More specifically, the 3rd signal processing 911(arithmetical unit equalization arithmetical unit 912) by by the unit of time of regulation to self having the signal of pending statistical disposition, the i.e. the 2nd filtering signal c(t) average, thus generate average 2nd filtering signal c_ave(t).Additionally, the 3rd signal processing 911(arithmetical unit normalization arithmetical unit 913) by with self having the signal of pending statistical disposition, the i.e. the 2nd filtering signal c(t) divided by average 2nd filtering signal c_aveT () (being normalized) generates normalization the 2nd filtering signal c_nrm(t).Here the average 2nd filtering signal c generated_aveT () is imported into signal intensity determinant 921, normalization the 2nd filtering signal c_nrmT () is imported into noise intensity determinant 931.

The signal intensity determinant 921 of present embodiment is according to average 1st filtering signal s_ave(t), averaged distorted signal p_ave(t) and average 2nd filtering signal c_aveAt least any one and the 2nd filtering signal c(t in (t)), it is determined that the reception signal r(t corresponding with the accuracy of error signal e (t)) signal intensity, and generate the 2nd correction signal x according to this signal intensity₂(t).

Figure 22 is the block diagram of the concrete structure example of the signal intensity determinant 921 illustrating and carrying out this action.Signal intensity determinant 921 shown in this Figure 22 has and is transfused to average 1st filtering signal s_ave(t), averaged distorted signal p_ave(t) and average 2nd filtering signal c_ave3rd comparator 922 of (t)；And be transfused to output and the 2nd filtering signal c(t of the 3rd comparator 922) the 3rd determinant 923.

3rd comparator 922 is according to average 1st filtering signal s_ave(t), averaged distorted signal p_ave(t) and average 2nd filtering signal c_aveIn (t) at least any one and obtain signal, and using this acquired signal as the 3rd comparison signal psc_aveT () exports.Such as, the 3rd comparator 922 is to average 1st filtering signal s_ave(t), averaged distorted signal p_ave(t) and average 2nd filtering signal c_aveT () compares, and according to its comparative result, using any one in these signals as the 3rd comparison signal psc_aveT () exports.Or, the 3rd comparator 922 can also will be averaged the 1st filtering signal s_ave(t), averaged distorted signal p_ave(t) and average 2nd filtering signal c_aveThe signal that t addition of () is average and obtain is as the 3rd comparison signal ps_caveT () exports.

3rd determinant 923 is except by the 1st comparison signal ps of explanation in embodiment 3_aveT () is changed to the 3rd comparison signal psc_aveBeyond the aspect of (t), carry out the action identical with the 1st determinant 893 of embodiment 3.That is, the 3rd determinant 923 is according to the 3rd comparison signal psc_ave(t) and the 2nd filtering signal c(t), it is determined that the reception signal r(t corresponding with the accuracy of error signal e (t)) signal intensity, and generate the 2nd correction signal x according to this signal intensity₂(t).Hereinafter, to the 2nd correction signal x in the 3rd determinant 923₂T the generation of () is specifically described.

3rd determinant 923 according to receive signal r(t) signal intensity judge in use signal in except the 2nd filtering signal c(t) in addition to signal, the i.e. the 3rd comparison signal psc_aveThe size of (t), it is determined that received electric field strength (i.e. receives electric field environment).Herein, the 3rd determinant 923 has for the 3rd comparison signal psc_aveT the size of () carries out two threshold values classified, and judge the 3rd comparison signal psc_aveT which in 3 the differentiation scopes being divided out by these threshold values () belong to.Further, the 3rd determinant 923, from " highfield ", " middle electric field " and " weak electric field " such 3 grades of received electric field strengths, selects and the 3rd comparison signal psc_aveOne grade corresponding to t differentiation scope that () is fallen into is as the result of determination of received electric field strength.It addition, in the above description, the 3rd determinant 923 is according to the 3rd comparison signal ps_caveT the size of () judges received electric field strength (reception electric field environment), but be not limited to this, it is also possible to judge to receive power level (reception power rating).

Additionally, the 3rd determinant 923 is according to the 2nd filtering signal c(t) size judge the 2nd filtering signal c(t) electric field grade.Such as, the 3rd determinant 923 has for the 2nd filtering signal c(t) two threshold values classifying of size, and judge the 2nd filtering signal c(t) belong in 3 the differentiation scopes being divided out by these threshold values which.Further, the 1st determinant 893, from " high-grade ", " middle grade " and " inferior grade " such 3 grades of electric field grades, selects and the 2nd filtering signal c(t) rank corresponding to the differentiation scope that fallen into be as the result of determination of electric field grade.It addition, in the above description, the 3rd determinant 923 is according to the 2nd filtering signal c(t) size judge the 2nd filtering signal c(t) electric field grade, but be not limited to this, it is also possible to judge the 2nd filtering signal c(t) power grade.

Additionally, the 3rd determinant 923 has the 2nd correction signal x₂The coefficient of (t) respectively be combined with for the 3rd comparison signal psc_ave3 grades of received electric field strengths of (t) and for the 2nd filtering signal c(t) the table that is mapped of 9 groups of 3 grades of electric field grades, i.e. there is the table identical with the table shown in Fig. 9.3rd determinant 923, with reference to this table, the received electric field strength determined according to self and an electric field grade, selects a 2nd correction signal x₂The coefficient of (t).

It addition, be described herein received electric field strength and the situation of electric field grade judgement accuracy that the 3rd determinant 923 determines according to self, but it is not limited to this, it is also possible to the reception power level determined according to self and power grade judge accuracy.Additionally, received electric field strength and the respective rank of electric field grade are not limited to 3 grades, it is also possible to be 2 grades or more than 4 grades.

Returning Figure 21, the noise intensity determinant 931 of present embodiment is according to normalization the 1st filtering signal s_nrm(t), normalization distorted signal p_nrm(t) and normalization the 2nd filtering signal c_nrmIn (t) at least any one, it is determined that the noise intensity of this any one signal corresponding with the accuracy of error signal e (t), and according to this noise intensity generate the 3rd correction signal x₃(t).

Figure 23 is the block diagram of the concrete structure example of the noise intensity determinant 931 illustrating and carrying out this action.Noise intensity determinant 931 shown in this Figure 23 has and is transfused to normalization the 1st filtering signal s_nrm(t), normalization distorted signal p_nrm(t) and normalization the 2nd filtering signal c_nrm4th comparator 932 of (t)；And it is transfused to the 4th determinant 933 of the output of the 4th comparator 932.

4th comparator 932 is according to normalization the 1st filtering signal s_nrm(t), normalization distorted signal p_nrm(t) and normalization the 2nd filtering signal c_nrmIn (t) at least any one and obtain signal, and using this acquired signal as the 4th comparison signal psc_nrmT () exports.Such as, the 4th comparator 932 is to normalization the 1st filtering signal s_nrm(t), normalization distorted signal p_nrm(t) and normalization the 2nd filtering signal c_nrmT () compares, and according to its comparative result, using any one in these signals as the 4th comparison signal psc_nrmT () exports.More specifically, the 4th comparator 932 is only at normalization the 1st filtering signal s_nrmT the variance yields size of () is less than normalization distorted signal p_nrmThe variance yields size of (t), and normalization the 2nd filtering signal c_nrmT the variance yields size of () is less than normalization the 1st filtering signal s_nrmIn the case of the variance yields size of (t), by normalization the 1st filtering signal s_nrmT () is as the 4th comparison signal psc_nrmT () exports.Further, the 4th comparator 932 is in the case of in addition, by the 4th comparison signal psc when the operating state of the 1st wave filter 302 grade sometime before is stablized_nrm(t) (hereinafter referred to as " the 4th comparison signal psc before_nrm(t) ") as the 4th current comparison signal psc_nrmT () exports.It addition, the 4th comparison signal psc before Gai_nrmT () is stored in the memorizer (not shown) of the such as the 4th comparator 932.

Or, the 4th comparator 932 can also will take normalization the 1st filtering signal s_nrm(t), normalization distorted signal p_nrm(t) and normalization the 2nd filtering signal c_nrmThe signal that t addition of () is average and obtain is as the 4th comparison signal psc_nrmT () exports.Further, the 4th comparator 932 is such as set to have predetermined threshold, at normalization the 1st filtering signal s_nrm(t), normalization distorted signal p_nrm(t) and normalization the 2nd filtering signal c_nrmT () is entirely below in the case of this predetermined threshold, any one in these 3 kinds of signals or carry out these signals is added the average and signal that obtains as the 4th comparison signal psc_nrmT () exports.

4th determinant 933 carries out the action roughly the same with the 2nd determinant 903 of embodiment 3.That is, the 4th determinant 933 is according to the 4th comparison signal psc_nrmT () judges the above-mentioned noise intensity corresponding with the accuracy of error signal e (t), and generate the 3rd correction signal x according to this noise intensity₃(t).Hereinafter, to the 3rd correction signal x in the 4th determinant 933₃T the generation of () is specifically described.

4th determinant 933 is according to the 4th comparison signal psc_nrmT () calculates variance yields psc_var(t).Herein, the 4th determinant 933 has for the variance yields psc calculated_varT the size of () carries out two threshold values classified, and judge variance yields psc_varT which in 3 the differentiation scopes being divided out by these threshold values () belong to.Further, the 4th determinant 933, from " very noisy ", " middle noise " and " small noise " such 3 grades of noise intensities (noise circumstance), selects and variance yields psc_varOne rank corresponding to t differentiation scope that () is fallen into is as the result of determination of noise intensity (noise circumstance).

Additionally, the 4th determinant 933 has the 3rd correction signal x₃The coefficient of (t) respectively with for variance yields psc_varT table that 3 grades of noise intensities of () are mapped, i.e. identical with the table shown in Figure 10 table.4th determinant 933, with reference to this table, the noise intensity determined according to self, selects a 3rd correction signal x₃The coefficient of (t).It addition, the rank of noise intensity is not limited to 3 grades, it is also possible to be 2 grades or more than 4 grades.

Balancer according to above such present embodiment and equalization methods, according to the 1st filtering signal s(t), distorted signal p(t) and the 2nd filtering signal c(t), judge the signal intensity corresponding with the accuracy of error signal e (t) and noise intensity, and correct error signal e (t) according to these signal intensitys and noise intensity.Therefore, even if in the case of the output of the 2nd wave filter 303 has been absorbed in labile state, also be able to the signal obtained by the comparative result that the output according to distortion detector 301, the output of the 1st wave filter 302 and the 2nd wave filter 303 export, as above-mentioned for obtain the signal of accuracy and with continued reference to.

Below in order to more specifically describe this effect in detail, it is set to the 4th comparator 932 the most only at normalization the 1st filtering signal s_nrmT the variance yields size of () is less than normalization distorted signal pnrm(t) variance yields size, and normalization the 2nd filtering signal c_nrmT the variance yields size of () is less than normalization the 1st filtering signal s_nrmIn the case of the variance yields size of (t), by normalization the 1st filtering signal s_nrmT () is as the 4th comparison signal psc_nrmT () exports.Further, the 4th comparator 932 is in the case of in addition, by the 4th comparison signal psc before above-mentioned_nrmT () is as the 4th comparison signal psc_nrmT () exports.

Now, when the operating state of the 1st wave filter 302 becomes instability, normalization the 1st filtering signal s_nrmT the variance yields of () compares the normalization distorted signal p corresponding with the input of the 1st wave filter 302_nrmT the variance yields of () increased.In the case of Gai, not by normalization the 1st filtering signal s_nrmT () is as the 4th comparison signal psc_nrmT () exports, and by the 4th comparison signal psc before_nrmT () is as the 4th current comparison signal psc_nrmT () exports.

Therefore, balancer according to above such present embodiment and equalization methods, even if in the case of the output of the 2nd wave filter 303 has been absorbed in labile state, also be able to according to receiving signal r(t) signal intensity (such as received electric field strength, receive power level) and CNR, consider transmission path status, thus control coefrficient computational algorithm exactly.Therefore, even if receiving signal r(t) noise power that comprised and desired signal power contend with in the environment of or the most weak weak electric field environment of received electric field strength under with the mobile high speed variation producing transmission path time, it is also possible to stably compensate and receive signal r(t).

Being described in detail the present invention, but described above is all to illustrate in form of ownership, the present invention is not limited to this.It is interpreted as the infinite variety example can not being exemplified without departing from the present invention.

Label declaration

11: reference signal；12: data signal；300: balancer；301: distortion detector；302: the 1 wave filter；303: the 2 wave filter；304: compensator；305: coefficient calculator；306,316,326: error transfer factor device；801: signal converter；811,851: the 1 signal processing arithmetical units；821,861,891,921: signal intensity determinant；831,871,901,931: noise intensity determinant；841: error corrector；881: the 2 signal processing arithmetical units；911: the 3 signal processing arithmetical units.

Claims (9)

1. a balancer, connecing of its OFDM modulation mode to comprising reference signal and data signal Transmission path distortion in the collection of letters number compensates, and this balancer has:

Distortion detector, it generates the distortion of the distortion components representing the described reference signal comprised in described reception signal Signal；

1st wave filter, it, according to described distorted signal, generates distortion components and the part institute representing described reference signal State the 1st filtering signal of the distortion components of data signal；

2nd wave filter, it, according to described 1st filtering signal and filter factor, generates the mistake representing described reference signal 2nd filtering signal of the distortion components of true composition and whole described data signal；

Compensator, the distortion components of described reception signal is compensated by it according to described 2nd filtering signal；

Error transfer factor device, it is according at least any one and the institute in described distorted signal and described 1st filtering signal State the 2nd filtering signal, it is determined that for calculating the accuracy of the error signal of described filter factor, and according to this accuracy Correct described error signal；And

Coefficient calculator, it is according to described 1st filtering signal with by the described error letter after the correction of described error transfer factor device Number, calculate the described filter factor that described 2nd wave filter uses.

Balancer the most according to claim 1, wherein,

Described error transfer factor utensil has:

Signal converter, it generates the 1st correction signal according to described 2nd filtering signal；

1st signal processing arithmetical unit, it carries out statistical disposition, generates the 1st statistical signal described 1st filtering signal With the 2nd statistical signal；

Signal intensity determinant, its according to described 1st statistical signal and described 2nd filtering signal judge with described accurately The signal intensity of the described reception signal that degree is corresponding, and generate the 2nd correction signal according to this signal intensity；

Noise intensity determinant, it judges described 1st corresponding with described accuracy according to described 2nd statistical signal The noise intensity of filtering signal, and generate the 3rd correction signal according to this noise intensity；And

Error corrector, it corrects described error signal according to described 1st correction signal～described 3rd correction signal.

Balancer the most according to claim 1, wherein,

Described error transfer factor utensil has:

Signal converter, it generates the 1st correction signal according to described 2nd filtering signal；

1st signal processing arithmetical unit, it carries out statistical disposition to described distorted signal, generates the 1st statistical signal and the 2 statistical signals；

Signal intensity determinant, its according to described 1st statistical signal and described 2nd filtering signal judge with described accurately The signal intensity of the described reception signal that degree is corresponding, and generate the 2nd correction signal according to this signal intensity；

Noise intensity determinant, it judges the described distortion letter corresponding with described accuracy according to described 2nd statistical signal Number noise intensity, and according to this noise intensity generate the 3rd correction signal；And

Error corrector, it corrects described error signal according to described 1st correction signal～described 3rd correction signal.

Balancer the most according to claim 1, wherein,

Described error transfer factor utensil has:

Signal converter, it generates the 1st correction signal according to described 2nd filtering signal；

1st signal processing arithmetical unit, it carries out statistical disposition, generates the 1st statistical signal described 1st filtering signal With the 2nd statistical signal；

2nd signal processing arithmetical unit, it carries out statistical disposition to described distorted signal, generates the 3rd statistical signal and the 4 statistical signals；

Signal intensity determinant, its most any one according in described 1st statistical signal and described 3rd statistical signal Individual and described 2nd filtering signal, it is determined that the signal intensity of the described reception signal corresponding with described accuracy, and root The 2nd correction signal is generated according to this signal intensity；

Noise intensity determinant, its most any one according in described 2nd statistical signal and described 4th statistical signal Individual, it is determined that the noise intensity of this any one signal corresponding with described accuracy, and generate the according to this noise intensity 3 correction signals；And

Error corrector, it corrects described error signal according to described 1st correction signal～described 3rd correction signal.

Balancer the most according to claim 1, wherein,

Described error transfer factor utensil has:

Signal converter, it generates the 1st correction signal according to described 2nd filtering signal；

1st signal processing arithmetical unit, it carries out statistical disposition, generates the 1st statistical signal described 1st filtering signal With the 2nd statistical signal；

2nd signal processing arithmetical unit, it carries out statistical disposition to described distorted signal, generates the 3rd statistical signal and the 4 statistical signals；

3rd signal processing arithmetical unit, it carries out statistical disposition, generates the 5th statistical signal described 2nd filtering signal With the 6th statistical signal；

Signal intensity determinant, it is according to described 1st statistical signal, described 3rd statistical signal and described 5th statistics At least any one and described 2nd filtering signal in signal, it is determined that the described reception letter corresponding with described accuracy Number signal intensity, and according to this signal intensity generate the 2nd correction signal；

Noise intensity determinant, it is according to described 2nd statistical signal, described 4th statistical signal and described 6th statistics In signal at least any one, it is determined that the noise intensity of this any one signal corresponding with described accuracy, and root The 3rd correction signal is generated according to this noise intensity；And

Error corrector, it corrects described error signal according to described 1st correction signal～described 3rd correction signal.

6. according to the balancer described in any one in claim 2～5, wherein,

Described signal intensity determinant according in the judgement of described signal intensity use signal in except the described 2nd Signal beyond filtering signal, it is determined that received electric field strength/reception power level, and according to this received electric field strength/connect Electric field grade/the power grade receiving power level and described 2nd filtering signal judges described signal intensity.

Balancer the most according to claim 6, wherein,

Described signal intensity determinant is judged to, described received electric field strength/reception power level and described 2nd filtering The absolute value of the difference of the described electric field grade/power grade of signal is the biggest, and described accuracy is the lowest, and reduces described The level of signal of the 2nd correction signal.

8. according to the balancer described in any one in claim 2～5, wherein,

Described noise intensity determinant judges according to the variance yields of the signal used in the judgement of described noise intensity Described noise intensity.

9. an equalization methods, the reception to the OFDM modulation mode comprising reference signal and data signal Transmission path distortion in signal compensates, and in this equalization methods, comprises the following steps:

Step a, generation represent the distorted signal of the distortion components of the described reference signal comprised in described reception signal；

Step b, according to described distorted signal, generate distortion components and the described data of part representing described reference signal 1st filtering signal of the distortion components of signal；

Step c, according to described 1st filtering signal and filter factor, generate the distortion components representing described reference signal The 2nd filtering signal with the distortion components of whole described data signals；

Step d, according to described 2nd filtering signal, the distortion components of described reception signal is compensated；

Step e, according in described distorted signal and described 1st filtering signal at least any one and the described 2nd Filtering signal, it is determined that for calculating the accuracy of the error signal of described filter factor, and correct institute according to this accuracy State error signal；And

Step f, according to described 1st filtering signal and by described step e correct after described error signal, calculate The described filter factor used in described step c.