CN100502374C - Channel balancer of VSB receiving system and channel balancing method - Google Patents

Abstract

The invention can make the periodically received synchronizing signal in each field trace impulse response of channel, and figure out absolute value or two power of difference between impulse response in current field and impulse response in previous field in order to detect variation of channel. After that, the variation of channel is compared with preset threshold to decide the character of channel is dynamical channel or static channel. Finally, according the character of channel controls the parameter in filter in channel equalizer to improve equalizing performance of channel.

Description

The channel equalization means of VSB receiving system and channel equalization method thereof

Technical field

The present invention relates to the radio-frequency (RF) digital broadcast receiver system, particularly the channel equalization means and the installation method thereof of the radio-frequency (RF) digital TV signal transmitted of VSB mode.

Background technology

The U.S. has selected ATSC (the Advanced Television standard committee) 8T-VSB transmission means in order to advance radio-frequency (RF) digital broadcasting for use in nineteen ninety-five, and since the formal operation second half year in 1998.Korea S also selects the ATSC 8T-VSB load mode identical with American Way for use, begun experimental broadcast in May nineteen ninety-five, entered the stage of piloting on August 31st, 2000, and since October calendar year 2001, each big broadcaster formally implements this broadcast mode.

In addition, digital communication system is with the digital information (voice, data or image) of sign map transmission end, and is each symbol transition and its size or the proportional analog signal of phase place to be sent to receiving terminal by transfer channel.For the example of a receiving terminal, the signal that arrives the ATSC8T-VSB receiving system produces and the interference of facing mutually between signal by the multi-path transfer channel time, and its state is seriously distorted.Therefore, be original signal, must be adopted as the channel equalizer of compensate for channel in order to restore from the received signal of distorting.

The general channel equalizer that the most generally uses has the decision zeedback equalizer (DFE) of suitable lowest mean square (LMS:Least MeanSquare) algorithm.Above-mentioned DFE is a kind of when received signal enters by the multi-path channel, the path of peak signal is made as main path, other paths are considered as the intersymbol interference (ISI) or the double-image signals (GhostSignal) that enter by reflection path, correction form signals phase place and size only to entering by main path, and extract, and remove the mode of the signal that enters from other paths.

Fig. 1 represents, the pie graph of the common decision zeedback equalizer that moves in the certain hour scope.

Simple observation Fig. 1 is as follows, go up the path signal that arrives earlier than main path by the 101 elimination times of front end filter, promptly eliminate the influence of in-plant dual imaging, to remove then by back end filter or feedback filter 102, than the path signal that arrives behind the main path, i.e. the influence of remote dual imaging.At this moment, adder 105 is exported to detection unit 103 with the dateout and the closing of feedback filter 102 dateouts of above-mentioned leading portion filter 101, above-mentioned detection unit 103 is compared the signal of above-mentioned adder 105 outputs with the standard signal grade of having set, electing decision content as with the nearest standard signal grade of adder 105 output signals.At this moment, the output of detection unit 103 will feed back to feedback filter 102 and control part 104.That is, as the input of above-mentioned feedback filter 102, not the dateout of adder 105, but the decision content that draws by detection unit 103.

On the other hand, control part 104 receives the output of equalizers, i.e. the output valve of adder 105 and decision content, and upgrade the coefficient of front end filter 101 and feedback filter 102.

In this decision zeedback equalizer, assign correct judgement as above-mentioned detection unit 103, will export under the situation of noise in the composition at the removing equalizer, re-enter feedback filter 102, noise can not take place increases phenomenon.Therefore, can show the performance more superior than common linear equalizer.

At this moment, above-mentioned decision zeedback equalizer will use the LMS algorithm as the method for upgrading filter coefficient.The step-length of using in the LMS algorithm (step-size) will be according to the length of filter and the distortion situation of channel, the higher limit of decision channel equalizer steady operation.In this higher limit, step-length is big more, in Dynamic Signal, the performance of the channel impulse response that compensation changes can be good more, but in static channel, residual (excess) mean square error (MSE:MeanSquare Errror) can increase, and the noise immunity (immunity) that channel is taken place reduces.

In addition, also according in the interval of not training order (training sequence), generally be used for blindness (blind) algorithm of applicable channel equalizer, the equalization performance in dynamic channel equalization performance and the static channel also can be influenced.

Summary of the invention

In view of the above, the purpose of this invention is, as described in the ATSC8T-VSB receiving system, utilizes channel impulse response to judge whether to be dynamic or static channel, and according to the parameter of the suitable control channel equalizer of characteristics of channel, so that channel equalization means and the method thereof that improves the channel equalization performance to be provided.

To achieve these goals, the channel equalization means in the ATSC 8T-VSB receiving system of the present invention divides following several sections to constitute characteristics generally: follow the trail of channel impulse response from received signal by transfer channel, and the tracking part of output; The channel impulse response that utilization is followed the trail of from the existing field of above-mentioned channel tracking portion output judges with the difference of the channel impulse response of former field tracking whether channel is the channel characteristic test section of dynamic or static characteristics; According to the characteristics that detect from above-mentioned channel characteristic test section, control the parameter of inner filter, compensate the channel equalization portion of the channel distortion that is included in received signal with this.Wherein, above-mentioned channel characteristic test section comprises: store above-mentioned channel tracking portion and follow the trail of, and behind the existing field channel impulse response of output, as the memory of former field channel impulse response output; Draw that above-mentioned channel tracking portion is followed the trail of and the former field channel impulse response of the existing channel impulse response of output and the output of above-mentioned memory poor, and the subtracter of output; Obtain the absolute value of difference of above-mentioned subtracter output or 2 powers of absolute value, draw the operational part of channel variation amount; The accumulative total portion of the channel variation amount of above-mentioned operational part output with the interval accumulative total of setting accumulative total; The comparator that the value of above-mentioned accumulative total portion accumulative total is compared with the critical value of having set.

Above-mentioned channel tracking portion characteristics are when the synchronizing signal of each input field, utilize the training order that is included in the synchronizing signal of above-mentioned field, follow the trail of the impulse response of channel.

Above-mentioned channel characteristic test section is to constitute characteristics with following part: after storing the existing field channel impulse response of above-mentioned channel tracking portion's tracking and output, as the memory of former field channel pulse information output; The subtracter of the difference output of the channel impulse response in the existing field of obtaining above-mentioned channel tracking portion's tracking and exporting and the former field channel response of above-mentioned memory output; Obtain absolute value or 2 powers of above-mentioned subtracter output difference, draw the operational part of the variable quantity of channel with this; In setting the accumulative total interval, the accumulative total portion of the channel variation amount of the above-mentioned operational part output of accumulative total; The value of accumulative total portion accumulative total is compared with first critical value of having set, if aggregate-value greater than first critical value, just can be judged as dynamic channel with channel characteristic; If not quite then, just be judged as static channel, and this judged result be transferred to the comparator of channel equalization portion.

Above-mentioned channel characteristic test section constitutes characteristics with following part: after storing the existing field channel impulse response of tracking of channel tracking portion and output, as the memory of former field channel pulse information output; Draw the subtracter of the difference output of the channel impulse response in existing field that channel tracking portion follows the trail of output and the impulse response of field channel in the past; Obtain absolute value or 2 powers of the difference of above-mentioned subtracter output, draw the operational part of the variable quantity of channel; Accumulative total portion according to the channel variation amount of setting the above-mentioned operational part output of the interval accumulative total of accumulative total; By a plurality of fields, it is average to obtain above-mentioned accumulative total portion aggregate-value, and the average portion of output; The mean value of average portion output is compared with second critical value of having set, is dynamic channel if mean value, is just judged channel characteristic greater than second critical value; If little, then be judged as static channel, judged result is outputed to the comparator of channel equalization portion.

Above-mentioned channel characteristic test section constitutes characteristics with following part: after storing the existing field channel impulse response of above-mentioned channel tracking portion's tracking and output, as the memory of former field channel pulse output; Draw that above-mentioned channel tracking portion is followed the trail of and the former field channel response of the existing field channel impulse response of output and the output of above-mentioned memory poor, and export these subtracter; Obtain absolute value or 2 powers of the difference of above-mentioned subtracter output, draw the operational part of the variable quantity of channel; The accumulative total portion of the channel variation amount of the above-mentioned operational part output of accumulative total in setting the accumulative total interval; The comparator that the value of above-mentioned accumulative total portion accumulative total is compared with the 3rd critical value of having set; Increasing or subtract the value of counting according to the comparative result of above-mentioned comparator, simultaneously this count value is compared with the counting critical value of having set, is dynamic channel if count value, is just judged channel characteristic greater than critical value; As not quite, just be judged as static channel, and this judged result outputed to the reliability counter of above-mentioned channel equalization portion.

The characteristics of above-mentioned channel equalization portion are to implement channel equalization with (blind) accommodation mode blindly, if channel characteristic belongs to dynamic channel, just utilize the G-Pseudo mode to obtain mistake; If belong to static channel, just obtain mistake in the Stop-and-Go mode, upgrade filter coefficient.

Above-mentioned channel equalization portion guides accommodation mode to carry out channel equalization to judge, if channel characteristic is judged as dynamic channel, just step value is heightened; If be judged as static channel, the level when just making step-length be lower than dynamic channel is implemented the renewal of filter factor.

The channel equalization method of the VSB receiving system of this invention is, as distorting from the digital broadcast signal compensate for channel that receives by channel, the channel equalization method that is intended to recover the channel equalizer in the VSB receiving system of original signal is described, and the branch following several stages constitutes characteristics:

(a) when obtaining field synchronizing signal input, from follow the trail of the stage of channel impulse response by the received signal of transfer channel at every turn;

The channel impulse response that follow the trail of in the existing field of (b) above-mentioned (a) being exported in the stage is compared with the critical value of having set with the difference of the channel impulse response of field tracking in the past, judges with this whether channel characteristic is the dynamic or static stage;

(c) according to above-mentioned detected channel characteristic of (b) stage, the inner filter parameter of control channel equalizer compensates the stage of the channel distortion that is included in the acknowledge(ment) signal with this.

Wherein, above-mentioned (b) stage comprises: obtain the poor of existing field channel impulse response that above-mentioned (a) stage follows the trail of and field channel impulse response in the past, and obtain the absolute value of this value or 2 powers of absolute value, the stage of calculating channel variable quantity; Will be in the accumulative total interval of having set, after the channel variation amount that the above-mentioned stage of accumulative total calculates, the stage of accumulative total; The value of above-mentioned accumulative total is compared with the critical value of having set,, otherwise be judged as the stage of static channel if aggregate-value greater than critical value, just is judged as dynamic channel with channel characteristic.

Above-mentioned (b) stage is to constitute characteristics as the next stage: obtain the poor of existing field channel impulse response that above-mentioned (a) stage follows the trail of and field channel impulse response in the past, and obtain absolute value or 2 powers of this value, the stage of calculating channel variable quantity; Will be after the channel variation amount of the above-mentioned stage calculating of the interval accumulative total of the accumulative total of having set, the stage of accumulative total; The value of above-mentioned accumulative total is compared with first critical value of having set,, otherwise be judged as the stage of static channel if aggregate-value just is judged as dynamic channel with channel characteristic greater than first critical value.

Above-mentioned (b) stage is to constitute characteristics as the next stage: obtain the poor of existing field channel impulse response that above-mentioned (a) stage follows the trail of and field channel impulse response in the past, and take out absolute value or 2 powers of above-mentioned difference, the stage of calculating channel variable quantity; After the channel variation amount that will calculate in the interval above-mentioned stage of accumulative total of the accumulative total set, aggregate-value is average with a plurality of fields, and the stage of output; The mean value in above-mentioned stage is compared with second critical value of setting, if mean value just is judged as dynamic channel to channel characteristic greater than second critical value; If not quite, just be judged as the stage of static channel.

Above-mentioned (b) stage is to constitute characteristics as the next stage: obtain the poor of existing field channel impulse response that above-mentioned (a) stage follows the trail of and field channel impulse response in the past, and draw absolute value or 2 powers of above-mentioned difference, the stage of calculating channel variable quantity; After the channel variation amount that will calculate in the interval above-mentioned stage of accumulative total of the accumulative total set, stage of aggregate-value and the 3rd critical value set relatively; In the above-mentioned stage, if aggregate-value greater than the 3rd critical value, just increases the reliability value, otherwise after reducing, will compare, if the reliability value just is judged as dynamic channel with channel characteristic greater than the counting critical value with the counting critical value of having set; Otherwise be judged as the stage of static channel.

Above-mentioned (c) stage is that above-mentioned channel equalizer is implemented channel equalization with blindness (blind) accommodation mode, if channel characteristic is judged as dynamic channel, just utilizes the G-Pseudo mode to obtain mistake; If be judged as static channel, just obtain mistake in the Stop-and-Go mode, the renewal filter factor is characteristics.

Above-mentioned (c) stage is that above-mentioned channel equalizer guides the accommodation mode operating factor to upgrade to judge, carries out in the channel equalization process, if channel characteristic is judged as dynamic channel, just enlarges step-length; If be judged as static channel, just step-length is turned down level when the above-mentioned dynamic channel, implement the characteristics that are updated to of filter coefficient with this.

Other purposes, characteristics and the strong point of this invention will more be expressly understood by the explanation of the example in the accompanying drawing.

As mentioned above, according to channel equalization means and the method thereof on the VSB receiving system of this invention, follow the trail of the channel impulse response of each the field synchronizing signal that periodically receives, and utilize the channel impulse response and the existing impulse response in the past of following the trail of, measure channel variance situation, based on this, judge whether to after dynamic channel or the static channel, will be according to the channel characteristic of judging, the parameter of suitable control channel equalizer is to improve the channel equalization performance.

Description of drawings

Fig. 1 is the formation integration map of common channel equalization means;

Fig. 2 is that channel equalization means of the present invention constitutes integration map;

Fig. 3 is the structural representation of the Frame of common VSB transmission means;

Fig. 4 is the structural representation of the synchronous segment in field of Fig. 3;

Fig. 5 is an embodiment of Fig. 2 channel characteristic test section integration map of explaining in detail;

Fig. 6 is another embodiment of Fig. 2 channel characteristic test section integration map of explaining in detail;

Fig. 7 is another embodiment of Fig. 2 channel characteristic test section integration map of explaining in detail.

Accompanying drawing major part symbol description

101: front end filter 102: feedback filter

103: detection unit 104: control part

105: adder 201: channel tracking portion

202: channel characteristic test section 203: channel equalization portion

501,601,701: memory 502,602,702: subtracter

503,603,703: operational part 504,604,704: accumulative total portion

505,606,705: comparator 605: average portion

706: the reliability counter

Embodiment

Below, formation and the effect to the embodiment of the invention is elaborated with reference to the accompanying drawings.As shown in the figure, and the formation of the present invention of explanation in view of the above is the explanation of carrying out with an example at least with effect, and above-mentioned the technology of the present invention thought and its core constitute and act on and can so not be restricted.

Characteristics of the present invention are: the channel impulse response of following the trail of each the field synchronizing signal that periodically receives, and utilize the former and present impulse response that tracks channel to measure channel variance situation, after judging dynamic or static channel based on this simultaneously, make it be fit to the channel characteristic of judging, control channel parametric equalizer.

Fig. 2 is as the formation integration map of an example of expression channel equalization means of the present invention, with when each reception field synchronizing signal is imported, from follow the trail of the channel tracking portion 201 of transfer channel impulse response output by the received signal of transfer channel; Channel impulse response according to 201 outputs of above-mentioned channel tracking portion judges whether channel characteristic is dynamic or static channel characteristic test section 202; According to channel characteristic test section 202 detected channel characteristics, the control filters parameter, compensation is included in channel equalization portion 203 formations that the channel in the received signal is distorted.

Above-mentioned channel equalization portion 203 can directly utilize above-mentioned Fig. 1, also can utilize the channel equalization portion of knowing altogether.

The present invention uses channel tracking portion and channel characteristic test section at as shown in Figure 1 decision zeedback equalizer, and improving the channel equalization performance is embodiment.That is, above-mentioned channel equalization portion 203 is made of front end filter 101, feedback filter 102, detection unit 103, control part 104, adder 105 as shown in Figure 1.

In the present invention who constitutes like this, channel tracking portion 201 when receiving field synchronizing signal input at every turn, follow the trail of the impulse response h (n) of the discontinuous equivalent channel that may pass through original signal x (n) from the input signal u (n) that receives after, the limited impulse response value H of channel _(n)Output to channel characteristic test section 202.

That is,, discontinuous equivalent channel impulse response is made as h (n) if be made as x (n) transmitting signal, when white noise is made as w (n), available following mathematical expression 1 expression of the input signal u (n) that the ATSC8T-VSB receiving system is received.

[mathematical expression 1]

$u\left(n\right)=\underset{k}{\mathrm{\Σ}}h\left(k\right)\·x(n-k)+w\left(n\right)$

So, the input that signal u (n) receives in above-mentioned channel tracking portion 201, follow the trail of the discontinuous equivalent channel impulse response h (n) that original signal x (n) may pass through after, will be the limited impulse response value H of channel _(n)Export to channel characteristic test section 202.

At this moment, the simplest method that the transfer channel impulse responses are followed the trail of by above-mentioned channel tracking portion 201 is, be assumed to be white signal transmitting the training signal that periodically adds in the signal number, and draw the simple relevant mode (Simple Correlation Method:SCM) of the cross correlation score (Cross Correlation Value) between the known training signal of the training signal that enters by channel and receiving terminal.This side is comparatively simple, so can utilize a small amount of hardware to realize.But at another side,, will exist, and the channel tracking field becomes wide more, the easier influence that is subjected to training signal two side datas, the correct tracking of impossible enforcement than big tracking error if training signal does not have white character.

In addition compared with simple relevant mode, least square method (the LSM:Least Square Method) trace mode that is acknowledged as correct method not under the situation with white character, also can correctly be followed the trail of channel at training signal.

Promptly, draw and detect the training time, and the cross correlation score p between the known training signal of the training signal that in the above-mentioned training time, enters and receiving terminal by channel, obtain the self-correlation matrix R of training signal after, in order to remove the self-relevant portion among the cross correlation score p between received signal and former training signal, carry out R ^-1The matrix computations of p can more correctly be followed the trail of channel with this.

But LS mode (LSM) is that self-correlation matrix R can become Identity (unanimity) matrix under the situation of white at training signal, therefore can draw SCM (the relevant merely mode) result identical with LSM.Simultaneously, though above-mentioned LS mode is presented as cost compared with simple relevant mode with complexity, can be followed the trail of channel more accurately, if but the channel tracking field broadens, just there is such shortcoming, promptly can be subjected to the influence that the data the same with simple relevant mode are brought.

Fig. 3 is made of odd data field and even data field with a Frame as the data frame structure of the common ATSC 8T-VSB of expression system, and each data field is divided into 313 data segments once more.Above-mentioned 313 data fragments are made of a synchronous segment in field and 312 general data segments that comprise training order (Training sequence) signal.Each data fragments is made of 4 segment synchronizing symbols and 828 data symbols.

Fig. 4 is as the drawing of the structure of the synchronous segment in expression field, form by 1 data fragment length, there are 4 symbols and data fragments synchronous mode during beginning, there are 511 PN511 and each 63 PN 63, PN 63, the PN 63 belong to pseudo random sequence (Pseudo Random Sequence) afterwards, have the information of relevant VSB pattern afterwards in 24 symbols.Here, in above-mentioned three PN63 intervals, the polarity of second PN63 all can change at every turn.That is, ' 1 ' become ' 0 ', ' 0 ' become ' 1 '.Therefore, according to the polarity of second PN63, a frame is divided into the even/odd field.

In addition, it is not use (Rererved) that there is other 104 symbol beyond 24 symbols in above-mentioned relevant VSB pattern information, at this not in the use field, and last 12 symbol datas (pre-code symbol) of segment before will duplicating in last 12 symbols.

At this moment, because of cycle in a data field is about 24.2ms, transmit 41.3 synchronous segments in field as can be known 1 second approximately.

So channel tracking of the present invention portion 201 utilizes training order such as PN511, PN63 when receiving the synchronous segment in above-mentioned field at every turn, the impulse response of also using simple relevant mode (SCM) or LS trace mode (LSM) to follow the trail of channel.

Simultaneously, the channel impulse response that above-mentioned channel characteristic test section 202 is followed the trail of existing field is compared with the critical value of having set with the difference of the channel impulse response of field tracking in the past, judges whether channel characteristic is dynamic channel or static channel.Afterwards, the channel characteristic of judging is exported to the control part 104 of channel characteristic equilibrium portion 203.

That is, the difference of the channel impulse response of the channel impulse response of existing field tracking and field tracking is in the past represented a channel variance situation between the field.Suppose that the channel impulse response that follow the trail of in existing field is H _(n), n=0,1......, N-1; The channel impulse response that follow the trail of in the field before supposing is H _{T-1 (n)}Here, N is the channel impulse response length that tracks.

The channel impulse response that follow the trail of in existing field and in the past the difference D of the channel impulse response followed the trail of of field (t, t-1) as the numerical value of expression channel impulse response, as following mathematical expression 2, can be with the absolute value or 2 the powers calculating of the difference of two channel impulse responses.

[mathematical expression 2]

$D_{(t,t-1)}=\underset{n=0}{\overset{N-1}{\mathrm{\Σ}}}|H_{t\left(n\right)}-H_{t-1\left(n\right)}|$

$D_{(t,t-1)}=\underset{n=0}{\overset{N-1}{\mathrm{\Σ}}}{|H_{t\left(n\right)}-H_{t-1\left(n\right)}|}^2$

At this moment,, promptly almost do not change if channel is static, the channel impulse response in the channel impulse response in so existing field and field in the past much at one, so the variable quantity of channel is as following mathematical expression 3, almost near 0.

[mathematical expression 3]

H _t(n)<H _t-1(n)rD(t，t-1)<0

So (t's above-mentioned channel characteristic test section 202 t-1) compares with the critical value of having set, if less than critical value, just as following mathematical expression 4, is judged as static channel the variable quantity D of channel; Greatly then be judged to be dynamic channel.

[mathematical expression 4]

IfD (t, t-1)≤critical value, the then static channel

The else dynamic channel

Fig. 5 to Fig. 7 is the detailed formation integration map that the above-mentioned channel characteristic test section 202 of expression is judged a plurality of embodiment of channel characteristic.

Fig. 5 is suitable for the integration map of the example that the channel characteristic test section of above-mentioned mathematical expression 4 does as detailed expression, substantially the memory 501 of the channel impulse response that is tracked by field before storing; Obtain the poor of the channel impulse response of the existing field tracking that will import and the channel impulse response that is stored in the former field tracking in the above-mentioned memory 501, and the subtracter 502 of output; Obtain the absolute value of difference of subtracter 502 output or the operational part 503 of 2 powers; Between with the setting district, in (N), add up the accumulative total portion 504 of (summation) above-mentioned operational part 503 outputs; And the value of above-mentioned accumulative total portion 504 accumulative totals compared with the critical value of having set, making is dynamic channel, or the judgement of static channel, and this judged result is exported to comparator 505 formations of channel equalization portion 203.At above-mentioned operational part 503, if the channel impulse response of following the trail of is a plural number, the absolute value that calculates just means Euclidean distance (Euclideandistance).

More than the subtracter 502 among the Fig. 5 of Gou Chenging is exported to operational part 503 to existing field channel impulse response of following the trail of and the difference that is stored in the channel impulse response of memory 501, after above-mentioned operational part 503 is obtained 2 powers of the absolute value of difference of two channel impulse responses or absolute value, output to accumulative total portion 504, through the length N accumulative total of impulse response.At this moment, the value of above-mentioned accumulative total portion 504 accumulative totals be equivalent to the variable quantity D of channel (t, t-1), comparator 505 is compared this aggregate-value with the critical value of having set, judgement is a dynamic channel, or after the static channel, judged result is outputed to channel equalization portion 203.

(t t-1) under near situation about often changing the critical value, in order to prevent the instability of channel characteristic test section 202 outputs, can be divided into several fields with the variable quantity of channel, draws mean value, and compares with critical value but at the variable quantity D of channel.As the averaging method that at this moment uses the forgetting factor average of the moving average of following mathematical expression 5 (Moving average) mode and mathematical expression 6 (forgetting factor average) mode is arranged.

[mathematical expression 5]

$D_{\mathrm{avg}}(t,t-1)=\frac{1}{K}\underset{k=t-K+1}{\overset{t}{\mathrm{\&Sigma;}}}D(k,k-1)$

[mathematical expression 6]

D _avg(t，t-1)＝λ·D _avg(t，t-1)+(1-λ)·D(t，t-1)

Here, 0≤forgetting factor λ≤1

Fig. 6 is the detailed integration map of an example that utilizes the channel characteristic detector of above-mentioned mathematical expression 5 or mathematical expression 6 mean values.

As shown in Figure 6, substantially by the memory 601 that stores the channel impulse response of following the trail of in former field; Obtain the channel impulse response of following the trail of in existing field and be stored in the poor of the channel impulse response the former field of above-mentioned memory 601 followed the trail of, and the subtracter 602 of output; Obtain the absolute value of difference of subtracter 602 output or the operational part 603 of 2 powers; In the accumulative total portion 604 that (N) accumulative total operational part 603 is exported between the setting district; And, and the average portion 605 of output average through a plurality of fields the value of above-mentioned accumulative total portion 604 accumulative totals; The mean value of above-mentioned average portion 605 outputs is compared with setting critical value, and judgement is a dynamic channel, or static channel, and simultaneously, the comparator 606 that this judged result is transferred to channel equalization portion 203 constitutes.

More than the subtracter 602 of the Fig. 6 of Gou Chenging is exported to operational part 603 to the channel impulse response of existing field tracking and the difference of the channel impulse response of the former field tracking that is stored in memory 601, after above-mentioned operational part 603 is obtained the absolute value or its 2 powers of difference of two channel impulse responses, export to accumulative total portion 604, through the length N accumulative total of impulse response.Above-mentioned aggregate-value will output to average portion 605.

Above-mentioned average portion 605 exports its mean value through the average above-mentioned aggregate-value in a plurality of fields, and the method for the arithmetic average of this moment has, as the moving average mode of above-mentioned mathematical expression 5 or as the forgetting factor average mode of mathematical expression 6.

Also have, the mean value of above-mentioned average portion 605 outputs will output to comparator 606, and above-mentioned comparator 606 more above-mentioned mean values and the critical value of having set judge whether into behind the dynamic channel judged result to be exported to channel equalization portion 203.

In addition, except the method for above-mentioned arithmetic average shown in Figure 6, with critical value relatively after, can utilize the reliability counting to carry out this result's final judgement.

Fig. 7 utilizes the detailed integration map of an example of the channel characteristic detector of this reliability counter as expression, the memory 701 of the channel impulse response of being followed the trail of by field before storing; Obtain the channel impulse response of following the trail of in existing field and be stored in the poor of the channel impulse response the former field of above-mentioned memory 701 followed the trail of, and the subtracter 702 of output; Obtain the absolute value of difference of subtracter 702 output or the operational part 703 of its 2 powers; In the accumulative total portion 704 that the above-mentioned operational part 703 of (N) accumulative total (summation) is exported between the setting district; And the comparator 705 that the value of above-mentioned accumulative total portion 704 accumulative totals is compared with the critical value of having set; Comparative result according to above-mentioned comparator 705, increase or subtract count value, and the count value of this moment compares with the counting critical value of having set (count_threshold), judge it is static channel with this, still after the static channel, the reliability counter 706 that judged result is outputed to channel equalization portion 203 constitutes.

More than the subtracter 702 among the Fig. 7 of Gou Chenging outputs to operational part 703 to existing field channel impulse response of following the trail of and the difference that is stored in the channel impulse response of memory 701, after above-mentioned operational part 703 is obtained 2 powers of the absolute value of difference of two channel impulse responses or absolute value, output to accumulative total portion 704, through the length N accumulative total of impulse response.The value of above-mentioned accumulative total portion 704 accumulative totals will output to comparator 705, above-mentioned comparator 705 this aggregate-value and the critical value of having set mutually relatively after, the result is outputed to reliability counter 706.

If the aggregate-value that is judged as comparator 705 is greater than the critical value of having set, above-mentioned reliability counter 706 just is equivalent to set point to the count value increase; Be judged as not quite, then count value reduced the value that is equivalent to set.Here, the value of the value of increase and minimizing can be set arbitrarily by the designer.Embodiments of the invention for example can be designed to increase and decrease and be respectively 1 or increase at 2 o'clock, reduce 1.This can be according to the self-defined and extensive use more of designer, so need not be confined to the example of above-mentioned prompting.

Afterwards, the critical value of count value with the counting of having set compared,, just be judged as dynamic channel if count value surpasses the critical value of counting.

That is, the mode of utilizing reliability counter of the present invention be if the variable quantity D of channel (t during t-1) greater than critical value, just increases the value of a reliability counter; As less than critical value the time, just reduce by one structure.As following mathematical expression 7, when the count value of reliability counter surpasses a certain counting critical value, will be judged as dynamic channel.

[mathematical expression 7]

If D (t, t-1)≤critical value, then reduces the value of reliability counter.

Increase the value of else reliability counter

Here, the value≤count_max of 0≤reliability counter

Value≤the count_threshold of if reliability counter, the then static channel

The else dynamic channel

Below, according to as the channel characteristic of above-mentioned judgement, the process of control part 104 Control Parameter of channel equalization portion 203 is described.Promptly, the output of control part 104 receive channel equalizers, be 105 the output of adder and the output of detection unit 103, and according to the judged result of channel characteristic test section 202, (for example select suitable channel equalizer parameter, improper value, step-length etc.), upgrade the coefficient of front end filter 101 and back end filter 102 with this.

In above-mentioned channel equalizer, time field decision zeedback equalizer and frequency field LMS adaptive equalizer all as the update method of filter coefficient, use the LMS algorithm.That is, in the channel equalizer as Fig. 2, the coefficient C (N) of front end filter 101 and feedback filter 102 upgrades following row mathematical expression 8.That is, after the input U (N) of filter sheet multiply by wrong e (n), multiply by the value that step size mu draws again and add existing filter coefficient c (n), to upgrade as next filter coefficient c (n+1).

[mathematical expression 8]

c(n+1)＝c(n)+μ·u(n)·e(n)

Here, e (n)=d (n)-y (n)

That is, in above-mentioned mathematical expression 8, d (n) is as the signal that needs, and interval use of synchronous segment trained order in the field; At the interval decision content that uses 103 judgements of Fig. 2 detection unit of data fragments.Like this, in the signal of needs, subtract the output y (n) that channel equalizer, and the mode of adaptive channel equalizer is called judgement guide (decision-directed) accommodation mode.

In addition, distort because of channel under the situation of the initial set that seriously is difficult to carry out channel equalizer, replace the judgement of above-mentioned explanation to guide accommodation mode to use with (blind) accommodation mode blindly.Normally used blindness accommodation mode has G-Pseudo mode (Albert Benveniste, MauriceGoursat, ＂ Blind Equalizers ＂, IEEE Tran.On Comm.Vol.COM-32, No.8, August 1984) and Stop-and-Go mode (Giorgio Picchi, GiancarloPrati, ＂ BlindEqualization and Carrier Recovery Using a Stop-and-Go Decision-DirectedAlgorithm ＂, IEEE Tran.On Comm.Vol.COM-35, No.9, September1987), as mathematical expression 9 and 10, replace the error items of mathematical expression 8 respectively, upgrade the coefficient of filter.

[mathematical expression 9]

e(n)＝k ₁e _DD(n)+k ₂|e _DD(n)|e _Blind(n)

[mathematical expression 10]

$e\left(n\right)=\left\{\begin{array}{cc}{e}_{\mathrm{DD}0}& \mathrm{if}\mathrm{sgn}\left[e_{\mathrm{DD}}\left(n\right)\right]=\mathrm{sgn}\left[e_{\mathrm{Blind}}\left(n\right)\right]\\ 0_0& \mathrm{Otherwise}\end{array}\right.$

Above-mentioned mathematical expression 9 is to obtain wrong formula in the G-Pseudo mode; Mathematical expression 10 is with stop﹠amp; The go mode is obtained wrong formula.At this moment, under the situation as channel equalizer and Fig. 2, above-mentioned wrong can obtaining at control part 104, and output to front end filter 101 and feedback filter 102.

Show as in above-mentioned mathematical expression 9 and 10:

e _DD(n)＝d(n)-y(n)

$e_{\mathrm{Blind}}\left(n\right)=\left\{\begin{array}{cc}5.25-y{\left(n\right)}_0& \mathrm{if}y\left(n\right)>0\\ -5.25-y{\left(n\right)}_0& \mathrm{Otherwise}\end{array}\right.$

Also have, k1 in mathematical expression 9 and k2 all are constants; The symbol of function sgn (x) the expression x of mathematical expression 10.Simultaneously, above-mentioned constant 5.25 and the-5.25th, on 8 rank VSB signals, add 7 ,-5 ,-3 ,-1 ,+1 ,+3 ,+5, the standard value during+7}.

At this moment, though as the G-Pseudo mode of above-mentioned mathematical expression 9 have initial set fast, the superior advantage of tracking performance in the dynamic channel, the output jitter of channel equalizer is bigger than normal and be unfavorable in static channel.On the contrary, though as the Stop-and-Go mode initial set speed of above-mentioned mathematical expression 10 slowly, the performance of dynamic channel is inferior slightly, the less than normal and static channel of the output jitter of the channel equalizer energy of playing superiority.

In view of the above, if the channel characteristic of judging according to above-mentioned channel characteristic test section 202, parameter that can the control channel equalizer also just can make the performance of channel equalizer reach the best.That is, if above-mentioned channel characteristic test section 202 is judged as dynamic channel, channel equalization portion 203 will use the G-Pseudo mode to obtain mistake as the blindness accommodation mode; If be judged as static channel, just obtain mistake in the Stop-and-Go mode, upgrade the parameter of filter.

Simultaneously,, compare, when the step-length of using when upgrading filter coefficient in the LMS mode is judged as dynamic channel, use, improve the channel equalization performance with this greater than value with static channel as above-mentioned mathematical expression 8.

In addition, overcome the defective of decision zeedback equalizer recently, improve the frequency field LMS adaptive equalizer of channel equalization performance, once proposed by applicant.

The channel characteristic monitor that utilizes the channel tracking device that the present invention proposes can only not be confined to decision zeedback equalizer, can be suitable in the same way in the LMS adaptive equalizer of said frequencies field yet.

In addition, as the output that utilizes the channel characteristic detector, the parameter of control equalizer inside, parameter-the step-length of equalizer and the blindly selection of accommodation mode have been described, but this is an embodiment only, and the testing result of channel characteristic detector can be widely used in the more areas of ATSC 8T-VSB receiver.

By above-mentioned description, the related work personnel can carry out various change and modification fully in the scope that does not depart from this invention technological thought.

Therefore, the technical scope of this invention is not limited to the content on the specification, must determine its technical scope according to interest field.

Claims (19)

1, a kind of channel equalization means of VSB receiving system as distorting from the digital broadcast signal compensate for channel that receives by channel, is intended to recover the channel equalization means of original signal, and described channel equalization means comprises:

From the signal that receives by transfer channel, follow the trail of the channel tracking portion of channel impulse response output;

Utilize the channel impulse response that the existing field of above-mentioned channel tracking portion output follows the trail of and the difference of the channel impulse response followed the trail of of field in the past, judge that channel characteristic is a dynamic channel, or the channel characteristic test section of static channel;

According to the channel characteristic that above-mentioned channel characteristic test section detects, control inner filter parameter, compensation is included in the channel equalization portion that the channel in the received signal is distorted;

It is characterized in that above-mentioned channel characteristic test section comprises:

Store above-mentioned channel tracking portion and follow the trail of, and behind the existing field channel impulse response of output, as the memory of former field channel impulse response output;

Draw that above-mentioned channel tracking portion is followed the trail of and the former field channel impulse response of the existing channel impulse response of output and the output of above-mentioned memory poor, and the subtracter of output;

Obtain the absolute value of difference of above-mentioned subtracter output or 2 powers of absolute value, draw the operational part of channel variation amount;

The accumulative total portion of the channel variation amount of above-mentioned operational part output with the interval accumulative total of setting accumulative total;

The comparator that the value of above-mentioned accumulative total portion accumulative total is compared with the critical value of having set.

2, the channel equalization means of VSB receiving system as claimed in claim 1 is characterized in that, above-mentioned channel tracking portion is a field synchronizing signal when at every turn importing, and utilizes the training order that is included in the synchronizing signal of above-mentioned field, follows the trail of channel impulse response.

3, the channel equalization means of VSB receiving system as claimed in claim 1, it is characterized in that, above-mentioned comparator is compared above-mentioned accumulative total portion value totally with first critical value of having set, if aggregate-value just is judged as dynamic channel to channel characteristic greater than first critical value; Otherwise be judged as static channel, and this judged result is exported to equalizer.

4, the channel equalization means of VSB receiving system as claimed in claim 1 is characterized in that, the above-mentioned accumulative total interval of having set is an impulse response length.

5, the channel equalization means of VSB receiving system as claimed in claim 1 is characterized in that, above-mentioned channel characteristic test section also comprises: by a plurality of fields, draw the average of above-mentioned accumulative total portion aggregate-value, and the average portion of output;

Above-mentioned comparator is compared the mean value of above-mentioned average portion output with second critical value of having set, if it is dynamic channel that mean value, is just judged channel characteristic greater than second critical value, otherwise, just be judged as static channel, and judged result is exported to channel equalization portion.

6, the channel equalization means of VSB receiving system as claimed in claim 5 is characterized in that, above-mentioned average portion is suitable for formula $D_{\mathrm{avg}}(t,t-1)=\frac{1}{K}\underset{k=t-K+1}{\overset{t}{\mathrm{\&Sigma;}}}D(k,k-1)$ The moving average mode, draw the mean value D of aggregate-value _Avg(t, t-1).

7, the channel equalization means of VSB receiving system as claimed in claim 5 is characterized in that, above-mentioned average portion is suitable for formula D _Avg(t, t-1)=λ D _Avg(t, t-1)+(1-λ) (t, forgetting factor average mode t-1) draws the mean value D of aggregate-value to D _Avg(t, t-1), here, 0≤λ≤1.

8, the channel equalization means of VSB receiving system as claimed in claim 1 is characterized in that,

Above-mentioned comparator is compared above-mentioned accumulative total portion value totally with the 3rd critical value of having set; Above-mentioned channel characteristic test section also comprises

According to the comparative result of above-mentioned comparator, increase or subtract count value, this count value is compared with the counting critical value of having set, be dynamic channel if count value, is just judged channel characteristic greater than critical value; Otherwise, just be judged to be static channel, and this judged result exported to the reliability counter of above-mentioned channel equalization portion.

9, the channel equalization means of VSB receiving system as claimed in claim 8 is characterized in that, above-mentioned reliability counter is if above-mentioned comparator judges that aggregate-value greater than the 3rd critical value of having set, just increases count value; Otherwise after the minimizing count value, count value is compared with the counting critical value of having set,, just be judged as dynamic channel, otherwise be judged as static channel if count value surpasses the counting critical value.

10, the channel equalization means of VSB receiving system as claimed in claim 1 is characterized in that, above-mentioned channel equalization portion is with the blindness accommodation mode, implements channel equalization, if channel characteristic is judged as dynamic channel, just utilizes the G-Pseudo mode to obtain mistake; If be judged as static channel, just obtain mistake in the Stop-and-Go mode, upgrade filter coefficient.

11, the channel equalization means of VSB receiving system as claimed in claim 1 is characterized in that, above-mentioned channel equalization portion guides accommodation mode to carry out channel equalization to judge, if channel characteristic is judged as dynamic channel, just step value is heightened; If be judged as static channel, just, turn down step-length than above-mentioned dynamic channel, upgrade filter coefficient.

12, a kind of channel equalization method of VSB receiving system as distorting from the digital broadcast signal compensate for channel that receives by channel, is intended to recover the channel equalization method of the channel equalizer in the VSB receiving system of original signal, and described method comprises:

When (a) obtaining field synchronizing signal input, from follow the trail of the stage of channel impulse response by the received signal of transfer channel at every turn;

(b) channel impulse response that the existing field of above-mentioned (a) stage output is followed the trail of is compared with the critical value of having set with the difference of the channel impulse response of field tracking in the past, judges with this whether channel characteristic is the dynamic or static stage;

(c) according to above-mentioned detected channel characteristic of (b) stage, the inner filter parameter of control channel equalizer compensates the stage of the channel distortion that is included in the acknowledge(ment) signal with this;

It is characterized in that above-mentioned (b) stage comprises:

(b1) obtain the poor of existing field channel impulse response that above-mentioned (a) stage follows the trail of and field channel impulse response in the past, and obtain the absolute value of this value or 2 powers of absolute value, the stage of calculating channel variable quantity;

(b2) will be in the accumulative total interval of having set, after the channel variation amount that the above-mentioned stage of accumulative total calculates, the stage of accumulative total;

(b3) value of above-mentioned accumulative total is compared with the critical value of having set,, otherwise be judged as the stage of static channel if aggregate-value greater than critical value, just is judged as dynamic channel with channel characteristic.

13, the channel equalization method of VSB receiving system as claimed in claim 12 is characterized in that, above-mentioned (b3) stage is:

The value of above-mentioned accumulative total is compared with first critical value of having set,, otherwise be judged as the stage of static channel if aggregate-value just is judged as dynamic channel with channel characteristic greater than first critical value.

14, the channel equalization method of VSB receiving system as claimed in claim 12 is characterized in that, above-mentioned (b2) and (b3) also comprise between the stage:

Through a plurality of fields average cumulative value, and the stage of output;

Above-mentioned (b3) stage is, the mean value in above-mentioned stage is compared with second critical value of having set, greater than second critical value, just channel characteristic is judged as dynamic channel as mean value, otherwise is judged as the stage of static channel.

15, the channel equalization method of VSB receiving system as claimed in claim 14 is characterized in that, the above-mentioned average stage is to be suitable for formula $D_{\mathrm{avg}}(t,t-1)=\frac{1}{K}\underset{k=t-K+1}{\overset{t}{\mathrm{\&Sigma;}}}D(k,k-1)$ The moving average mode obtain the mean value D of aggregate-value _Avg(t, t-1).

16, the channel equalization method of VSB receiving system as claimed in claim 14 is characterized in that, the above-mentioned average stage is to be suitable for formula D _Avg(t, t-1)=λ D _Avg(t, t-1)+(1-λ) (t, forgetting factor average mode t-1) draws the mean value D of aggregate-value to D _Avg(t, t-1), here, 0≤λ≤1.

17, the channel equalization method of VSB receiving system as claimed in claim 12 is characterized in that, above-mentioned (b3) stage is:

The stage that aggregate-value is compared with the 3rd critical value of having set;

If the aggregate-value in above-mentioned stage greater than the 3rd critical value, just increases the reliability value, otherwise after reducing the reliability value, to compare with the counting critical value of having set, if the reliability value just is judged as dynamic channel with channel characteristic, otherwise is judged as the stage of static channel greater than the counting critical value.

18, the channel equalization method of VSB receiving system as claimed in claim 12, it is characterized in that, above-mentioned (c) stage is to carry out channel equalization with above-mentioned channel equalizer with the blindness accommodation mode, if channel characteristic is judged as dynamic channel, just utilizes G-Pseudo to obtain mistake; If be judged as static channel, just obtain mistake in the Stop-and-Go mode, upgrade filter coefficient with this.

19, the channel equalization method of VSB receiving system as claimed in claim 12, it is characterized in that above-mentioned (c) stage is to guide the accommodation mode update coefficients with above-mentioned channel equalizer to judge, carries out channel equalization, if channel characteristic is judged as dynamic channel, just heighten step value; If be judged as static channel, the level when just step value being transferred to less than dynamic channel is upgraded filter coefficient with this.

Images