CN1806395A - System of using pilot counteracting equilibrinm - Google Patents

Abstract

An adaptive procedure that optimizes the parameters of a receiver filter such as a Multiuser Detection (MUD) applied to Direct-Sequence Code Division Multiple Access (DS-CDMA) is disclosed. This procedure takes into account the constraints imposed by the absence of training data sequences sent by the transmitter and required to adapt the filter parameters at the receiver. The adaptation consists in using two distinct data sequences transmitted through the same channel; one data sequence is transmitted as payload data and a second data sequence is transmitted as training data used to adapt the filter parameters at the receiver. Parameters of the receiver filter are adapted in presence of varying channels at the same time as the data information sequences are transmitted. The adaptation is realized following a mixed adaptation procedure based on a direct (without channel identification) and indirect (with channel identification) scheme. The invention is described for UMTS (Universal Mobile Telecommunications System) application in cellular communications system.

Description

Utilize the system of pilot counteracting equilibrinm

Technical field

The present invention relates to the digital telecommunication field.Or rather, the present invention relates to directly-indirectly adaptation procedure of a kind of mixing that is applied to the receiver filter.

Background technology

Channel equalization is one of basic problem in the digital telecommunication.Fig. 1 has shown the model of the exemplary communications systems that has merged equilibrium/detection technique.

Different with TDMA (time division multiple access) equalizer, the DS-CDMA equalizer is to remove inter symbol interference (InterSymbol Interference from the data that receive by telecommunication channel, ISI) and multiple access disturb (Multiple Access Interference, MAI).

With reference to Fig. 2 a, model (the M.Latva-aho and M.J.Juntti that wherein shows the DS-CDMA baseband system, " LMMSE Detection for DS-CDMA System in Fading Channel ", IEEETransactions of Communications, Vol.48, No.2,2000, pp194-199 and M.J.Juntti, and M.Latva-aho, " Multiuser Receivers for CDMA Systemsin Rayleigh Fading Channels ", IEEE Transactions on VehicularTechnology, Vol.49, No.3, May 2000, pp.885-899, and D.Massicotte A.0.Dahmaneand, " Nonlinear Multiuser Receiver for UMTSCommunications ", IEEE-Semiannual Vehicular Technology Conference, Vancouver 24-29 September 2002, pp.252-256).In this model, K user sending from alphabet Ξ=1, the code element of 1}.Each user's code element is corresponding by it, length is N _c, be expressed as s _kCode sequence expansion.This code sequence by OVSF code and short scramble code combination are generated (Verd ù S., Multiuser Detection, Cambridge University Press, 1998).Code-element period is represented that by T chip (chip) cycle is by T _cExpression, wherein N _c=T/T _cBe integer.

Suppose that all users use identical chip pulse shaping filter 230, by Ψ (t) expression, it is the square root raised cosine of roll-off factor (roll-off factor) β=0.22 in this case.Each user k passes through by h _k(t) Biao Shi L _kThe Rayleigh in individual path (Rayleigh) attenuation channel 240 transmits its data, and wherein maximum delay spread (delay spread) is τ _mExcept as otherwise noted, baud-spaced index (Baud spaced index) represents that with n the chip-spaced index is represented that by m n the transmission code element of user k is b _k ⁽ⁿ⁾Employed model is the baud-spaced form, but can easily expand to fractional spaced form.

K user's spreading wave form continuous time is $s_k^{\left(n\right)}\left(t\right)=\underset{m=0}{\overset{N_c-1}{\mathrm{\Σ}}}{s}_{k,m}^{\left(n\right)}\mathrm{\ψ}(t-m{T}_c)$ (formula 1).

All users' baseband receiving signals is $\tilde{r}\left(t\right)=\underset{k=1}{\overset{K}{\mathrm{\Σ}}}\left(\underset{n=0}{\overset{N_b-1}{\mathrm{\Σ}}}((A_k{b}_k^{\left(n\right)}{s}_k^{\left(n\right)}(t-\mathrm{nT})+j{\mathrm{\β}}_k{P}_k^{\mathrm{train}\left(n\right)}{p}_k^{\left(n\right)}\left(t\right))*h_k^{\left(n\right)}\left(t\right))\right)+\mathrm{\η}\left(t\right)$ (formula 2), and by adder unit 260 outputs.

In formula 2, N _bThe number of the code element that expression is received, A _kThe amplitude that is received of expression user k, η (t) expression variance is σ _η ²Additive gaussian noise, * represents linear convolution, p _k ⁽ⁿ⁾(t) for have a short scramble code identical with a full business datum with n business (traffic) bit b _k ⁽ⁿ⁾K overlapping user periodic Control (pilot tone) waveform (3GPP-TS 25.213 v4.1.0 (2001-06): Spreading and Modulation (FDD)), shown in Fig. 2 b, and $j=\sqrt{-1}.$

The transmitting channel h of the user k of Rayleigh attenuation channel 240 _k ⁽ⁿ⁾(t) be defined as $h_k^{\left(n\right)}\left(t\right)=\underset{l=1}{\overset{L_k}{\mathrm{\Σ}}}{h}_{k,j}^{\left(n\right)}\mathrm{\δ}(t-{\mathrm{\τ}}_{k,1})$ (formula 3), wherein L _kBe the number of propagation path, h _{K, j} ⁽ⁿ⁾For at the complex gain (complex gain) of time n to the path 1 of user k, τ _{K, 1}Be propagation delay, δ (t) is Di Lake (Dirac) pulse.

The signal that is received can be written as then: $\tilde{r}\left(t\right)=H(b\left(t\right),p\left(t\right))+\mathrm{\η}\left(t\right)$ (formula 4), wherein H () expression channel model 270 is the first of (formula 2).

With discrete form baud-spaced, last formula be $\tilde{r}=H(b,p)+\mathrm{\η}$ (formula 5), wherein $\tilde{r}={\left[\begin{array}{ccc}{\tilde{r}}^{{\left(0\right)}^T},& \·\·\·,& {\tilde{r}}^{{(N_b-1)}^T}\end{array}\right]}^T$ And ${\tilde{r}}^{\left(n\right)}={\left[\begin{array}{ccc}\tilde{r}\left(T_c(n{N}_c+1)\right),& \·\·\·,& \tilde{r}\left(T_c(n+1)N_c\right)\end{array}\right]}^T$ (formula 6), the code element that is sent is $b={\left[\begin{array}{ccc}{b}^{{\left(0\right)}^T},& \·\·\·,& b^{{(N_b-1)}^T}\end{array}\right]}^T$ And $b^{\left(n\right)={\left[\begin{array}{ccc}{b}_1^{\left(n\right)}& ,\·\·\·,& b_K^{\left(n\right)}\end{array}\right]}^T}$ (formula 7), and the control that is sent (pilot tone) is $p={\left[\begin{array}{ccc}{p}^{\left(0\right)T},& \·\·\·,& p^{(N_p-1)T}\end{array}\right]}^T$ And $p^{\left(n\right)}={\left[\begin{array}{ccc}{p}_1^{\left(n\right)}& ,\·\·\·,& p_K^{\left(n\right)}\end{array}\right]}^T$ (formula 8).

Summary of the invention

One object of the present invention is to provide a kind of device that regeneration (regenerated) data sequence is provided.

Another object of the present invention is to provide a kind of playback of data sequence.

Another object of the present invention is to provide a kind of device that the regeneration control sequence is provided.

Another object of the present invention is to provide a kind of device that the playback of data sequence is provided.

According to an aspect of the present invention, provide a kind of device that the playback of data sequence is provided, this device comprises: the channel identification unit, it receives the signal that is sent from communication channel

With training control sequence (p ^Train), so that a plurality of channel coefficients of expression communication channel to be provided And the Channel Modeling unit, it utilizes known training data sequence (X) to filter a plurality of channel coefficients of described expression communication channel So that described playback of data sequence (Y) to be provided.

According to another aspect of the present invention, provide a kind of method that the playback of data sequence is provided, described method comprises: receive the signal that is sent from communication channel

With training control sequence (p ^Train), so that a plurality of channel coefficients of the described communication channel of expression to be provided And utilize known training data sequence (X) to filter a plurality of channel coefficients of described expression communication channel So that described playback of data sequence (Y) to be provided.

Disclose a kind of adaptation procedure, it optimizes the receiver filter parameter, for example is applied to the Multiuser Detection (MUD) of direct-sequence code division multiple access (DS-CDMA).That this process is considered is that reflector sends by lacking, the required training data sequence of adaptive receiver place filter parameter and the restriction of forcing.

This adaptive being: utilize two different data sequences that send by same channel; One of them data sequence sends as effective load data, and another data sequence sends as being used for the training data of adaptive receiver place filter parameter.There is adaptive receiver filter parameter under the situation that changes channel simultaneously with the transmission data information sequence.By based on direct (not having channel identification) the adaptation procedure of mixing, realize described adaptive with indirect (channel identification is arranged) scheme.Use at the UMTS in the cellular communications system (Universal Mobile Telecommunications System), described the present invention.

Description of drawings

Other features of the present invention and advantage become clear by following detailed description with reference to accompanying drawing, wherein:

Fig. 1 is for showing the block diagram of exemplary communications systems model;

Fig. 2 a is the block diagram of the prior art baseband model of demonstration DS-CDMA system;

Fig. 2 b is the data on flows in other data of demonstration and the block diagram of pilot tone;

Fig. 3 is for showing that it comprises channel identification unit and Channel Modeling unit according to the block diagram of the data sequence generator unit of an embodiment of the present invention;

Fig. 4 for show according to an embodiment of the present invention, for the DS-CDMA system, that do not have pilot tone, the direct block diagram of adaptation method filter receiver architecture;

Fig. 5 for show according to an embodiment of the present invention, for the DS-CDMA system, that do not have pilot tone, the block diagram of adaptation method filter receiver architecture indirectly;

Fig. 6 for show according to an embodiment of the present invention, for the block diagram DS-CDMA system, mixed method cascade (cascade) filter receiver architecture;

Fig. 7 for show according to an embodiment of the present invention, for the block diagram DS-CDMA system, that do not have pilot tone, mixed method cascading filter receiver architecture;

Fig. 8 is for showing according to block diagram an embodiment of the present invention, receiver place training data sequence generator device.

Embodiment

People have developed various algorithms, to solve following problem: according to

Obtain original series b _k ⁽ⁿ⁾Estimation send data  _k ⁽ⁿ⁾Most of algorithm can be reduced to the digital filtering of the sequence of symhols of being hindered and damaged by inter symbol interference and multiple access, ${\hat{b}}_k^{\left(n\right)}=F\left[\begin{array}{c}{\tilde{r}}^{\left(n\right)}\end{array}\right]$ (formula 9), wherein F[] the MUD filter 250 of presentation graphs 2a.

Still with reference to Fig. 2 a,, must consider N in order to have discrete linear model _f Dimension MUD filter 250, the output of channel model will be applied to this MUD filter 250.Introduce vector ${\tilde{r}}^{\left(n\right)}={\left[\begin{array}{cccc}{\tilde{r}}^{\left(m\right)},& {\tilde{r}}^{(m-1)},& ...,& {\tilde{r}}^{(m-N_f+1)}\end{array}\right]}^T$ (formula 10).

For first receiver that IS-95 proposed is the Rake receiver, to utilize the attenuation characteristic of channel.Yet, nearly far away problem (near-far problem) make this receiver efficient not high (Nolma H., etToskala A., WCDMA for UMTS:Radio Access For Third Generation Mobile Communications, John Wiley ﹠amp; Sons LTD, 2000).

People have proposed linear and Nonlinear M UD filter 250 (F[]) and have overcome this near-far problem.It is at first main that what consider is that application extension with the TDMA equalizer is to DS-CDMA.This algorithm be for Sequence Detection the PRML sequencal estimation (Maximum Likelihood Sequence Estimation, MLSE) and for by the maximum a posteriori probability of symbol detection (Maxium a-posteriori, MAP).Unfortunately, these algorithms and impracticable, this be because complexity with number of users be exponential increase (Verd ù S., Multiuser Detection, Cambridge University Press, 1998).Extensively other algorithms that propose are ZF (Zero Forcing, zero forces) and MMSE (Minimum Mean Square Error), it needs accurate impulse response (the Klein S. of all subscriber channels, Kaleh G.K., et Baier P.W., " Zero Forcing and MinimumMean-Square-Error Equalization for Multiuser Detection in Code-DivisionMultiple-Access Channels " IEEE Transactions on Vehicular Technology, Vol.45, No.2, Mai 1996, pp.276-287).This is unpractical.Can not in using, adopt real life even the self adaptation version of lms algorithm is also too complicated.PIC (ParallelInterference Cancellation, parallel Interference Cancellation) algorithm and SIC (SuccessiveInterference Cancellation, Interference Cancellation continuously) algorithm was before also disclosed.The optimization version of these two kinds of receivers need be known the amplitude that does not have multiple access user that disturb, that all receive.This is not easy to obtain, and disturbs because difficulty just is to remove multiple access.

People have proposed other receivers based on linear filter and neural net, but do not obtain a kind of all users' of equilibrium universal architecture (Das K., et Morgera S.D., " Adaptive InterferenceCancellation for DS-CDMA Systems Using Neural Network Techniques " IEEE Journal on Selected Areas in Communications, Vol.16, No.9,1998, pp.1774-1784).

If the independent variable of the function F [] of formula 9 is linear, think that then the MUD filter is linear, under other all situations, all is used as it as Nonlinear Processing.For example, the MUD filter based on neural net is considered to nonlinear.

If in order to estimate that its coefficient is necessary to send the known array that is called the training sequence maker and (is respectively b for sending data and control (pilot tone) ^TrainWith p ^Train), then the MUD filter has supervision.

For time varying channel, this causes the loss of available bandwidth, and the adaptive technology that can use at judgement, and it is corresponding to first approach of blind (Blind) method.Different with the MUD filter that supervision is arranged, estimate its coefficient under the situation of the training sequence that blind MUD filter sends not knowing, improved bandwidth efficiency thus.

If with received signal

Directly send to the MUD filter input end, then the MUD filter has control assembly, for example shown in Fig. 4,6,7.

Otherwise, if from received signal

Middle removal is controlled (pilot tone) signal and is applied to after the MUD filter input end, and then not control (pilot tone) of MUD filter is for example shown in Fig. 4,5.

MUD filter 250 is based on two kinds of known methods.

If its coefficient obtains from data available, then the MUD filter is directly, MUD filter 251 for example shown in Figure 6.

If by utilizing such as channel identification unit 12 such as correlators according to H[] parameter (tap (tap) of 270 previous identification With delay

And try to achieve its coefficient, then the MUD filter is indirect, MUD filter 252 for example shown in Figure 5 (Bhashyam, S., Aazhang, B., " Multiuser channelestimation and tracking for long-code CDMA systems ", IEEE Transactions on Communications, Volume:50, Issue:7, July 2002, pp.1081-1090).

In practice, if channel model is linear, then generally use indirect method, this is because therefore channel identification unit 12 can be easy to realize.

As described below, can use two or the more a plurality of different data sequence that send by same channel.

The first data sequence b sends as the effective load data sequence.

The second data sequence p ^TrainUtilize training control (pilot tone) sequence of channel identification method identification channel parameter to send at the receiver place as being used for.

Referring now to Fig. 3, wherein shown example according to the data sequence maker 10 of an embodiment of the present invention.

Data sequence maker 10 comprises channel identification unit 12 and Channel Modeling unit 14.

Channel identification unit 12 receives the signal that is sent

With training control sequence p ^Train, carry out channel identification, so that a plurality of channel coefficients of expression communication channel to be provided

Channel Modeling unit 14 receives a plurality of channel coefficients of this expression communication channel

And known training data sequence (X), and filter a plurality of channel coefficients of expression communication channel with known training data sequence (X)

So that playback of data sequence (Y) to be provided.

Be appreciated that in one embodiment known training data sequence (X) can be training control sequence p ^Train(referring to Fig. 4).In this case, Channel Modeling unit 14 comprises channel control modeling unit 410, and playback of data sequence (Y) comprises regeneration control sequence r ^PolitBelow will further describe the example that merges this embodiment.

In another embodiment, known training data sequence (X) can be training control sequence b ^Train(referring to Fig. 6).In this case, Channel Modeling unit 14 comprises channel data modeling unit 610, and playback of data sequence (Y) comprises regeneration training sequence r ^TrainBelow will further describe the example that merges this embodiment.

Referring now to Fig. 8, shown that wherein data sequence maker 10 (DSGA) is positioned at the example at receiver 80 places.

By channel H[] 270 transmission training control (pilot tone) sequence p ^Train, to carry out for training sequence by the channel identification unit 12 of K user-defined all channels.

Training control (pilot tone) sequence p that receiver is known ^TrainBe used for identification channel model H[] parameter.

Carry out channel identification method, for example correlator, PRML or the like, and/or following adaptation algorithm, for example LMS, RLS, Kalman filter, reverse transmittance nerve network or the like by channel identification unit 12.

A plurality of channel coefficients of expression communication channel have been discerned when channel identification unit 12 The time, send a plurality of channel coefficients of representing communication channel to Channel Modeling unit 14

Generate one group of training data sequence (X) at the receiver place.Use this group training data sequence (X), generate playback of data sequence (Y) to utilize Channel Modeling unit 14.Use this group training data sequence (X) at the receiver place, with adaptive receiver filter.

At Y=r ^DataAnd X=b ^TrainSituation under, Channel Modeling unit 10 can by $r^{\mathrm{Data}}\left(t\right)=\underset{k=1}{\overset{K}{\mathrm{\Σ}}}\left(\underset{n=0}{\overset{N_b-1}{\mathrm{\Σ}}}(\left({b^{\mathrm{train}}}_k^{\left(n\right)}{s}_k^{\left(n\right)}(t-\mathrm{nT})\right)*{\hat{h}}_k^{\left(n\right)}\left(t\right))\right)$ (formula 11) definition.

Perhaps, at X=b ^TrainAnd Y=r ^PilotSituation under, Channel Modeling unit 10 can by $r^{\mathrm{pilot}}\left(t\right)=\underset{k=1}{\overset{K}{\mathrm{\Σ}}}\left(\underset{n=0}{\overset{N_p-1}{\mathrm{\Σ}}}(\left({{{\mathrm{jP}}^{\mathrm{train}}}_k^{\left(n\right)}P}_k^{\left(n\right)}(t-\mathrm{nT})\right)*{\hat{h}}_k^{\left(n\right)}\left(t\right))\right)$ (formula 12) definition.

Under the situation of MUD filter adaptation receiver, the MUD filter designs with hybrid mode, wherein obtains coefficient according to indirect and direct method, shown in Fig. 5,6.

Referring now to Fig. 4, wherein show receiver according to the direct adaptation method operation that utilizes control (pilot tone) to offset.

Or rather, receiver comprises Channel Modeling unit 10, and it has channel identification unit 12 and channel control modeling unit 410.

This receiver also comprises control signal offset unit 420, direct MUD filter 251 and K switch ^1-2

By using directly and handling both indirectly, carry out the adaptive of direct MUD filter simultaneously and offset with control (pilot tone).

According to the first step, the parameter of the direct MUD filter 251 of initialization and the parameter of channel identification unit 12.

According to second step, send training control sequence p by channel ^Train, to obtain for training sequence by the channel identification unit 12 of K user-defined all channels.Simultaneously, send training data sequence b ^TrainAnd/or the effective load data sequence b that does not show, comprise the information that sends by all users.Training data sequence p ^TrainFor receiver is known, and training data sequence p ^TrainBe used for the parameter of identification channel model H ().Channel identification unit 12 uses the channel identification algorithms, for example correlator, PRML or the like, and/or following adaptation algorithm, and for example LMS, RLS, Kalman filter, reverse transmittance nerve network or the like are to determine a plurality of channel coefficients of expression communication channel

According to the 3rd step, when a plurality of channel coefficients of having discerned the expression communication channel

The time, utilize channel control modeling unit 410 and training control sequence p ^Train, the regeneration training sequence r at generation receiver place ^Train

According to the 4th step, the effect of pilot data is disturbed training data or effective load data, and must offset.This counteracting is finished by following: utilize control signal offset unit 420, with the data that received

Deduct regeneration control sequence r ^PilotControl signal offset unit 420 provides new reception data

With second while in step, K switch ^1-2Be positioned at position A, to send in order to carry out to the required training data sequence b of training sequence by K user-defined all channels ^Train

According to the 6th step, use the known training data sequence b of receiver ^TrainAnd data

With according to LMS, RLS, reverse transmittance nerve network or the like adaptation algorithm, the parameter of adaptive direct MUD filter 251.

According to the 7th step, when the adaptive parameter of direct MUD filter 251, K switch ^1-2Be positioned at position B.By direct MUD filter 251, utilize data Estimate to comprise the effective load data b of information.On the B position, do not send training data, and the parameter constant of direct MUD filter 251.

With the 7th while in step, utilized for second step followed the tracks of channel parameter, and go on foot by using the 3rd and the 4th, calculate the reception data that do not have pilot tone.

Periodically, K switch ^1-2Alternate between position A and B is with the parameter of adaptive direct MUD filter 251 under various channel H () condition.

Referring now to Fig. 5, wherein shown receiver according to the indirect adaptation method operation that comprises control (pilot tone) counteracting.By following steps this pilot counteracting is described.

According to the first step, the parameter of Initial Channel Assignment recognition unit 12.

According to second step, send training control sequence p ^Train, to carry out for training sequence by the channel identification of K user-defined all channels.Simultaneously, send the effective load data sequence b that does not show by all users' transmissions.Training control sequence p ^TrainFor receiver is known, and it is used for the parameter of identification channel model H ().Utilize to use the channel identification unit 12 of following method, for example correlator, PRML or the like, and/or following adaptation algorithm, for example LMS, RLS, Kalman filter, reverse transmittance nerve network or the like are to determine a plurality of channel coefficients of expression communication channel

According to the 3rd step, when a plurality of channel coefficients of having discerned the expression communication channel The time, utilize channel control modeling unit 410 and training control sequence p ^Train, the pilot frequency sequence r at generation receiver place ^Pilot

According to the 4th step, do not having under the condition of pilot tone, the effect of control (pilot tone) data is disturbed effective load data, and therefore must offset.This counteracting is finished by following: utilize control signal offset unit 420, with the data that received Deduct regeneration control sequence r ^PilotControl signal offset unit 420 provides

With the 3rd step and the 4th while in step, and when a plurality of channel coefficients of having discerned the expression communication channel

The time, calculate the parameter of indirect MUD filter 252.

When having calculated the parameter of indirect MUD filter 252, use data by indirect MUD filter 252 Estimate to comprise the effective load data b of information.

With the final step while, utilized for second step followed the tracks of a plurality of channel coefficients of expression communication channel

And, calculate the reception data that do not have pilot tone by using for the 3rd and the 4th step.

Under the time varying channel condition, utilize second step, the 3rd step, follow the tracks of a plurality of channel coefficients of expression communication channel with final step

Referring now to Fig. 6, wherein shown following receiver, it, and generates and is used for the training sequence (not needing to send training data) of data parameter, the receiver place of adaptive direct MUD filter 251 according to mixing the adaptation method operation based on channel identification unit 12.

According to the first step, the parameter of the direct MUD filter 251 of initialization and the parameter of channel identification unit 12.

According to second step, send training control sequence p by channel ^Train, to carry out for channel identification unit 12 training sequences by K user-defined all channels.Simultaneously, send the effective load data sequence b that comprises the information that sends by all users.Training data sequence p ^TrainFor receiver is known, and in order to the parameter of identification channel model H ().Use the channel identification method by channel identification unit 12, for example correlator, PRML or the like, and/or following adaptation algorithm, for example LMS, RLS, Kalman filter, reverse transmittance nerve network or the like are with parameter (a plurality of channel coefficients of expression communication channel of identification channel model H ()

According to the 3rd step, when having discerned the parameter of channel model H (), generate one group of training data sequence b at the receiver place ^TrainWith training control sequence p ^TrainThis group training data sequence b ^TrainBe used for utilizing channel data modeling unit 610 to generate receiving data sequence r ^Data, and training control sequence p ^TrainBe used for utilizing channel control modeling unit 410 to generate regeneration control sequence r ^Pilot

According to the 4th step, utilize adder unit 620, r ^DataWith r ^PilotAdd and produce received signal r ^Train

According to the 5th step, K switch ^1-1With K ^1-2All be positioned at position A, to carry out to training sequence by K user-defined all channels.

According to the 6th step, use the one group of training data sequence b that is generated ^TrainAnd received signal r ^Train, with according to LMS, RLS, reverse transmittance nerve network or the like adaptation algorithm for example, the parameter of adaptive direct MUD filter 251.

According to the 7th step, when the parameter of adaptive direct MUD filter 251, K switch ^1-1With K ^1-2All be placed in position B.By direct MUD filter 251, utilize data Estimate to comprise the effective load data b of information.

Periodically, K switch ^1-1With K ^1-2Alternate between position A and B, with repeating step 2 to 7, thereby under various channel H () condition the parameter of adaptive direct MUD filter 251.

Referring now to Fig. 7, wherein shown following receiver, it is adding under the condition of not having the signal of controlling (pilot tone) based on operating according to mixing adaptation method with the described identical scheme of Fig. 6.

According to the first step, the parameter of the direct MUD filter 251 of initialization and the parameter of channel identification unit 12.

According to second step, send training control sequence p by channel ^Train, to obtain for training sequence by the channel identification unit 12 of K user-defined all channels.Simultaneously, send the effective load data sequence b that does not show, comprise the information that sends by K user.Training control sequence p ^TrainFor receiver is known, and be used for the parameter of identification channel model H ().Channel identification unit 12 can use the channel identification method, for example correlator, PRML or the like, and/or following adaptation algorithm, for example LMS, RLS, Kalman filter, reverse transmittance nerve network or the like.

According to the 3rd step, when having discerned all parameters of channel data modeling unit 610 and channel control modeling unit 410, generate one group of training data sequence b at the receiver place ^TrainWith training control sequence p ^TrainThis group training data sequence b ^TrainBe used for utilizing channel data modeling unit 610 to generate receiving data sequence r ^Data, and training control sequence p ^TrainBe used for utilizing channel control modeling unit 410 to generate regeneration control sequence r ^Pilot

According to the 4th step, K switch ^1-1With K ^1-2All be positioned at position A, to carry out to training sequence by K user-defined all channels.

According to the 5th step, use the one group of training data sequence b that is generated ^TrainAnd control sequence r ^Train, with according to LMS, RLS, reverse transmittance nerve network or the like adaptation algorithm, the parameter of adaptive direct MUD filter 251.

According to the 3rd the 6th step of carrying out simultaneously in step, channel data modeling unit 610 produces one group of data sequence r ^Pilot

According to the 7th step, under the condition of not having control (pilot tone), the effect of control (pilot tone) data is disturbed effective load data, and therefore must offset.This counteracting is finished by following: utilize control signal offset unit 420, with the data that received Deduct r ^PilotControl signal offset unit 420 provides new reception data

When the adaptive parameter of direct MUD filter 251, and according to the 8th step, K switch ^1-1With K ^1-2All be positioned at position B.By direct MUD filter 251, utilize data Estimate to comprise the effective load data b of information.

According to the 7th the 9th step of carrying out simultaneously in step, utilize the parameter that second step followed the tracks of channel identification unit 12, and use does not have the reception data of pilot tone in the 6th and seven steps.

According to the tenth step, K switch ^1-1With K ^1-2Alternate between position A and B, with repeating step 2 to 9, thereby under various channel H () condition the parameter of adaptive direct MUD filter 251.

In addition, above-mentioned blind adaptation procedure can be used for these adaptation methods, to increase bandwidth efficiency.

Though preferred embodiment is so that the block diagram of the discrete component group of intercommunication shows as connect mutually by the different pieces of information signal, but those skilled in the art should understand that preferred embodiment combines with nextport hardware component NextPort by software provides, wherein some assembly is realized by the given function or the operation of hardware or software systems, and shown many data paths are realized by the data communication in computer applied algorithm or the operating system.Thus, structure shown in provides for ease of explanation the preferred embodiments of the present invention.

Be noted that the present invention can be implemented as method, can be in system, in computer-readable medium or electric or electromagnetic signal, realize.

One or more embodiment of the invention described above are just in order to illustrate.Therefore, scope of the present invention only is defined by the claims.

Claims (12)

1. device that the playback of data sequence is provided, described device comprises:

The channel identification unit, it receives the signal that is sent from communication channel With training control sequence (p ^Train), so that a plurality of channel coefficients of the described communication channel of expression to be provided And

The Channel Modeling unit, it utilizes known training data sequence (X) to filter a plurality of channel coefficients of the described communication channel of described expression So that described playback of data sequence (Y) to be provided.

2. device as claimed in claim 1, wherein said training sequence (X) comprise described training control sequence (p ^Train), and wherein said playback of data sequence (Y) comprises regeneration control sequence (r ^Polit), and wherein said Channel Modeling unit comprises channel control modeling unit, and it utilizes described training control sequence (p ^Train), filter a plurality of channel coefficients of the described communication channel of described expression

So that described regeneration control sequence (r to be provided ^Polit).

3. device as claimed in claim 2 also comprises the control signal offset unit, its signal from being sent

In deduct described regeneration control sequence (r ^Polit), so that the control sequence of no described control sequence to be provided

4. device as claimed in claim 1, wherein said training sequence (X) comprises training data sequence (b ^Train), and wherein said playback of data sequence (Y) comprises regeneration training sequence (r ^Train), and wherein said Channel Modeling unit comprises the channel data modeling unit, and it utilizes described training data sequence (b ^Train), filter a plurality of channel coefficients of the described communication channel of described expression So that described regeneration training sequence (r to be provided ^Train).

5. device as claimed in claim 4, wherein said Channel Modeling unit also comprise channel control modeling unit, and it utilizes described training control sequence (p ^Train), filter a plurality of channel coefficients of the described communication channel of described expression

So that regeneration control sequence (r to be provided ^Pilot).

6. one kind provides the direct adaptive receiver of estimating effective load data sequence (), and described receiver comprises:

Generate the device of the playback of data sequence of no control sequence, this device comprises:

The channel identification unit, it receives the signal that is sent from communication channel

With training control sequence (p ^Train), so that a plurality of channel coefficients of the described communication channel of expression to be provided

The Channel Modeling unit, it utilizes described training control sequence (p ^Train) filter a plurality of channel coefficients of the described communication channel of described expression So that regeneration control sequence (r to be provided ^Polit);

The control signal offset unit, it is from the described signal that sends In deduct described regeneration control sequence (r ^Polit), so that the control sequence of no described control sequence to be provided

And

Filter unit, it receives the playback of data sequence of described no control sequence, and optionally receives training data sequence (b ^Train), so that described estimation effective load data sequence to be provided

And

Wherein said filter unit is according to described training data sequence (b ^Train) by adaptive.

7. method that the playback of data sequence is provided, described method comprises:

Receive the signal that is sent from communication channel With training control sequence (p ^Train), so that a plurality of channel coefficients of the described communication channel of expression to be provided

And

Utilize known training data sequence (X) to filter a plurality of channel coefficients of the described communication channel of described expression So that described playback of data sequence (Y) to be provided.

8. method as claimed in claim 7, wherein said training sequence (X) comprise described training control sequence (p ^Train), and wherein said playback of data sequence (Y) comprises regeneration control sequence (r ^Polit), and described method also comprises: utilize described training control sequence (p ^Train), filter a plurality of channel coefficients of the described communication channel of described expression So that described regeneration control sequence (r to be provided ^Polit).

9. method as claimed in claim 8 also comprises: from the signal that is sent In deduct described regeneration control sequence (r ^Polit), so that the control sequence of no described control sequence to be provided

10. method as claimed in claim 7, wherein said training sequence (X) comprises training data sequence (b ^Train), and wherein said playback of data sequence (Y) comprises regeneration training sequence (r ^Train), and wherein said method also comprises: utilize described training data sequence (b ^Train), filter a plurality of channel coefficients of the described communication channel of described expression So that described regeneration training sequence (r to be provided ^Train).

11. method as claimed in claim 10 also comprises: utilize described training control sequence (p ^Train), filter a plurality of channel coefficients of the described communication channel of described expression

So that regeneration control sequence (rpliot) to be provided.

12. an adaptive approach of optimizing receiver place filter parameter, described method comprises:

First and second data sequence that utilization sends by same communication channel, wherein said first data sequence comprises effective load data, and described second data sequence comprises training data;

Utilize described training data, with adaptive receiver place filter parameter;

Wherein with the described data sequence while adaptive described filter parameter under the situation of the variation channel that exists the receiver place to receive of transmission.

Images