WO1995012926A1 - Procede de demodulation adaptative produisant des repliques et demodulateur l'utilisant - Google Patents
Procede de demodulation adaptative produisant des repliques et demodulateur l'utilisant Download PDFInfo
- Publication number
- WO1995012926A1 WO1995012926A1 PCT/JP1994/001862 JP9401862W WO9512926A1 WO 1995012926 A1 WO1995012926 A1 WO 1995012926A1 JP 9401862 W JP9401862 W JP 9401862W WO 9512926 A1 WO9512926 A1 WO 9512926A1
- Authority
- WO
- WIPO (PCT)
- Prior art keywords
- vector
- candidate
- matrix
- input signal
- sequence
- Prior art date
Links
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04B—TRANSMISSION
- H04B3/00—Line transmission systems
- H04B3/02—Details
- H04B3/20—Reducing echo effects or singing; Opening or closing transmitting path; Conditioning for transmission in one direction or the other
- H04B3/23—Reducing echo effects or singing; Opening or closing transmitting path; Conditioning for transmission in one direction or the other using a replica of transmitted signal in the time domain, e.g. echo cancellers
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L25/00—Baseband systems
- H04L25/02—Details ; arrangements for supplying electrical power along data transmission lines
- H04L25/03—Shaping networks in transmitter or receiver, e.g. adaptive shaping networks
- H04L25/03006—Arrangements for removing intersymbol interference
- H04L25/03012—Arrangements for removing intersymbol interference operating in the time domain
- H04L25/03019—Arrangements for removing intersymbol interference operating in the time domain adaptive, i.e. capable of adjustment during data reception
- H04L25/03031—Arrangements for removing intersymbol interference operating in the time domain adaptive, i.e. capable of adjustment during data reception using only passive components
Definitions
- the present invention relates to a demodulation method for demodulating an input signal by using adaptively generated Lebric force with respect to a transmission signal from a transmission line whose transmission characteristics fluctuate, and a demodulator using the same.
- transmission characteristics for example, an impulse response from a transmitter to a receiver may fluctuate greatly with time.
- Microwave wireless transmission and mobile communication are typical examples.
- the desired signal is received with time-varying waveform distortion, and further fluctuating co-channel interference and adjacent channel interference are superimposed and transmitted. Characteristics may deteriorate significantly. Therefore, it is necessary to realize a highly reliable receiving system while suppressing deterioration due to these causes.
- An adaptive receiver using an adaptive algorithm has been used to receive a transmission signal from a transmission line whose transmission characteristics fluctuate in this way.
- an adaptive equalizer has been used as an adaptive receiver in a transmission path where a fluctuating distortion occurs.
- This adaptive equalizer is classified into a linear equalizer and a non-linear equalizer based on its structure.
- Figure 1 shows the configuration of a conventional linear equalizer.
- the real component of the input signal x (i) to the equalizer at discrete time i at the discrete time of one symbol interval represents the amplitude of the in-phase component of the received signal
- the imaginary component represents the amplitude of the quadrature component of the received signal.
- the input signal x (i) is input from the input terminal 11 to the transversal filter 12 having M taps, and the distortion is controlled by controlling the tap coefficients w, (i), "', and w M (i). Is removed and input to the determiner 13.
- the determination signal d (i) determined by the determiner 13 is output from the output terminal 14. Both the input and output signals of the determiner 13 Is input to the error calculation unit 15 to calculate an error signal e (i). This error signal is sent to the control unit 16. The control unit 16 generates the error signal e (i) and the input signal x ( From (i), the tap coefficient of the transversal filter 12 is updated. This is described in detail in G. Proakis, Digital Communications, 2nd edition, McGraw-Hill, 1989.
- a column vector in which M tap coefficients w, (i),, w M (i) given to the transversal filter 12 with M taps are denoted by a coefficient vector W (i ), And a sequence of M input signals x (i), '", x (i-M + l) from time i to M—one time before corresponding to each tap position.
- the vector be the input vector Z (i), where the input signal x (i), which is an element of the vector Z (i), The interfering signal and noise are superimposed and fluctuate every moment on the wireless transmission path
- the coefficient vector W (il) is the input vector Z (i) and the error signal e It is sequentially updated to W (i) using (i).
- the output s (i) of the transversal filter 12 is expressed by the following linear equation.
- H represents the complex conjugate transpose.
- XH (i) (Z (i), Z (i-1), ⁇ , Z (i-K + 1)
- SH (i) (s (i), ⁇ , s (i- K + l)), where K is an integer of 1 or greater.
- Equation (03) X H (i) X (i) represents the autocorrelation matrix of the input signal, and X H (i) D (i) represents the cross-correlation vector between the input signal and the decision signal. Represents. Expressing them as R (i) and V (i), respectively, equation (03) becomes
- R (i) W (i) V (i) (04)
- W (i) in the system of Fig. 1 is a least squares solution.
- the autocorrelation matrix R (i) obtained from the input vector Z (i) consisting of M elements is an M-dimensional square matrix, and the input vector Z (i) and the decision output signal vector D Ci)
- the cross-correlation vector V (i) with is an M-dimensional vector.
- R (i) is said to be regular if the inverse of R (i) exists.
- the inverse matrix R one 1 of the autocorrelation matrix R (i) (i) had always nonexistent say limiting. Therefore, when a received wave whose autocorrelation matrix rank is smaller than M, for example, when a modulated wave with the same code continues under low noise conditions, the elements of R (i) have the same value.
- Adaptive algorithms are also known for sequentially solving the equations. For example, carman filters, RLS, and LMS are typical examples. For more information on adaptive algorithms, see Haykin, "Adaptive Filter Theory", Prentice-Hall, 1991. Even in these methods, if the autocorrelation matrix is not regular, that is, if it is singular, the solution diverges.
- Figure 2 shows an example of a demodulator configured as a nonlinear equalizer.
- the input signal x (i) is input from the input terminal 11 to the difference circuit 1 ⁇ , and the difference signal is calculated from the replied force signal y m (i) output from the transversal filter 18 to obtain the error signal e.
- the square of the absolute value of the error signal e citi(i) is calculated by the squaring circuit 19, and the calculation result is input to the maximum likelihood sequence estimation circuit 21 and the signal Is estimated.
- a code sequence candidate ⁇ a hail ⁇ is output from the maximum likelihood sequence estimation circuit 21.
- the modulation circuit 22 performs the same modulation as that performed on the transmission side by the code sequence candidate ⁇
- the modulation circuit 22 outputs a modulated wave candidate s administrat(for example, a complex symbol candidate). Is the modulated wave candidate s n are input to the transformer parser Ruch I filter 1 8 of power strips number M, the desired signal replica signal y ra (i) is generated.
- the state at the time point i is defined by the immediately preceding sequence of M—one complex symbol candidate s (i ⁇ l), s (i ⁇ 2), to s (i ⁇ M + l). Takes a transition state according to the code sequence candidate ⁇ aflower ⁇ .
- the candidates ⁇ aggi ⁇ output from the maximum likelihood sequence estimation circuit 21 are Since there are only the number of state transitions for each state, the above-described calculation is performed for all candidates corresponding to the product of the number of states and the number of state transitions for the same input signal x (i).
- the maximum likelihood sequence estimating unit 21 selects the maximum likelihood among a plurality of state transitions merged from the previous time to the current time based on the error signal corresponding to each candidate.
- the control circuit 2 along the state transition of the maximum likelihood for each state of the present time point, with the corresponding error signal e m (i) modulated wave candidate s m, bets La Nsubasarufu I filter 1 8 Update the tap coefficient.
- This nonlinear equalizer is described in, for example, K. Fukavm, H. Suzuki, "Adaptive equalization with R and S—M and SE for frequency-selective fast fading mobile radio channels, 'IEEE Globecom' 91, PP. 1-16.6.5, Dec. 1991, and H. Yoshino, K. Fukawa, H. Suzuki, "Adaptive equalization with RLS-MLSE for fast fading mobile radio channels," IEEE Inter. Symp. Circuit and Sys. , PP.501-504, San Diego, May, 1992.
- the nonlinear equalizer shown in Fig. 2 uses a transversal filter 18 to approximate the propagation characteristics of the transmission path. Characteristics, that is, the Tub coefficient Controlling.
- the number of taps of the transversal filter 18 is M, and the tap coefficient w perhaps., (I),-, w m of the transversal filter 18 for the modulated wave candidate s prepare (i). and M (i) coefficient base-vector ⁇ column base click preparative le formed by arranging) modulated wave candidate s m (modulated wave from i) to M- 1 time before a candidate s m (i-M + l) is the modulated wave candidate vector S m (i), the coefficient vector W i-1) is the error of the modulated wave candidate vector S m (i). It is updated to Wêt(i) using the signal estein (i).
- the rank of the autocorrelation matrix of the modulated wave candidate vector S fur(i) is smaller than M, and the coefficient vector Wn> (i) is diverged.
- a blind Viterbi equalization method is known as a type of non-linear equalizer that does not require a training signal.
- the operation of this equalizer is described in Y. Furuya, A. Ushiro kawa, H. Isa, Y. Sato, "A study of blind Viterbi equalization algorithm," 1991 Spring Nat. Conv. of IEICE, A-141, March 1991.
- Even in such an equalizer there is a drawback that the coefficient vector diverges in a modulated wave candidate having the same code as described above. For this reason, in this equalization, a method is used in which the coefficient vector is not updated when a modulated wave candidate having the same code continues.
- an interference canceller or a diversity configuration of an equalizer and a canceller.
- These are also well known for their linear and ⁇ linear constructs.
- An object of the present invention is to generate a replica from an input signal and diverge the coefficient vector even if a modulated wave candidate having the same code is given to a trans-spar filter.
- An object of the present invention is to provide an adaptive demodulation method that does not cause the error to gradually converge to the original solution and a demodulator using the same.
- An adaptive demodulator includes: a maximum likelihood sequence estimating unit for estimating a maximum likelihood sequence based on an error signal sequence; and a coefficient vector corresponding to each state of the maximum likelihood sequence estimating unit. And a modulating wave candidate vector corresponding to the state transition in each state to generate a repli- cation force by calculating the inner product, and a replica vector having a series of the generated repli- cation force as an element.
- the error calculation means for generating an error vector obtained by calculating the difference between the input signal vector and the input signal vector whose elements are the input signal series, and the maximum likelihood sequence estimation means select using the error vector. Updating means for updating the coefficient vector using the generalized inverse matrix generated from the modulated wave candidate corresponding to the state transition.
- the adaptive demodulation method includes the following steps: (1) From the maximum likelihood sequence estimating means, outputs a signal sequence candidate corresponding to a state transition for each state to the Levi force generating means, and (2 :) Lebris
- a modulated wave candidate vector is generated from the signal sequence candidate corresponding to the state transition, and a replica is generated by calculating an inner product with the coefficient vector corresponding to each state of the maximum likelihood sequence estimation means.
- the error calculation means generates a replica vector having a sequence of the generated replica as an element and an input signal vector having a sequence of the input signal as an element.
- An error vector is generated by a vector difference operation, and (4) the maximum likelihood sequence estimation means selects a state transition candidate with a high likelihood from an error signal sequence corresponding to the state transition for each state, and Judgment signal based on high degree status (5)
- the maximum likelihood sequence estimating means selects the inner product vector between the generalized inverse matrix generated from the modulated wave candidate corresponding to the state transition selected using the error vector and the error vector. Update the coefficient vector for each state in Torr.
- the adaptive demodulator and demodulation method according to the present invention are different from the conventional maximum likelihood sequence estimation demodulator and demodulation method in the following points.
- the conventional error calculating means performs the difference calculation of the scalar amount, and the error is also a scalar amount.In the present invention, however, the difference calculation of the vector amount is performed, and the error is also the vector amount. . However, when the number of elements is 1, it is equivalent to the conventional error calculation means.
- the conventional maximum likelihood sequence estimation means performs state estimation using the error of the scalar quantity.
- state estimation is performed using the error of the vector amount.
- the number of elements is 1, it is equivalent to the conventional state estimation.
- Conventional updating means finds the coefficient vector for each state as the minimum norm solution of a normal equation or a solution that converges with a successive updating formula such as Kalman filter, RLS, or LMS.
- the coefficient vector for each state is sequentially updated using a general inverse matrix, which is essentially different from the conventional updating means.
- FIG. 1 is a block diagram of a demodulator configured as a conventional linear equalizer.
- Figure 2 is a block diagram of a demodulator configured as a conventional nonlinear equalizer.
- FIG. 3 is a block diagram showing a configuration example of an adaptive demodulator according to the present invention when maximum likelihood sequence estimation is used.
- Fig. 4 is a diagram showing an example of a trellis in the maximum likelihood sequence estimator.
- Figure 5 is a diagram for explaining the updating of coefficient vectors.
- FIG. 6 is a block diagram showing the configuration of an adaptive demodulator according to the present invention when using decision feedback.
- Figure 7 is a block diagram showing the configuration of an adaptive demodulator configured as an interference canceller.
- Figure 8 is a block diagram showing the configuration of an adaptive demodulator when applied to diversity reception.
- FIG. 3 shows an embodiment of the present invention.
- the adaptive demodulator of this embodiment is configured as a nonlinear equalizer using maximum likelihood sequence estimation in the same manner as in FIG. 2, and includes a maximum likelihood sequence estimation unit 31, an error calculation unit 32, It comprises a unit 37, a repliing force generation unit 38, and an updating unit 41.
- the error calculation unit 32 calculates the complex conjugate of the M input signals x (i), x (il),.
- the operation of the maximum likelihood sequence estimator 31 is described by a state.
- a state For example, when the number of taps M of the trans-perspective filter 34 is 2 (that is, one symbol delay) and the modulator 39 performs QPSK modulation, there are four states, and the state transition diagram (Train) Li scan view), as shown in FIG. 4, each state S ,, S if Ss at time i one 1, either from S 4 of the four states of the time i can be a transition.
- the coefficient vector ⁇ ) is set in the transversal filter 34 corresponding to each state m.
- the input signal x (i) is input from the input terminal 11 to the error calculator 32, and the input signal memory 33 stores N input signal sequences at successive time points. From the input signal memory 33, the following equation using the sequence of the input signal x (i) as an element
- the input signal vector X (i) represented by (1) is output.
- the input signal vector X (i) has a different definition from the input vector (i) in the above description of the linear equalizer.
- N repetition force signal sequences at N consecutive points are accumulated in the replica memory 35 in the error calculator 32.
- the replicator memory 35 outputs a replicator vector Ym (i) having a sequence of the replica signal y ra (i) as an element.
- a Y m H (i) ( yi y-Ci-l .ymCi- M + 1)).
- the subtractor 36 performs a difference operation between the input signal vector X (i) and the replica vector Y m (i) to calculate an error vector E m (i). .
- the Bruno square of Lum error base-vector E m (i) is input is calculated in the maximum likelihood sequence estimation tough 3 1 squarer 3 7.
- the signal is estimated by maximum likelihood sequence estimation section 31 and output from output terminal 14.
- the maximum likelihood sequence estimation unit 31 outputs a code sequence candidate (a m ) to the replica generation unit 38.
- Replicon force generation unit 3 8 includes a modulator 3 9 and transformer bar sulfates I filter 3 4, the modulator 3 9 to the code sequence candidate ⁇ a m ⁇ , by performing the same modulation and transmission-side modulation Output the wave sequence candidate ⁇ s m ).
- Modulated wave sequence candidate ⁇ s snake ⁇ is input to trans-parser filter 34, and is applied to each state transition candidate m of maximum likelihood sequence estimator 31.
- a replica signal (i) is generated by performing an inner product operation (convolution operation) with the corresponding coefficient vector Wi-1).
- each state S at time i there are state transitions from each state at the time i one 1 in ⁇ S 4, 4 one transition are merged. From these merged ones, the state transition with the highest likelihood, that is, the small state transition II E vigorous(i) II 2 , is selected.
- the updating unit 41 to which the present invention is applied includes a generalized inverse matrix generated from the modulated wave sequence candidate (s n ) corresponding to the state transition selected by the maximum likelihood sequence estimating unit 31 and an error-based matrix.
- the coefficient vector W m (i-1) for each state is updated to Wêt(i) by finding the dot product vector with the vector E Kh (i). since candidate ⁇ a m) which is outputted from the one to the number of state transitions for each state, performed for the same input signal x (i), the above-mentioned Starring temperature corresponding to the number of states and state transitions.
- the update of W (i) is specifically performed as follows. First, the basic principle of the update algorithm is described. Modulated wave candidate vector (complex symbol candidate vector) corresponding to state transition candidate m corresponding to code sequence candidate (aong) selected by maximum likelihood sequence estimator 31
- the modulation matrix candidate matrix corresponds to N modulation wave candidate vectors at successive points in time.
- N is an integer of 1 or greater.
- VacheCi A, preparation H Ci) X (i) (3)
- I M is the unit matrix of MX M.
- Equation (4) the by R ra (i) to the original, Moor-Penrose- ⁇ matrix R "+ (i) it is necessary to calculate.
- R m (i) from R m + (i) There are several mathematically known methods of calculating, for example, in the above-mentioned A. Albert document, Chapter 5 shows four methods.
- the first term on the right-hand side of the above equation (4) is a special solution and a minimum norm solution.
- R ⁇ (i) is regular
- R (i) RnT i) the second term on the right-hand side of equation (4) disappears, and the minimum norm solution of the first term becomes a solution.
- m + (i) does not always become RnT '(i).
- W m (i) can be obtained as follows by using this general solution as follows.
- Equation (6) represents the prediction error, and equation (5) corrects Wm (i-1) based on the prediction error and updates it to W ⁇ dress(i).
- the cross-correlation vector V m (i) in equation (6) is calculated from the input signal vector X (i) and the modulated wave candidate transpose A Ci) using equation (3).
- the autocorrelation matrix R m (i) and its general inverse matrix R » + (i) of the modulated wave candidate defined by Equation (2) are obtained for the state transition candidate m. It is not based on the received input signal, but based on a self-correlation function for a pure signal component without fluctuations in the transmission path and no addition of noise. Therefore, it is attached to the state, and the state transition candidate is Corresponding calculations can be made in advance.
- the previously calculated matrices R m (i) and R (i) are stored in the matrix memory 41 M in the updating unit 41 in correspondence with m, and the maximum likelihood sequence ,
- the corresponding state transition candidate m is given to the updating unit 41, and the corresponding autocorrelation matrix R m (i) and its general inverse matrix R m + (i) are obtained from the matrix memory 41M.
- the second update algorithm is
- Equations (5) and (6) are rewritten using the fact that
- the third update algorithm introduces a step size) u as shown below, with respect to equation (5) in the first update method.
- V "(i) can be defined by A m H (i) AX ( i).
- ⁇ is a constant forgetting factor that satisfies 0 ⁇ ⁇ 1.
- Equation (4) or Equations (5) and (6) which are the first updating methods.
- equations (7) and (8) hold only when the rank of (i) is ⁇ and ⁇ , and generally do not hold. Updating by equations (9) and (10) is possible only if) and (8) hold.
- the modulated wave candidate vector S m (i) is also fixed for each state transition. That is, the modulated wave candidate matrix A ⁇ Ci) can be regarded as a matrix A researchinghaving no temporal variation.
- the updating unit 41 can calculate the general inverse matrix R or A réelle + from A m or A m in advance, and store it in the matrix memory 41 M of the updating unit 41.
- the amount of calculation processing can be reduced and the processing time can be reduced. This is a major difference from the orthogonal projection method in the linear equalizer that inputs the received received signal to a transversal filter.
- matrices such as R m + and R »that can be calculated from Am can be calculated in advance and stored in the matrix memory 41 M.
- the transversal filter 34 operates so as to generate a delayed wave component by the multipath transmission line, and the input signal x (i) supplied through the feedforward filter 43 is used.
- the delayed wave component is removed by the subtractor 36, and the obtained desired signal component is determined by the determiner 13 and the result of the determination is output as a code s (i).
- the error e (i) between the input and output of the decision unit 13 is obtained by the error calculator 15, the square of the absolute value of the error is calculated by the squarer 37, and the updating unit 4 to which the present invention is applied Given to 1.
- Updating unit 4 1 I e I s to determine the door la Nsubasarufu I filter 3 4 data class tap coefficient vector preparative Le W (i) to minimize.
- a feedforward filter 43 is provided on the input side, and the updating unit 41 determines that the level of the direct wave component (desired signal component) in the input signal is greater than the level of the delayed wave component.
- the data-up number M of the capital La Nsupasarufu Lee torque 3 4 to give, et al. Is the in which the signal sequence s "(i), s" (i- l), ⁇ , and s m (i-M + l ), If the output of the feedforward filter 43 is x (i), the modulating wave candidate vector S (i), the autocorrelation matrix R (i), and the cross-correlation vector V ( i) can be defined similarly. However, we consider one state where m 1. In this way, the coefficient vector W (i) of the transversal filter 34 can be updated by the updating algorithm used in the above-described equations (4) to (11).
- FIG. 7 shows a case where the demodulator of the present invention is configured as a non-linear interference canceller
- FIG. 8 shows a case where the demodulator is configured as a combination of a non-linear equalizer and dipersity.
- the maximum likelihood sequence estimator 31 sequentially generates all possible sets of desired signal sequence candidates and interference signal sequence candidates and supplies them to the modulators 39, 39 2 , respectively. Modulated signal sequence candidates are generated and given to the trans- versal filters 34,, 34 s. Updating unit 4 1 to which the invention is applied DOO La Nsubasa Ruch I filter 3 4, 3 4 2 of data class tap coefficients, so as to approximate the propagation characteristics of each desired signal channel and interference signal channel The repetition power of the desired signal and the interference signal generated by the control is supplied to a subtractor 3636z.
- the input signals x (i) are sequentially subtracted by the subtracters 36, 36, and the subtraction result is given to the squarer 37 as an error vector £ terme(In the embodiment of FIG. By canceling the interference signal as an unnecessary signal component in the received signal x (i), the determination error of the desired signal is reduced.
- the input terminal 1 1,, 1 1 input signals from the two branches given to x, with respect to (i :), x 2 (i ) each subtractor 3 6, 3, 6 t s squarer 3 7, 3 7 2, preparative La Nsubasarufu I filter 3 4, 3 4 2, updating unit 4 1, 4 1 8 is provided.
- an optimum coefficient vector is set for each signal candidate, and the coefficient vector is updated successively using the Moor-Penrose-general inverse matrix in updating the coefficient vector. Since it has been updated, it has good tracking performance and stable operation is expected. Since the Moor-Penrose generalized inverse matrix prepared for each state transition candidate is determined for each state transition candidate, it can be calculated and stored in memory, and the amount of calculation is a general adaptive signal for calculating the inverse matrix. Less than processing. Also, since there is a maximum likelihood sequence estimation unit, the signal detection performance is excellent.
- the adaptive demodulator and the demodulation method according to the present invention can be applied to fixed radio such as mobile communication and mobile satellite communication with fast fusing fluctuation, and microphone mouth wave which requires a high-precision demodulation circuit for large capacity transmission. Suitable for the field of transmission. Adaptive receivers such as adaptive equalizers, adaptive interference cancellers, and dipersities in these fields can be easily manufactured.
Landscapes
- Engineering & Computer Science (AREA)
- Computer Networks & Wireless Communication (AREA)
- Signal Processing (AREA)
- Power Engineering (AREA)
- Cable Transmission Systems, Equalization Of Radio And Reduction Of Echo (AREA)
- Filters That Use Time-Delay Elements (AREA)
- Digital Transmission Methods That Use Modulated Carrier Waves (AREA)
Description
Claims
Priority Applications (3)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
EP94931682A EP0684706B1 (en) | 1993-11-05 | 1994-11-04 | Replica producing adaptive demodulating method and demodulator using the same |
US08/411,748 US5602507A (en) | 1993-11-05 | 1994-11-04 | Adaptive demodulating method for generating replica and demodulator thereof |
DE69432100T DE69432100T2 (de) | 1993-11-05 | 1994-11-04 | Replikherstellendes adaptives demodulationsverfahren und dieses verwendender demodulator |
Applications Claiming Priority (2)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
JP27672393 | 1993-11-05 | ||
JP5/276723 | 1993-11-05 |
Publications (1)
Publication Number | Publication Date |
---|---|
WO1995012926A1 true WO1995012926A1 (fr) | 1995-05-11 |
Family
ID=17573444
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
PCT/JP1994/001862 WO1995012926A1 (fr) | 1993-11-05 | 1994-11-04 | Procede de demodulation adaptative produisant des repliques et demodulateur l'utilisant |
Country Status (4)
Country | Link |
---|---|
US (1) | US5602507A (ja) |
EP (1) | EP0684706B1 (ja) |
DE (1) | DE69432100T2 (ja) |
WO (1) | WO1995012926A1 (ja) |
Cited By (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP2001244855A (ja) * | 1999-12-14 | 2001-09-07 | Stmicroelectronics Sa | 遠端漏話を除去するdsl伝送システム |
JP2012090239A (ja) * | 2010-10-22 | 2012-05-10 | Nec Corp | 分散推定方法、分散推定装置、変調信号生成方法、変調信号生成装置、および、コンピュータ・プログラム |
JP5859147B2 (ja) * | 2013-01-25 | 2016-02-10 | 三菱電機株式会社 | ビット尤度演算装置およびビット尤度演算方法 |
WO2019142912A1 (ja) * | 2018-01-19 | 2019-07-25 | 日本電信電話株式会社 | シンボル判定装置及びシンボル判定方法 |
Families Citing this family (55)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US5870051A (en) * | 1995-08-14 | 1999-02-09 | William K. Warburton | Method and apparatus for analog signal conditioner for high speed, digital x-ray spectrometer |
US5873054A (en) * | 1995-08-14 | 1999-02-16 | William K. Warburton | Method and apparatus for combinatorial logic signal processor in a digitally based high speed x-ray spectrometer |
US5999355A (en) | 1996-04-30 | 1999-12-07 | Cirrus Logic, Inc. | Gain and phase constrained adaptive equalizing filter in a sampled amplitude read channel for magnetic recording |
US6819514B1 (en) * | 1996-04-30 | 2004-11-16 | Cirrus Logic, Inc. | Adaptive equalization and interpolated timing recovery in a sampled amplitude read channel for magnetic recording |
JP3275779B2 (ja) * | 1997-06-16 | 2002-04-22 | 日本電気株式会社 | 遅延判定帰還型系列推定受信装置 |
JPH1141115A (ja) * | 1997-07-23 | 1999-02-12 | Mitsubishi Electric Corp | ディジタルデータ復調装置 |
US6115418A (en) | 1998-02-09 | 2000-09-05 | National Semiconductor Corporation | Simplified equalizer for twisted pair channel |
US6381290B1 (en) * | 1998-05-15 | 2002-04-30 | Ericsson Inc. | Mobile unit for pilot symbol assisted wireless system and method of improving performance thereof |
US6426972B1 (en) * | 1998-06-19 | 2002-07-30 | Nxtwave Communications | Reduced complexity equalizer for multi mode signaling |
US6415003B1 (en) | 1998-09-11 | 2002-07-02 | National Semiconductor Corporation | Digital baseline wander correction circuit |
US6438163B1 (en) | 1998-09-25 | 2002-08-20 | National Semiconductor Corporation | Cable length and quality indicator |
CA2318929A1 (en) * | 1998-12-22 | 2000-06-29 | Nortel Networks Corporation | Stable adaptive filter and method |
US6754340B1 (en) | 1998-12-22 | 2004-06-22 | Nortel Networks Limited | Stable adaptive filter and method |
JP3258973B2 (ja) * | 1999-03-24 | 2002-02-18 | 三洋電機株式会社 | 伝送チャネルの割当方法およびそれを用いた無線装置 |
US6418172B1 (en) * | 1999-04-21 | 2002-07-09 | National Semiconductor Corporation | Look-ahead maximum likelihood sequence estimation decoder |
US6463099B1 (en) * | 1999-05-18 | 2002-10-08 | The Hong Kong University Of Science And Technology | Blind channel equalizers and methods of blind channel equalization |
US6370191B1 (en) * | 1999-11-01 | 2002-04-09 | Texas Instruments Incorporated | Efficient implementation of error approximation in blind equalization of data communications |
US7254198B1 (en) | 2000-04-28 | 2007-08-07 | National Semiconductor Corporation | Receiver system having analog pre-filter and digital equalizer |
US7050517B1 (en) | 2000-04-28 | 2006-05-23 | National Semiconductor Corporation | System and method suitable for receiving gigabit ethernet signals |
US20020106040A1 (en) * | 2001-02-02 | 2002-08-08 | Sarnoff Corporation | Method and apparatus for reducing multipath distortion in a wireless ian system |
WO2001091331A2 (en) * | 2000-05-22 | 2001-11-29 | Sarnoff Corporation | Method and apparatus for reducing multipath distortion in a wireless lan system |
US7319719B1 (en) * | 2000-06-21 | 2008-01-15 | Adc Telecommunications, Inc. | Parallel equalization for systems using time division multiple access |
US20050063487A1 (en) * | 2001-05-08 | 2005-03-24 | Soheil Sayegh | Method and apparatus for parameter estimation, modulation classification and interference characterization in satellite communication systems |
DE10157247B4 (de) * | 2001-11-22 | 2007-06-14 | Rohde & Schwarz Gmbh & Co. Kg | Verfahren zur gemeinsamen Schätzung von Parametern |
US20030161258A1 (en) * | 2002-02-22 | 2003-08-28 | Jianzhong Zhang | Apparatus, and associated method, for a multiple-input, multiple-output communications system |
ES2376743T3 (es) * | 2002-03-04 | 2012-03-16 | Glowlink Communications Technology, Inc. | Detección y medición de interferencia contenida dentro de una portadora digital |
US6842496B2 (en) * | 2002-10-24 | 2005-01-11 | Interdigital Technology Corporation | Algorithm for multiple-symbol differential detection |
US7016406B1 (en) * | 2003-04-29 | 2006-03-21 | Scintera Networks | Adaptation structure and methods for analog continuous time equalizers |
US20110121179A1 (en) * | 2007-06-01 | 2011-05-26 | Liddiard Steven D | X-ray window with beryllium support structure |
KR20100037615A (ko) * | 2007-07-09 | 2010-04-09 | 브라이엄 영 유니버시티 | 대전된 분자 조작을 위한 방법 및 이를 위한 장치 |
US8498381B2 (en) | 2010-10-07 | 2013-07-30 | Moxtek, Inc. | Polymer layer on X-ray window |
US9305735B2 (en) | 2007-09-28 | 2016-04-05 | Brigham Young University | Reinforced polymer x-ray window |
US20100285271A1 (en) * | 2007-09-28 | 2010-11-11 | Davis Robert C | Carbon nanotube assembly |
WO2009152852A1 (en) * | 2008-06-18 | 2009-12-23 | Telefonaktiebolaget L M Ericsson (Publ) | Intercell interference reduction |
US20100098216A1 (en) * | 2008-10-17 | 2010-04-22 | Moxtek, Inc. | Noise Reduction In Xray Emitter/Detector Systems |
US8247971B1 (en) | 2009-03-19 | 2012-08-21 | Moxtek, Inc. | Resistively heated small planar filament |
US20100239828A1 (en) * | 2009-03-19 | 2010-09-23 | Cornaby Sterling W | Resistively heated small planar filament |
US7983394B2 (en) * | 2009-12-17 | 2011-07-19 | Moxtek, Inc. | Multiple wavelength X-ray source |
US8804910B1 (en) | 2011-01-24 | 2014-08-12 | Moxtek, Inc. | Reduced power consumption X-ray source |
US8750458B1 (en) | 2011-02-17 | 2014-06-10 | Moxtek, Inc. | Cold electron number amplifier |
US8929515B2 (en) | 2011-02-23 | 2015-01-06 | Moxtek, Inc. | Multiple-size support for X-ray window |
US8559494B1 (en) | 2012-06-20 | 2013-10-15 | MagnaCom Ltd. | Timing synchronization for reception of highly-spectrally-efficient communications |
US9166834B2 (en) | 2012-06-20 | 2015-10-20 | MagnaCom Ltd. | Method and system for corrupt symbol handling for providing high reliability sequences |
WO2014016677A2 (en) | 2012-06-20 | 2014-01-30 | MagnaCom Ltd. | Highly-spectrally-efficient transmission using orthogonal frequency division multiplexing |
US8781008B2 (en) | 2012-06-20 | 2014-07-15 | MagnaCom Ltd. | Highly-spectrally-efficient transmission using orthogonal frequency division multiplexing |
US9088400B2 (en) | 2012-11-14 | 2015-07-21 | MagnaCom Ltd. | Hypotheses generation based on multidimensional slicing |
US8811548B2 (en) | 2012-11-14 | 2014-08-19 | MagnaCom, Ltd. | Hypotheses generation based on multidimensional slicing |
US9118519B2 (en) | 2013-11-01 | 2015-08-25 | MagnaCom Ltd. | Reception of inter-symbol-correlated signals using symbol-by-symbol soft-output demodulator |
US8804879B1 (en) | 2013-11-13 | 2014-08-12 | MagnaCom Ltd. | Hypotheses generation based on multidimensional slicing |
US9130637B2 (en) | 2014-01-21 | 2015-09-08 | MagnaCom Ltd. | Communication methods and systems for nonlinear multi-user environments |
US9496900B2 (en) | 2014-05-06 | 2016-11-15 | MagnaCom Ltd. | Signal acquisition in a multimode environment |
US8891701B1 (en) | 2014-06-06 | 2014-11-18 | MagnaCom Ltd. | Nonlinearity compensation for reception of OFDM signals |
US9246523B1 (en) | 2014-08-27 | 2016-01-26 | MagnaCom Ltd. | Transmitter signal shaping |
US9276619B1 (en) | 2014-12-08 | 2016-03-01 | MagnaCom Ltd. | Dynamic configuration of modulation and demodulation |
US9191247B1 (en) | 2014-12-09 | 2015-11-17 | MagnaCom Ltd. | High-performance sequence estimation system and method of operation |
Citations (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JPH05110617A (ja) * | 1991-10-14 | 1993-04-30 | Oki Electric Ind Co Ltd | 適応最尤系列推定器 |
JPH06268540A (ja) * | 1993-03-12 | 1994-09-22 | Fujitsu Ltd | ビタビ等化方法 |
Family Cites Families (8)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US5251233A (en) * | 1990-12-20 | 1993-10-05 | Motorola, Inc. | Apparatus and method for equalizing a corrupted signal in a receiver |
JP2876856B2 (ja) * | 1991-10-31 | 1999-03-31 | 日本電気株式会社 | 系列推定方法および装置 |
US5363408A (en) * | 1992-03-24 | 1994-11-08 | General Instrument Corporation | Mode selective quadrature amplitude modulation communication system |
JP2720721B2 (ja) * | 1992-08-21 | 1998-03-04 | 日本電気株式会社 | 変復調装置 |
JPH0744473B2 (ja) * | 1993-02-02 | 1995-05-15 | 日本電気株式会社 | 復調システム |
US5490169A (en) * | 1993-06-16 | 1996-02-06 | Universal Data Systems, Inc. | Decision feedback equalizer method and apparatus |
US5371471A (en) * | 1993-10-29 | 1994-12-06 | General Electric Company | Low complexity adaptive equalizer radio receiver employing direct reference state updates |
US5471504A (en) * | 1994-04-14 | 1995-11-28 | Computer & Communication Research Laboratories | Bilinear decision feedback equalizer |
-
1994
- 1994-11-04 EP EP94931682A patent/EP0684706B1/en not_active Expired - Lifetime
- 1994-11-04 US US08/411,748 patent/US5602507A/en not_active Expired - Lifetime
- 1994-11-04 WO PCT/JP1994/001862 patent/WO1995012926A1/ja active IP Right Grant
- 1994-11-04 DE DE69432100T patent/DE69432100T2/de not_active Expired - Lifetime
Patent Citations (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JPH05110617A (ja) * | 1991-10-14 | 1993-04-30 | Oki Electric Ind Co Ltd | 適応最尤系列推定器 |
JPH06268540A (ja) * | 1993-03-12 | 1994-09-22 | Fujitsu Ltd | ビタビ等化方法 |
Cited By (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP2001244855A (ja) * | 1999-12-14 | 2001-09-07 | Stmicroelectronics Sa | 遠端漏話を除去するdsl伝送システム |
JP2012090239A (ja) * | 2010-10-22 | 2012-05-10 | Nec Corp | 分散推定方法、分散推定装置、変調信号生成方法、変調信号生成装置、および、コンピュータ・プログラム |
JP5859147B2 (ja) * | 2013-01-25 | 2016-02-10 | 三菱電機株式会社 | ビット尤度演算装置およびビット尤度演算方法 |
WO2019142912A1 (ja) * | 2018-01-19 | 2019-07-25 | 日本電信電話株式会社 | シンボル判定装置及びシンボル判定方法 |
US11349524B2 (en) | 2018-01-19 | 2022-05-31 | Nippon Telegraph And Telephone Corporation | Symbol-determining device and symbol determination method |
Also Published As
Publication number | Publication date |
---|---|
US5602507A (en) | 1997-02-11 |
EP0684706A4 (en) | 1996-11-20 |
DE69432100T2 (de) | 2003-09-25 |
EP0684706A1 (en) | 1995-11-29 |
EP0684706B1 (en) | 2003-02-05 |
DE69432100D1 (de) | 2003-03-13 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
WO1995012926A1 (fr) | Procede de demodulation adaptative produisant des repliques et demodulateur l'utilisant | |
JP3560991B2 (ja) | 適応型最尤系列推定装置 | |
US5175747A (en) | Equalizer | |
US5068873A (en) | Equalizer | |
Malik et al. | Adaptive equalization algorithms: an overview | |
JPH06197032A (ja) | 適応型等化器 | |
Shah et al. | Design of fractional-order variants of complex LMS and NLMS algorithms for adaptive channel equalization | |
EP0453814B1 (en) | Adaptive channel impulse response estimation system using maximum likelihood sequence estimation | |
US5272726A (en) | Blind type sequence estimator for use in communications system | |
JPH077461A (ja) | 受信信号の等化方法 | |
Guo et al. | Frequency-domain set-membership filtering and its applications | |
Pickholtz et al. | The recursive constant modulus algorithm; a new approach for real-time array processing | |
JP3808311B2 (ja) | 受信方法及び受信機 | |
Miranda et al. | Avoiding divergence in the Shalvi–Weinstein algorithm | |
JP3256646B2 (ja) | 適応干渉キャンセル受信機 | |
Mohanty et al. | Block NLMS/F-based equalizer design and channel capacity analysis for indoor IEEE 802.11 fading wireless channels | |
JPH10173573A (ja) | 適応等化器 | |
JP2846958B2 (ja) | レプリカを生成する適応形復調方法及びそれを使った復調器 | |
Cid-Sueiro et al. | Recurrent radial basis function networks for optimal blind equalization | |
JP3304035B2 (ja) | ダイバーシチ受信装置 | |
US6704354B1 (en) | Method and apparatus of providing adaptive equalization using sign permutation filtering | |
JP2986488B2 (ja) | 等化器 | |
JP3180240B2 (ja) | 適応等化器 | |
JP3424724B2 (ja) | 干渉キャンセラ | |
Katwal et al. | A simple Kalman channel Equalizer using adaptive algorithms for time variant channel |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
WWE | Wipo information: entry into national phase |
Ref document number: 08411748 Country of ref document: US |
|
WWE | Wipo information: entry into national phase |
Ref document number: 1994931682 Country of ref document: EP |
|
AK | Designated states |
Kind code of ref document: A1 Designated state(s): JP US |
|
AL | Designated countries for regional patents |
Kind code of ref document: A1 Designated state(s): AT BE CH DE DK ES FR GB GR IE IT LU MC NL PT SE |
|
121 | Ep: the epo has been informed by wipo that ep was designated in this application | ||
WWP | Wipo information: published in national office |
Ref document number: 1994931682 Country of ref document: EP |
|
WWG | Wipo information: grant in national office |
Ref document number: 1994931682 Country of ref document: EP |