• 专利附录
  • 专利说明
  • 权利要求
  • 类似技术
  • 同族专利
  • 引用文献
  • 法律状态
  • 优先权
    • 专利摘要:
    A method of processing an analog channel signal from a transmission channel, the analog channel signal (SAC) being capable of comprising a useful signal (SU) modulated on a subset of a set of available carriers, carrying at least one frame of symbols according to a frame structure and optionally noisy by at least one narrow-band noise signal (SB), the method comprising an analog-to-digital (CAN) conversion of the channel analog signal and a synchronization process comprising a filtering processing (MFL) including a time domain on-the-fly determination of a limited number of coefficients of a predictive filter of a self regressive model of a digital channel signal (CNS) derived from said conversion analog / digital and on-the-fly filtering of the digital channel signal in the time domain by a finite impulse response digital filter whose coefficients are those of the filter predictive, and detection of at least one indication (IND) for identifying at least one location of said frame structure, from the filtered channel digital signal (SNF) and a reference signal (SREF).
    公开号:FR3034274A1
    申请号:FR1552588
    申请日:2015-03-27
    公开日:2016-09-30
    发明作者:Mark Wallis；Yoann Bouvet；Pierre Demaj
    申请人:STMicroelectronics Rousset SAS；
    IPC主号:
    专利说明:

    [0001] Process for processing an analog signal originating from a transmission channel, in particular an in-line carrier current signal Embodiments and embodiments of the invention relate to the transmission of information over a transmission channel. communication, and especially when this channel is a power line, the transmission of information by line power line (PLC: Power Line Communications), and more particularly the improvement of the processing of such a signal in reception when it is noisy by a narrow band noise signal (Narrow Band Interferer). Embodiments and embodiments of the invention are compatible with the various standards governing online powerline communication, in particular but not exclusively the PLC-G3, PRIME (PoweRline Intelligent Metering Evolution) standards or the IEEE standard. 1901-2. Powerline carrier technology aims to transmit digital data by exploiting the existing infrastructure of the power grid. It allows remote reading of electricity meters, exchanges between electric vehicles and charging stations, and management and control of smart grids. narrow-band power line communication (N-PLC) which is generally defined as communication over a power line operating at transmission frequencies up to 500 KHz. The N-PLC communication thus generally uses the frequency bands defined in particular by the European Committee for Electrotechnical Standardization (CENELEC) or by the Federal Communications Commission (FCC).
    [0002] 3034274 2 Thus, if we consider the CENELEC A frequency band (395 kHz), the transmission frequencies are between 42 and 89 KHz in the PRIME standard whereas they are between 35 and 91 KHz for the PLC standard. G3.
    [0003] In such frequency bands, the electrical cables carrying the in-line carrier signals are in a very harsh environment. They are particularly subject to disturbances of the type white noise, colored noise or noise pulse. In addition, they are not protected against interference. As a result, any FM / AM radio signal or wireless communication can lead to the presence of harmonics of these signals in the useful frequency band used by narrow band PLC communications. Moreover, the properties and characteristics of the electrical networks are not known a priori and are variable in the time. Thus, interference can be created on a power line when a user plugs any device such as a hairdryer or a washing machine. This results in a propagation of strong frequency harmonics which can also be located in the useful band of PLC communications.
    [0004] And such noise signals, which are generally Narrow Band Interferer signals, that is, having a lower frequency band than the frequency band of the wanted signal, then disturb the synchronization phase. the receiver connected to the power line, during which the receiver must be able to synchronize to find in particular the beginning of the useful data of the frame of symbols carried by the useful signal. Brian Michael Donlan's thesis entitled "Ultra-Broadband Narrowband Interference Cancellation and Channel Modeling for Communications", 31 January 2005, Blacksburg, Virginia, describes 30 different techniques for suppressing narrow-band noise signals from an ultra-wideband signal ( UWB) particularly in the context of Spread Spectrum.
    [0005] The techniques mentioned in this document use, for some of them, predictive filters in order to estimate the noise signal before subtracting it from the received signal. The signals discussed in this document have very different characteristics of the signals used in line carrier communications. Indeed, the UWB signals (and in particular those with direct sequence spectrum spreading) have a spread of the power of the signal transmitted over a wide frequency band in order to drown this power in the ambient noise or in the other communications. . Thus, the Power Spectral Density (PSD) of a UWB signal is generally defined as less than -41 dBm / MHz. The signals used in the PLC communications are signals modulated according to a multicarrier modulation, for example a quadrature modulation on orthogonal carriers (OFDM modulation: Orthogonal Frequency Division Multiplexing, according to an Anglosaxon name commonly used by those skilled in the art. ), but using only a subset of carriers out of a larger set of available carriers. Thus, for example, if we consider the CENELEC A frequency band, the size of the inverse Fourier transform and the direct Fourier transform is equal to 512, while only 97 subcarriers (the subcarriers 86 to 182 ) are used for transmission in the PRIME standard. If we consider the CENELEC A frequency band, the size of the inverse Fourier transform and the direct Fourier transform is equal to 256 whereas only 36 subcarriers (the subcarriers 23 to 58) are used in the PLC-G3 standard.
    [0006] Furthermore, it is important, during the synchronization phase, not to miss any symbol coming from the channel itself when it is noisy. According to an implementation and implementation, it is proposed a processing of an analog signal from the transmission channel which makes it possible to improve the performance of the synchronization phase whether the useful analog signal is noisy or not by at least one narrowband noise signal. According to another embodiment and embodiment, it is proposed to also improve the performance of the decoding of the remaining part of the frame after synchronization. According to an embodiment and embodiment, in order to facilitate the synchronization of the receiver, especially when the wanted signal is noisy by at least one narrow-band noise signal, the coefficients in the time domain are determined on the fly. predictive filter of an autoregressive model of the signal and the signal is filtered on the fly in the time domain by a finite impulse response (FIR) digital filter whose coefficients are those of the predictive filter.
    [0007] This is therefore different from the filtering methods of the prior art in which the noise signal is estimated before subtracting it from the signal from the channel, which requires perfect synchronization, both in time and in phase, and this which is difficult to achieve in practice, especially in the case of a synchronization phase where it is important not only to lose no symbol and secondly not to introduce phase errors. According to this implementation and embodiment, instead of estimating and then subtracting the estimated noise signal, the overall signal coming from the channel is filtered, whether or not this signal contains the noise signal, which makes it possible to to perform a non-coherent processing that is to say not requiring a perfect synchronization in time and in phase between the moment when the noise signal is estimated and the time when this signal noise signal is subtracted global. The receiver will thus be able to synchronize from the filtered signal and a reference signal, for example a known symbol. According to one aspect, it is therefore proposed a method of processing an analog channel signal from a transmission channel, for example a power line.
    [0008] The analog channel signal is capable of comprising a useful signal modulated on a subset of a set of available carriers, for example the useful signals complying with the PRIME or G3-PLC standards.
    [0009] This useful signal carries at least one frame of symbols according to a frame structure and this useful signal is optionally noisy by at least one narrow-band noise signal. In general, the noise signal is a single-frequency noise peak contained in the frequency band of the wanted signal, but more generally, a narrow-band noise signal is a noise signal whose frequency band is less than the frequency band of the wanted signal. Thus, the channel analog signal may quite at any given time have no useful signal, or may comprise only at least one noise signal, or a non-noisy useful signal or a noisy useful signal. The method according to this aspect then comprises a digital analog conversion of the analog channel signal and a synchronization processing including a filtering process.
    [0010] The analog channel signal that will undergo the analog / digital conversion may be, for example, the analog signal directly from the channel or, as is generally the case, the analog signal delivered by an analog input stage (including in particular bandpass filters, low pass filters and an amplifier) connected to the transmission channel. The filtering processing includes a time-domain on-the-fly determination of a limited number of coefficients of a predictive filter of an autoregressive model of a digital channel signal from said analog-to-digital conversion and a filtering at 30. the flight of the digital channel signal in the time domain by a finite impulse response digital filter whose coefficients are those of the predictive filter. The digital channel signal on which the filtering processing is carried out is not necessarily the digital signal directly derived from the analog / digital conversion but may be for example the digital signal resulting from the analog / digital conversion and having undergone possibly subsampling.
    [0011] The signals modulated on a subset of carriers from a set of available carriers have characteristics totally different from the UWB or spread spectrum signals. They have in fact in particular a power level much higher than that of a UWB signal or spread spectrum and it is then preferable to take precautions in the filtering so as to avoid completely filtering the useful signal in the absence of a narrow band noise. In fact, especially in view of the fact that only a few carriers are used for modulation among the available set of carriers (size of the inverse or direct Fourier transform), it is preferable to limit the number of filter coefficients so as to avoid having too much attenuation of the useful signal in the absence of a noise signal. In other words, the number of coefficients of the filter is advantageously less than or equal to a limit number which is chosen so as to form a finite impulse response filter whose frequency response comprises, in the presence of the noise signal, a notch at the level of the frequency band of the noise signal, and whose frequency response has, in the absence of a noise signal, a relatively flat profile in the frequency band of the wanted signal so as to allow attenuation of the lower useful signal at a chosen value, for example 6 dB, which of course depends on the intended application. The number of coefficients will moreover preferentially be equal to this limit number so as to more easily take into account several narrow-band noise signals with different tones. Those skilled in the art will be able to adapt the limit number of coefficients according to the intended application and the desired characteristics for the maximum acceptable attenuation of the useful signal in the absence of a noise signal. Having said this, the inventors have observed that when each symbol comprises a cyclic prefix, the acceptable limit number of filter coefficients is of the order of three quarters, and preferably of half, the length of the cyclic prefix expressed as a number of samples. It will be recalled here that the cyclic prefix mainly makes it possible to eliminate intersymbol interference and is a technique which consists of copying a part of a symbol to place it upstream of this symbol. In practice, the poles of the predictive filter resulting from the autoregressive model of the signal become the zeros of the finite impulse response filter. Also, the FIR filter can only attenuate the frequencies corresponding to these zeros. Also, so as to further limit the attenuation of the useful signal during the filtering, it is advantageous that the coefficients of the predictive filter are calculated at a calculation frequency of between 2 and 5 times, and preferably between 2 and 3 times, the frequency maximum of the digital channel signal. This allows the useful signal to be considered for the filter as somehow white noise with respect to the noise signal, which would not be the case if the calculation frequency was much higher than the maximum frequency of the digital signal. of channel. The determination of the coefficients of the filter and the actual filtering are performed in the time domain and on the fly, that is to say, as the analog channel signal arrives. This makes it possible not to miss a symbol whether the signal is noisy or not. The method according to this aspect also includes detecting at least one indication for identifying at least one location of said frame structure from the filtered channel digital signal and a reference signal.
    [0012] Said indication may be for example the recognition of a known symbol of the preamble of a frame and the reference signal may be this known symbol, the detection then taking place for example by sliding correlation operations.
    [0013] According to one embodiment, said on-the-fly determination of said coefficients and said on-the-fly filtering comprise a grouping of the samples of said digital signal into successive groups of samples, a determination of a current block of coefficients using the group. sample stream, and an application on said current group of the finite impulse response filter having said current block of coefficients so as to obtain a filtered current group of samples. According to one embodiment, each frame comprises a preamble comprising known symbols and preceding the remaining part of the frame, and the filtering treatment is applied at least to detect said at least one indication at the level of the preamble of FIG. least one frame. When the transmission is totally asynchronous, that is to say that the receiver does not know the time difference between successive frames, it is particularly advantageous for the filtering processes to be applied at least to detect said indication at the level of the message. preamble of each frame. At the level of the filtered digital signal, it can not be known whether this filtered signal results from a noisy useful signal or a noisy useful signal. Also, it is particularly advantageous to perform after the detection of said indication, a check of the presence or absence of the noise signal, for example from at least one known symbol of the unfiltered useful signal. Indeed, this will make it possible to improve the processing of the subsequent symbols of the frame.
    [0014] This verification may include direct Fourier transform processing on said known unfiltered symbol and analysis of the power of each carrier. This verification is for example carried out on at least one symbol of the preamble.
    [0015] And, in the absence of the noise signal, the processing of the remaining part of the frame is advantageously performed on the unfiltered digital channel signal, which makes it possible to decode the symbols of the remaining portion of the signal. the frame on an unfiltered signal 5 that is to say not attenuated. On the other hand, in the presence of the noise signal, the processing of the remaining part of the frame will be performed on the filtered channel digital signal. This processing of the remaining part comprises a direct Fourier transform processing, a demapping process providing for each carrier a value of the modulation coefficient (each symbol comprises modulation coefficients, or "bins", respectively associated with the carriers) and a determination for each coefficient of modulation of a soft indication of said value. It is then particularly advantageous to force the confidence indications of the modulation coefficients associated with the carriers whose frequencies correspond to those of the noise signal to zero. This makes it possible to improve the performance of the decoding 20 especially when a Viterbi-type de-interleaver-decoder pair (or any other decoder based on soft decisions) is used, It is also preferable, so as not to disturb the decoding too much. In another aspect, there is provided a receiver, comprising an input stage intended to be connected to a transmission channel and configured to receive the filter coefficients while processing the remaining portion of the frame. supplying an analog channel signal from the transmission channel, the channel analog signal being capable of comprising a useful signal modulated on a subset of a set of available carriers, carrying at least one frame of symbols according to a structure of frame and possibly noisy by at least one narrowband noise signal, an analog / digital conversion stage for performing a analog-to-digital conversion of the analog channel signal, and a processing stage comprising filtering means including a calculation module configured to determine on the fly a limited number of coefficients of a predictive filter of a self regressive model of a digital channel signal from the analog / digital conversion stage, a finite impulse response digital filter whose coefficients are those of the predictive filter, for on-the-fly filtering of the digital channel signal in the time domain and detection means configured to detect at least one indication for identifying at least one location of said frame structure from the filtered digital signal and a reference signal. According to one embodiment, each symbol comprises a cyclic prefix and the number of coefficients of the filter is less than or equal to a limit number which of the order of three quarters, preferably of the order of one-half, of the length of the cyclic prefix expressed in number of samples. According to one embodiment, the calculation module is configured to calculate the coefficients of the predictive filter at a calculation frequency of between 2 and 5 times, preferably between 2 and 3 times, the maximum frequency of said digital channel signal. According to one embodiment, the processing stage comprises grouping means configured to group samples of said digital channel signal into successive groups of samples, the calculation module is configured to determine a current block of coefficients using the current sample group, and the digital filter is configured to input said current group of samples so as to output a filtered current group of samples. According to one embodiment, each frame comprises a preamble comprising known symbols and preceding the remaining part of the frame, and the processing stage comprises control means configured to deliver said digital channel signal 3034274 11 to the filtering means at the same time. least so that the detection means detect said at least one indication at the level of the preamble of at least one frame. According to one embodiment, the control means are configured to deliver said digital channel signal to the filtering means at least so that the detection means detect said at least one indication at the preamble of each frame. According to one embodiment, the receiver further comprises verification means configured to perform after the detection of said at least one indication, a verification of the presence or absence of said noise signal from at least one symbol of the unfiltered useful signal. According to one embodiment, the verification means comprise a direct Fourier transform stage configured to carry out a Fourier transform process directly on the said at least one symbol and analysis means configured to perform a power analysis. each carrier. According to one embodiment, the verification means are configured to carry out said verification on at least one symbol of the preamble. According to one embodiment, the processing stage further comprises additional processing means configured to perform a processing of said remaining part of each frame and in the absence of the noise signal, the control means are configured to delivering said remaining portion of the frame directly to the additional processing means without passing through the filtering means. According to one embodiment, each symbol comprising modulation coefficients respectively associated with the carriers, the additional processing means comprise a direct Fourier transform stage, a demapping means providing for each carrier a value of said modulation coefficient and a module able to determine for each modulation coefficient a confidence indication of said value, and forcing means 3034274 12 configured to, in the event of presence of said noise signal, force to zero the confidence indications of the modulation coefficients associated with the carriers whose the frequencies correspond to those of the noise signal.
    [0016] According to one embodiment, the control means are configured to deactivate the module for calculating the coefficients of the filter during the processing of said remaining portion of the frame (freezing of the coefficients of the filter). According to one embodiment, the useful signal is a signal modulated according to an OFDM modulation. Furthermore, the transmission channel may be an electrical line, the useful signal then being an in-line carrier current signal. In yet another aspect, it is proposed, independently of an application to a synchronization, a method of filtering an analog channel signal from a transmission channel, the analog channel signal being capable of comprising a signal useful, for example modulated on a subset of a set of available carriers, and possibly noisy by at least one narrowband noise signal, the method comprising analog / digital conversion of the analog channel signal and a processing of filtering including on-the-fly time-domain determination of coefficients of a predictive filter of a self regressive model of a digital channel signal from said analog-to-digital conversion and on-the-fly filtering of the digital channel signal in the time domain by a finite impulse response digital filter whose coefficients are those of the predictive filter. The number of coefficients is advantageously limited as indicated above and / or the coefficients of the predictive filter are calculated at a calculation frequency of between 2 and 5 times, preferably between 2 and 3 times, the maximum frequency of the digital channel signal. . According to yet another aspect, there is provided a receiver intended to be connected on the transmission channel and comprising means, such as those defined above, configured to implement such a filtering method. Other advantages and characteristics of the invention will appear on examining the detailed description of embodiments and embodiments of the invention, in no way limiting, and the appended drawings in which: FIGS. 1 to 10 schematically illustrate different modes of implementation and embodiment of the invention. The modes of implementation and of realization that will now be described are in the context of an information transmission by line carrier (CPL), although the invention is not limited to this type of application . In what follows, each time that we cite to non-limiting examples the PLC-G3 or PRIME standards, it will be assumed that we consider the frequency band CENELEC A (3-95 kHz). Reference is now made to FIG. 1 to diagrammatically illustrate an example of a transmitter capable of transmitting a useful signal SU on an electric line LE by an in-line carrier current.
    [0017] The transmission chain comprises, for example, an ENC encoder, for example a convolutional encoder, receiving the data to be transmitted from source encoding means. INTL interleaving means are connected to the output of the encoder and are followed by mapping means which transform the bits into symbols according to a transformation scheme depending on the type of modulation used, for example BPSK type modulation or more generally QAM modulation. Each symbol contains modulation coefficients associated with carriers that will be modulated accordingly. The symbols are delivered at the input of MTFI processing means for performing a Fast Reverse Fourier Transform (IFFT) operation. It will be noted here, with particular reference to FIG. 2, that the modulated carriers form an SNS subassembly of the title 3034274 14 carrying among an available set of ENSs of carriers (set which corresponds to the size of the inverse Fourier transform) . Thus, in the PLC-G3 standard, the size of the inverse Fourier transform is equal to 256 while the modulated carriers of the subset SNS lie between the ranks 23 and 58, which corresponds to a frequency band F1- F2 between 35 and 91 KHz. The sampling frequency here is equal to 400 KHz leading to a carrier spacing equal to 1.5625 KHz, which thus makes the orthogonal frequencies (OFDM modulation).
    [0018] In the PRIME standard, the size of the inverse Fourier transform is equal to 512 while the number of carriers of the subset SNS is equal to 97, which provides for the useful signal a frequency band ranging between 42 and 89 KHz. The modulation coefficients associated with the unused carriers are equal to 0. The OFDM signal in the time domain is generated at the output of the processing means MTFI, and means MCP add to each OFDM symbol in the time domain a cyclic prefix which is a duplicate at the head of the OFDM symbol of a number of samples at the end of this symbol. For example, in the PLC-G3 standard, the length of the cyclic prefix is 30 samples for a sampling frequency of 400 KHz while it is 48 samples for a sampling frequency of 250 KHz in the PRIME standard.
    [0019] The signal is then converted into a DAC digital-to-analog converter and then processed in an ETA stage, commonly referred to by those skilled in the art as "Analog Front End", where it undergoes, in particular, power amplification, before being transmitted on the electric line LE.
    [0020] In reception, it will be seen, with particular reference to FIG. 3, that the receiver RCP here comprises an analog input stage ET1 whose input terminal BE is connected to the electric line LE.
    [0021] This analog input stage ET1 conventionally comprises a bandpass filter BPF, a low pass filter LPF, as well as AMP amplification means. The output of the stage ET1 is connected to a CAN analog / digital conversion stage whose output is connected to the input of a processing stage ET2. The processing stage ET2 here comprises automatic gain control means AGC making it possible to control the value of the gain of the amplification means AMP of the stage ET1. The signal SAC delivered at the output of the analog stage ET1 and at the input of the analog / digital conversion stage CAN designates an analog channel signal coming from the transmission channel (electrical line) LE. By way of non-limiting example, FIG. 4 diagrammatically illustrates the frequency spectrum of such an analog SAC channel signal. It can be seen that this signal SAC comprises the useful signal SU conveying the data transmitted from the transmitter and whose frequency band is located between the frequencies F 1 and F 2 corresponding to the numbers of the modulated carriers.
    [0022] The signal SAC also optionally includes a narrow-band noise signal SB, which possibly noises the useful signal SU. Generally, the noise signal SB comprises a single tone located at the frequency F3. It may, however, in practice be distributed on the frequency carrier F3 as well as on some adjacent carriers. It can be seen that the signal SU is a dome-shaped signal whose level is much higher than the noise level AWGN of the channel in the absence of a signal. The level of the noise signal SB is higher than the level of the useful signal SU.
    [0023] Turning now to FIG. 3, it can be seen that the ET2 processing stage also comprises a LPF2 low pass filter followed, although this is not indispensable, by means of MSCH sub-sampling means. The sampling rate of the upstream signal of the MSCH means is noted Fs while the sampling frequency of the output signal of the MSCH means is noted Fss. The SNC signal at the output of the MSCH means then here a digital channel signal which is derived from the analog / digital conversion of the analog signal SAC channel and on which will be applied in particular a filtering treatment in MSL filtering means as we will see it in more detail below. The frequency Fc designates in the following the calculation frequency at which will be calculated in particular the filter coefficients of the MFL filtering means. In the G3-PLC standard, for example, the specified sampling frequency Fs is 400 KHz for an FFT size of 256. Although it would have been possible to carry out all the filtering operations at a calculation frequency Fc equal to the sampling frequency Fs of 400 KHz, subsampling the signal at a frequency Fss lower than Fs and performing all the filtering operations at the calculation frequency Fc equal to Fss makes it possible to reduce the complexity implementation of the processing stage, and in particular filtering means, and also makes it possible to carry out a fast fast Fourier transform (FFT) processing having a reduced size compared to the specified size of 256. Before see in more detail the structure of the filter means MFL and the other means which are incorporated in the processing stage ET2, reference is now made more particularly to FIG. structure of a frame carrying symbols, for example within the framework of the PLC-G3 standard. The TRM frame comprises a preamble PRM comprising here eight known symbols SYNCP followed by an opposite phase symbol SYNCM itself followed by a half symbol SYNCM. The TRM frame then comprises a header (header) HD followed by a PLD field containing useful data symbols to be decoded and more known by those skilled in the art under the Anglo-Saxon name of "payload".
    [0024] The symbols of the header HD contain, in particular, control information for the decoding of the data of the PLD field as well as the number of bytes to be decoded in the PLD field. The preamble PRM of the TRM frame allows the receiver to synchronize, that is to say to obtain an indication allowing to find the structure of the frame in order to be able to locate the beginning of the header HD. The MFL filtering means will be applied, at least during the synchronization phase of the receiver and possibly as will be seen in more detail hereinafter during the decoding phase of the remaining portion of the TRM frame (header and field PLD) in the case where the presence of a noise signal is proved. The MFL filtering means will determine on the fly the coefficients of a predictive filter of an autoregressive model of the digital signal of the SNC channel and then filter on the fly the digital signal of the channel in the time domain by a digital response filter. finite impulse whose coefficients are those of the predictive filter. As is well known to those skilled in the art, a signal can be modeled using convoluted white noise with an autoregressive filter. The model parameters (the predictive filter coefficients and the variance of the prediction error) can be estimated from the autocovariance of the signal by solving the Yule Walker equations: Rn + 1An - r0 r1 r1 r1 * - 1 2 - rn - Crf, n - o - - r1 ro 3034274 18 in which An = 1 are the n coefficients of the filter An, 1 An, n predictive of the autoregressive model of order n and 6f2 is the variance of the error prediction. The sign * denotes the conjugate complex. The autocovariance sequence Rn, = [ro r1-ro] can be estimated by the following formula: 1 N-1-k rk-I); where n is a sequence of N samples of the input signal.
    [0025] In general, N must be large enough to include any periodic content of the signal and to randomize any non-periodic content. However, in practice, N may be equal to the size of the symbol, possibly downsampled, which also corresponds to the size of the Fourier transform. Several algorithms exist to solve the Yule Walker equations. One can quote the Levinson algorithm, or the Durbin-Watson algorithm or the Burg algorithm or a less square type algorithm.
    [0026] Those skilled in the art may refer in this regard to page 879 of Appendix A of John G. Proakis, 3rd edition, entitled Digital Communications or to chapters 11-42 and 11-1. 2 of this same work. When the Levinson algorithm is used, it is a recursive algorithm that calculates the coefficients one by one according to for example the following sequence: A o = [110-f 2 o = ro repeat for m = 0 to n: 3034274 A m + 1 [rm +1 - - - K m + 1- m +1 2 f, m Am Am + 1 = 0 5 20_ 2 K 2) Cr f, m +1 f, mm +1 One Once the coefficients A. of the predictive filter are determined, then a finite impulse response filter 10 (FIR filter) is constructed whose z-transfer function is defined by the formula below: 1 + A1z-1 + A2z- 2 + A3z-2 + ... In this formula, the coefficients A. of the FIR filter are the coefficients A. of the predictive filter of the autoregressive model mentioned above. The set of coefficients of the filter is advantageously limited, that is to say less than or equal to a limit number and preferably equal to this limit number. Indeed, especially in view of the fact that the useful signal is modulated on only a subset of carriers of a set of available carriers, too many filter coefficients could cause too much attenuation of the signal during the filtering especially when there is no noise signal. In general, the limit number of coefficients is chosen by those skilled in the art, taking into account the application and the envisaged specifications, so that, as illustrated schematically in FIG. 6, the frequency response H1 of the filter in FIG. presence of a narrow band noise signal has a notch around the frequency F3 of the noise signal and so that the frequency response H2 30 of this filter has, in the absence of the noise signal, a relatively flat profile in the frequency band F 1, F 2 of the useful signal of 19 r114m + K m + 1 0 JA * m 3034274 so as to obtain a signal attenuation lower than an acceptable limit attenuation. This acceptable limit attenuation depends on the implementation and the dynamics supported by the different processing means.
    [0027] Those skilled in the art will be able to choose this acceptable limit attenuation according to these conditions. However, by way of nonlimiting example, the acceptable limit attenuation can be of the order of 6 dB. In practice, to satisfy this condition, it will be possible for example to choose a number of coefficients of the filter less than or equal to three quarters, and preferably to half, of the length of the cyclic prefix expressed in number of samples taking into account the Fc calculation frequency used. In the PLC-G3 standard, the length of the cyclic prefix for a sampling frequency Fs of 400 KHz is 30 samples. Thus, if we consider the PLC-G3 standard, we can choose for example a number of coefficients equal to 15Fc / Fs. In practice, as illustrated in FIG. 7, the on-the-fly filtering process in the time domain provides a grouping of samples (step 70) so as to form a current group GR of N samples. Then, in step 71, the coefficients of the predictive filter are calculated by running the Levinson algorithm according to the above-mentioned sequence for m varying from 0 to the limit value of the number of coefficients. Then, in step 72, the current group GR of N samples in the time domain is filtered with the finite impulse response filter whose coefficients are those which have just been calculated for the predictive filter.
    [0028] A GRF group of N filtered samples is then obtained. Moreover, it is particularly advantageous for the calculation frequency Fc of the filter coefficients (equal to the frequency Fs or possibly to the frequency Fss in the case of sub-sampling) to be not too great compared with the maximum frequency of the filter. If the computation frequency Fc is too high relative to this maximum frequency, the digital channel signal will not be seen as a "white" noise with respect to the signal of the channel. noise and the risk of having too much attenuation of the useful signal, it being possible to choose a calculation frequency Fc between 2 and 5 times, and preferably between 2 and 3 times, the maximum frequency of the digital channel signal on which In practice, as illustrated in FIG. 8, the filtering means MFL functionally comprise means MGR configured to group the filter coefficients. samples in groups of samples, an MCL module for calculating the coefficients of the predictive filter and an FIR module implementing the finite impulse response filter. In practice, these various means and modules may for example be made in software in a microprocessor. Referring again more particularly to FIG. 3, it can be seen that the filtered digital signal SNF delivered by the filtering means MFL is used in particular by means of synchronization means MSYNC, of conventional structure and known per se, for allow the RCP receiver to synchronize, that is to say for example to find the structure of the frame and its temporal timing so as to correctly decode the HD header and the PLD field.
    [0029] More specifically, the synchronization means MSYNC performs sliding correlation processing between the filtered digital signal SNF and a reference signal SREF which is in this case a known symbol of the frame, for example a known symbol of the preamble such as the symbol SYNCP.
    [0030] In the example described here, the IND indication representative of the frame structure and of a synchronization performed will for example be the occurrence of the transition between the last symbol SYNCP of the preamble and the symbol SYNCM.
    [0031] This indication IND will be transmitted to the additional processing means MTRS of the processing stage ET2 so as to allow the decoding of the symbols of the header HD and the field PLD of the frame.
    [0032] However, by simply observing the filtered digital signal, it is very difficult, if not impossible, to know if this filtered digital signal results from a noisy useful signal or a noiseless useful signal. However, it is particularly advantageous, as will be seen in more detail below, to know this information so as to further improve the performance of the decoding of the remaining part of the frame. In this regard, the processing stage ET2 comprises verification means MVRF configured to check for the presence or absence of the noise signal in the useful signal, once the synchronization is performed. More specifically, this verification will be performed on the preamble of the unfiltered SNC channel digital signal, and more particularly on one of the symbols of the preamble, for example the unfiltered SYNCP symbol.
    [0033] As illustrated in FIG. 9, a direct FFT fast Fourier transform is carried out in a step 86 so as to perform a time-domain transformation to the frequency domain, and then a power analysis is carried out (step 91). lines of the frequency spectrum obtained at the output of the Fourier transform.
    [0034] In this regard, it is examined in step 92 whether certain frequency lines have a power or a level greater than a fixed threshold TH. If none of the frequency lines of the spectrum has a level above the threshold TH, then it is concluded that there is an absence of noise signal SB in the digital channel signal CNS. In the opposite case, if at least one line has a power greater than the threshold TH, then it is concluded that it is in the presence of a narrowband noise signal SB.
    [0035] And this analysis also makes it possible to know the frequency position of the noise signal, that is to say what are the bins concerned. Materially, as diagrammatically illustrated in FIG. 3, the verification means MVRF comprise MTFD means configured to carry out the direct Fourier transform processing as well as MAL analysis means. In practice, these means can again be made for example in a software manner within a microprocessor.
    [0036] Moreover, as will be seen below, the MTFD means are advantageously those which are already present in the additional processing means MTRS. For further processing, i.e., the decoding of the symbols of the remaining portion of the frame, this will be performed on the unfiltered SNC channel digital signal if it is apparent from the aforementioned verification. that this SNC signal was in fact not noisy by the SB narrowband noise signal. On the other hand, if the verification shows that the noise signal was present, then the processing of the remaining portion of the frame will continue to be performed on the filtered SNF digital signal delivered by the MFL filtering means. In this respect, control means, embodied herein by way of illustration, by a multiplexer MUX controlled by a signal SC delivered by a control module MC connected at the output of the verification means MVRF and representative of the presence or absence of the noise signal, will activate or deactivate the MFL filtering means for the subsequent processing of the remaining part of the frame. The control module can be made for example by a logic circuit or in a software way.
    [0037] More precisely, as illustrated diagrammatically in FIG. 3, in the absence of a noise signal, the digital signal of the SNC channel is delivered directly to the additional processing means MTRS while in the presence of the noise signal SB, it is is the filtered digital signal SNF which is delivered to the additional processing means MTRS. Referring now more particularly to FIG. 10, it can be seen that these complementary processing means MTRS comprise means MCPR configured to remove from each symbol the cyclic prefix, followed by MTFD means configured to perform the fast Fourier transform. direct FFT. The MTFD means are followed by demapping means DMP (demapping means) providing for each carrier a value 10 of the corresponding modulation coefficient (bin). These DMP removal means are followed by an MCSM module configured to determine for each modulation coefficient an indication of confidence (soft decision: "soft decision") of said value. This module is conventional and known per se and uses for example an algorithm of the LogMAP type. The additional MTRS processing means also comprise DINTL deinterleaving means followed by a DCD decoder, for example a Viterbi type decoder, followed by CRC means able to carry out a parity check. The output of the CRC means is connected to the output terminal BS of the MTRS means which is connected to the means forming the MAC layer of the receiver. When the additional processing means MTRS receives the filtered digital signal, that is to say in the presence of a noise signal on the digital signal of the SNC channel, it is particularly advantageous that the indications of confidence (decision soft) associated with the bins on which is present the noise signal as well as those possibly associated with neighboring bins, are set to zero. Indeed, such soft null decisions are seen as being neutral decisions for the error correction algorithm implemented in the Viterbi decoder. This makes it possible to further improve the decoding performance of the de-interlacer-decoder pair. In this regard, the MTRF processing means comprise MFC forcing means configured to perform this forcing to zero.
    [0038] Here again, materially, all the means and modules of the additional processing means MTRS can be realized by software modules within a microprocessor. Moreover, so as not to disturb the decoding of the remaining part of the frame when the MFL filtering means are activated, it is preferable, for the decoding of this remaining part of the frame, to freeze the coefficients of the FIR filter. that is to say, not to update them as and decoding the remaining bet of the frame.
    [0039] In this respect, the control module MC can deliver a gel signal SC1 to the calculation module MCL of the filter coefficients. According to one aspect of the invention, it is thus possible to obtain a noticeable improvement in the performances in the synchronization phases and in the decoding of the symbols of the frame, especially in the presence of narrow-band noise signals which have levels which can up to 50 to 60 dB above the useful OFDM signal level, whereas current standards require robust decoding only in the presence of a noise signal whose level exceeds that of the wanted signal by only 20 dB.
    权利要求:
    Claims (28)
    [0001]
    REVENDICATIONS1. A method of processing an analog channel signal from a transmission channel, the analog channel signal (SAC) being capable of comprising a useful signal (SU) modulated on a subset of a set of available carriers, carrying at least one frame of symbols according to a frame structure and optionally noisy by at least one narrow-band noise signal (SB), the method comprising an analog-to-digital (CAN) conversion of the channel analog signal and a synchronization process comprising a filtering processing (MFL) including a time domain on-the-fly determination of a limited number of coefficients of a predictive filter of a self regressive model of a digital channel signal (CNS) derived from said conversion analog / digital and on-the-fly filtering of the digital channel signal in the time domain by a finite impulse response digital filter whose coefficients are those of the filter e predictive, and detection of at least one indication (IND) for identifying at least one location of said frame structure, from the filtered channel digital signal (SNF) and a reference signal (SREF) .
    [0002]
    2. Method according to claim 1, wherein each symbol comprises a cyclic prefix and the number of coefficients of the filter is less than or equal to a limit number of the order of three quarters, preferably of the order of one half, the length of the cyclic prefix expressed in number of samples.
    [0003]
    3. Method according to claim 1 or 2, wherein the coefficients of the predictive filter are calculated at a calculation frequency of between 2 and 5 times, preferably between 2 and 3 times, the maximum frequency of said digital channel signal (CNS).
    [0004]
    4. Method according to one of the preceding claims, wherein said on-the-fly determination of said coefficients and said on-the-fly filtering comprise a grouping (70) of the samples of said digital signal in successive groups of samples (GR), a 3034274 27 determining (71) a current block of coefficients using the current group of samples, and an application (72) on said current group of the finite impulse response filter having said current block of coefficients so as to obtain a filtered current group 5 samples (GRF).
    [0005]
    5. Method according to one of the preceding claims, wherein each frame comprises a preamble (PRM) comprising known symbols and preceding the remaining part (HD, PLD) of the frame, and the filtering treatment is applied at least to detect said at least one indication at the preamble of at least one frame.
    [0006]
    The method of claim 5, wherein the filtering process is applied at least to detect said at least one indication (IND) at the preamble of each frame.
    [0007]
    The method according to one of the preceding claims, further comprising after detecting said at least one indication, verifying (MVRF) the presence or absence of said noise signal from at least one symbol. useful unfiltered signal.
    [0008]
    8. The method of claim 7, wherein said verifying comprises a Fourier transform processing (90) on said at least one symbol (SYNCP) and analyzing (91) the power of each carrier.
    [0009]
    The method of claim 7 or 8 taken in conjunction with claim 5 or 6, wherein said checking is performed on at least one symbol (SYNCP) of the preamble.
    [0010]
    The method of claim 9, further comprising processing said remaining portion of each frame and wherein in the absence of the noise signal, processing (MTRS) of said remaining portion of the frame is performed on the signal digital channel of unfiltered channel.
    [0011]
    The method of claim 9, wherein each symbol has modulation coefficients respectively associated with said carriers, and the method further comprises processing the remaining portion (MTRS) of the frame performed in case of presence of said signal. noise on the filtered channel digital signal and comprising a direct Fourier transform processing, a demapping process providing for each carrier a value of said modulation coefficient and a determination for each modulation coefficient of a confidence indication of said value and a zero forcing of the confidence indications of the modulation coefficients associated with the carriers whose frequencies correspond to those of the noise signal.
    [0012]
    The method of claim 11, wherein the values of the filter coefficients are frozen during processing of the remaining portion of the frame.
    [0013]
    13. The method as claimed in one of the preceding claims, in which the useful signal is an OFDM modulated signal. 15
    [0014]
    14. Method according to one of the preceding claims, wherein the transmission channel is an electric line (LE) and the useful signal is a line carrying current signal.
    [0015]
    A receiver, comprising an input stage (ET1) for connection on a transmission channel (LE) and configured to output an analog channel signal (SAC) from the transmission channel, the channel analog signal being capable of comprising a useful signal (SU) modulated on a subset of a set of available carriers, carrying at least one frame of symbols according to a frame structure and possibly noisy by at least one noise signal (SB) in narrow band, an analog / digital conversion (CAN) stage for performing an analog / digital conversion of the analog channel signal and a processing stage (ET2) comprising filtering means (MFL) including a configured calculation module (MCL) to determine on the fly a limited number of coefficients of a predictive filter of a self regressive model of a digital channel signal (CNS) from the analog / digital conversion stage, a naked filter finite impulse response (FIR) method whose coefficients are those of the predictive filter, for on-the-fly filtering of the digital time-domain signal and detection means (MSYNC) configured to detect at least one indication (IND) for identifying at least one location of said frame structure from the filtered digital signal (SNF) and a reference signal (SREF). 5
    [0016]
    16. The receiver of claim 15, wherein each symbol comprises a cyclic prefix and the number of coefficients of the filter is less than or equal to a limit number of the order of three quarters, preferably of the order of one half, of the length of the cyclic prefix expressed in number of samples. 10
    [0017]
    17. Receiver according to one of claims 15 or 16, wherein the calculation module (MCL) is configured to calculate the coefficients of the predictive filter at a calculation frequency of between 2 and 5 times, preferably between 2 and 3 times, the maximum frequency of said digital channel signal. 15
    [0018]
    18. The receiver according to one of claims 15 to 17, wherein the filtering means comprise grouping means (MGR) configured to group samples of said digital channel signal into successive groups of samples, the calculation module (MCL). ) is configured to determine a current block of coefficients using the current group of samples, and the digital filter (FIR) is configured to input said current group of samples so as to output a filtered current group of samples .
    [0019]
    Receiver according to one of Claims 15 to 18, in which each frame comprises a preamble (PRM) comprising known symbols and preceding the remaining part of the frame, and the processing stage (ET2) comprises means for control (MC, MUX) configured to output said digital channel signal to the filtering means at least so that the detection means detects said at least one indication at the preamble of at least one frame.
    [0020]
    The receiver of claim 19, wherein the control means (MC, MUX) is configured to output said digital channel signal to the filtering means at least so that the detection means detect said at least one indication at the of the preamble of each frame.
    [0021]
    The receiver according to one of claims 19 to 20, further comprising verification means (MVRF) configured to perform after the detection of said at least one indication (IND), a verification of the presence or absence said noise signal from at least one symbol of the unfiltered useful signal.
    [0022]
    The receiver of claim 21, wherein the verification means comprises a direct Fourier transform (MTFD) stage configured to perform a Fourier transform processing directly on said at least one symbol and analysis means (MAL). ) configured to perform a power analysis of each carrier.
    [0023]
    Receiver according to claim 21 or 22 taken in combination with claim 19 or 20, wherein the verification means (MVRF) are configured to perform said verification on at least one symbol of the preamble.
    [0024]
    The receiver of claim 23, wherein the processing stage (ET2) further comprises additional processing means (MTRS) configured to perform a processing of said remaining portion of each frame and in the event of absence of the signal The control means (MC, MUX) are configured to deliver said remaining portion of the frame directly to the additional processing means (MTRS) without passing through the filtering means (MFL).
    [0025]
    A receiver according to claim 23, wherein each symbol having modulation coefficients respectively associated with the carriers, and the processing stage further comprises additional processing means configured to perform a processing of said remaining portion of each frame, the additional processing means (MTRS) comprising a direct Fourier transform stage (MTFD), a demapping means (DMP) providing for each carrier a value of said modulation coefficient and a module (MCSM) capable of determining for each modulation coefficient an indication of confidence of said value, and forcing means (MFC) configured for, in case of presence of said noise signal, force to zero confidence indications of the modulation coefficients associated with the 5 carriers whose frequencies correspond to those of the noise signal.
    [0026]
    The receiver of claim 25, wherein the control means (MC) is configured to disable the calculation module (MCL) of the filter coefficients during processing of said remaining portion of the frame. 10
    [0027]
    Receiver according to one of Claims 15 to 26, in which the wanted signal is a signal modulated according to an OFDM modulation.
    [0028]
    28. Receiver according to one of claims 15 to 27, wherein, the transmission channel being an electric line (LE), the useful signal is a line carrier current signal.
    类似技术:
    公开号 | 公开日 | 专利标题
    EP3073695B1|2018-01-24|Method for processing an analog signal coming from a transmission channel, in particular a signal transmitted through powerline communication
    EP0580216A1|1994-01-26|System and receiver for orthogonal frequency division multiplexed signals provided with a frequency synchronisation circuit
    FR2857802A1|2005-01-21|Multi-carrier signal propagation channel such as OFDM channel, estimation method for electromagnetic digital broadcasting system, involves correcting reference pilots by calculating error vector of amplitude and/or phase for each pilot
    FR2915840A1|2008-11-07|DECODING SYMBOLS OF A SIGNAL DISTRIBUTED FOLLOWING FREQUENCY AND TIME DIMENSIONS
    EP1810470B1|2017-08-23|Method for receiving a multicarrier signal using at least two estimations of a propagation channel and corresponding reception device
    EP0549445B1|1998-01-28|Method for the transmission of reference signals in a multicarrier data transmission system
    EP3116135A1|2017-01-11|Method for processing a signal from a transmission channel, in particular a signal carried over power line, and in particular the channel estimation, and corresponding receiver
    EP3116136B1|2020-07-08|Method for estimating a time-invariant transmission channel, and corresponding receiver
    FR2742613A1|1997-06-20|Quality factor evaluation method for digital transmission system
    EP3116137B1|2019-06-26|Method for estimating a cyclostationary transmission channel, and corresponding receiver
    WO2000011799A1|2000-03-02|Iterative rake receiver and corresponding reception method
    EP3101814A1|2016-12-07|Method and device for transmitting information on a communication medium having a variable impedance, in particular for plc transmission
    EP1206045A1|2002-05-15|Method of correcting frequency error
    EP3780529A1|2021-02-17|Device and method for communications signalling through a fragmented spectrum
    CA2060413C|1996-07-09|Interfering signal detection method and device for a digital data demodulator
    FR2826208A1|2002-12-20|System and method for transmitting an audio or phonic signal, by utilizing a digital modem for transmitting analogue signals
    EP1438815A1|2004-07-21|Inter-symbol interference canceller
    FR2916920A1|2008-12-05|METHOD FOR RECEIVING A MULTIPLEX FREQUENCY TRANSMITTED SIGNAL.
    FR2753592A1|1998-03-20|Digital burst signal demodulation method e.g. for TDMA system
    WO2019012227A1|2019-01-17|Transmission and reception of an ofdm signal on an electrical two-wire transmission channel of an aircraft
    EP2093889A1|2009-08-26|Method for processing a first and second signal superimposed within an incident aggregate signal and corresponding device.
    FR2675001A1|1992-10-09|Method and device for digital modulation with components in phase and in quadrature and transmission installation including an application thereof
    FR2815794A1|2002-04-26|Device to identify interference signals in a radio electric signal with a constant module
    EP2128998B1|2015-12-16|Method for detecting an OFDM signal
    WO2017098170A1|2017-06-15|Method and device for combining frames of complex symbols
    同族专利:
    公开号 | 公开日
    CN205249269U|2016-05-18|
    FR3034274B1|2017-03-24|
    CN106027439A|2016-10-12|
    CN106027439B|2019-07-23|
    US20170288918A1|2017-10-05|
    US9729199B2|2017-08-08|
    CN110224721B|2021-06-04|
    CN110224721A|2019-09-10|
    US20160285509A1|2016-09-29|
    EP3073695B1|2018-01-24|
    EP3073695A1|2016-09-28|
    US10050672B2|2018-08-14|
    引用文献:
    公开号 | 申请日 | 公开日 | 申请人 | 专利标题
    US20030063596A1|2001-09-19|2003-04-03|Huseyin Arslan|Methods and apparatus for canceling co-channel interference in a receiving system using spatio-temporal whitening|
    DE3811066A1|1988-03-31|1989-10-12|Philips Patentverwaltung|NUCLEAR RESONANCE SPECTROMETER|
    US5673210A|1995-09-29|1997-09-30|Lucent Technologies Inc.|Signal restoration using left-sided and right-sided autoregressive parameters|
    US6590932B1|1999-11-24|2003-07-08|Ericsson Inc.|Methods, receiver devices and systems for whitening a signal disturbance in a communication signal|
    US6314131B1|2000-03-24|2001-11-06|Cadence Design Systems, Inc.|Method and system for channel estimation using multi-slot averaged interpolation|
    SG99310A1|2000-06-16|2003-10-27|Oki Techno Ct Singapore Pte|Methods and apparatus for reducing signal degradation|
    US6594098B1|2000-09-12|2003-07-15|Marvell International Ltd.|Acquistion timing loop for read channel|
    FR2820507B1|2001-02-07|2003-03-28|Onera |REJECTION OF CLOTS IN A PASSIVE RADAR RECEIVER OF OFDM SIGNALS|
    DE10135089A1|2001-07-11|2003-01-23|Siemens Ag|Transducer signal detection/transmission method, e.g. in transformer substation protection, involves converting output of transducers into digital values, which are transmitted as transmit message with reduced sampling rate|
    AU2003302114A1|2002-11-21|2004-06-15|Nippon Telegraph And Telephone Corporation|Digital signal processing method, processor thereof, program thereof, and recording medium containing the program|
    US6868276B2|2003-06-17|2005-03-15|Nokia Corporation|Method and apparatus for estimating carrier frequency offset and fading rate using autoregressive channel modeling|
    CN1529421A|2003-09-28|2004-09-15|欣 耿|Digital demodulating-receiving device|
    CN101610372A|2004-12-20|2009-12-23|三星电子株式会社|Digital multimedia receiver and method of reseptance thereof|
    JP4457041B2|2005-05-20|2010-04-28|株式会社日立製作所|Verify processing method and optical disc apparatus|
    US7746970B2|2005-11-15|2010-06-29|Qualcomm Incorporated|Method and apparatus for filtering noisy estimates to reduce estimation errors|
    JP2007258897A|2006-03-22|2007-10-04|Mitsubishi Electric Corp|Power line carrier communication apparatus, and frame|
    US20100067629A1|2006-12-01|2010-03-18|Koninklijke Philips Electronics, N.V.|Performance in a time-frequency interleaved orthogonal frequency division multiplexing system|
    JP2009089146A|2007-10-01|2009-04-23|Sharp Corp|Communication apparatus, communication system, communication method, communication program, and recording medium|
    US8472576B2|2008-06-06|2013-06-25|Maxim Integrated Products, Inc.|Jammer canceller for power-line communication|
    US8483313B2|2008-11-14|2013-07-09|Qualcomm Incorporated|Transmission suppression|
    US9060222B2|2009-06-11|2015-06-16|Sda Software Design Ahnert Gmbh|Method for determining an averaged frequency-dependent transmission function for a disturbed linear time-invariant system, evaluation device and computer program product|
    CN101662433B|2009-06-23|2013-03-20|中山大学|Channel prediction method based on particle filtration correction|
    CN102098258B|2009-12-14|2013-11-06|中兴通讯股份有限公司|Method for removing narrow-band interference and self-adapting filter|
    US8559485B2|2010-04-08|2013-10-15|Andrew Llc|Autoregressive signal processing for repeater echo cancellation|
    CN101986629B|2010-10-25|2013-06-05|华为技术有限公司|Method and device for estimating narrowband interference as well as receiving equipment thereof|
    JP2012156861A|2011-01-27|2012-08-16|Renesas Electronics Corp|Power line communication device and noise detection method|
    US8743974B2|2011-08-29|2014-06-03|Texas Instruments Incorporated|Adaptive modulation and coding with frame size adjustment for power line communications |
    TWI436616B|2011-12-29|2014-05-01|Ind Tech Res Inst|Communication device capable of channel estimation and method thereof|
    EP2713522A1|2012-09-06|2014-04-02|Andrew LLC|Digital post-distortion compensation in telecommunication systems|
    WO2014037066A1|2012-09-10|2014-03-13|Sony Corporation|Device for power line communication, power line communications system, method for operating a device for power line communication and for adapting plc transmissions to noise|
    CN102868385B|2012-09-25|2015-07-08|北京东方联星科技有限公司|Digital filter capable of configuring coefficient and implementation method|
    EP2750130B1|2012-12-31|2015-11-25|Nxp B.V.|Signal processing for a frequency modulation receiver|
    US9325483B2|2013-03-15|2016-04-26|Wi-Lan Labs, Inc.|Flexible MIMO resource allocation through cross-correlation nulling and frequency domain segmented receiver processing|
    US9413575B2|2013-03-15|2016-08-09|Echelon Corporation|Method and apparatus for multi-carrier modulation packet detection based on phase differences|
    US9537641B2|2013-05-30|2017-01-03|Qualcomm Incorporated|Channel adaptation to compensate for interference from neighbor powerline communication networks|
    CN103532901B|2013-10-14|2017-06-23|国家电网公司|A kind of time synchronization method and system for power line carrier communication|
    FR3034274B1|2015-03-27|2017-03-24|Stmicroelectronics Sas|METHOD FOR PROCESSING AN ANALOGUE SIGNAL FROM A TRANSMISSION CHANNEL, ESPECIALLY AN ONLINE CARRIER CURRENT VEHICLE SIGNAL|FR3034274B1|2015-03-27|2017-03-24|StmicroelectronicsSas|METHOD FOR PROCESSING AN ANALOGUE SIGNAL FROM A TRANSMISSION CHANNEL, ESPECIALLY AN ONLINE CARRIER CURRENT VEHICLE SIGNAL|
    US9838077B2|2015-07-09|2017-12-05|StmicroelectronicsSas|Method for estimating a cyclostationary transmission channel, and corresponding receiver|
    FR3038801B1|2015-07-09|2017-07-21|StmicroelectronicsSas|METHOD OF ESTIMATING A TEMPORALLY INVARIANT TRANSMISSION CHANNEL AND CORRESPONDING RECEIVER|
    FR3038800A1|2015-07-09|2017-01-13|StmicroelectronicsSas|METHOD FOR PROCESSING A SIGNAL FROM A TRANSMISSION CHANNEL, PARTICULARLY AN IN-LINE CARRIER CURRENT VEHICLE SIGNAL, AND ESPECIALLY CHANNEL ESTIMATION, AND CORRESPONDING RECEIVER|
    TWI610545B|2016-06-21|2018-01-01|晨星半導體股份有限公司|Detecting method and detecting device for detecting notch-band|
    US10164684B2|2016-09-09|2018-12-25|Hong Kong Applied Science and Technology Research Institute Company Limited|Interference detection and mitigation in power line communication|
    CN107995133B|2016-10-26|2019-12-13|电信科学技术研究院|Method and device for generating channel frequency and channel evaluation circuit|
    FR3082970B1|2018-06-21|2020-05-29|Zodiac Data Systems|METHOD FOR DEMODULATING A CARRIER SIGNAL USING SEVERAL DIGITAL DEMODULATORS|
    法律状态:
    2016-02-22| PLFP| Fee payment|Year of fee payment: 2 |
    2016-09-30| PLSC| Search report ready|Effective date: 20160930 |
    2017-02-21| PLFP| Fee payment|Year of fee payment: 3 |
    2018-02-20| PLFP| Fee payment|Year of fee payment: 4 |
    2019-11-29| ST| Notification of lapse|Effective date: 20191106 |
    优先权:
    申请号 | 申请日 | 专利标题
    FR1552588A|FR3034274B1|2015-03-27|2015-03-27|METHOD FOR PROCESSING AN ANALOGUE SIGNAL FROM A TRANSMISSION CHANNEL, ESPECIALLY AN ONLINE CARRIER CURRENT VEHICLE SIGNAL|FR1552588A| FR3034274B1|2015-03-27|2015-03-27|METHOD FOR PROCESSING AN ANALOGUE SIGNAL FROM A TRANSMISSION CHANNEL, ESPECIALLY AN ONLINE CARRIER CURRENT VEHICLE SIGNAL|
    EP15190536.1A| EP3073695B1|2015-03-27|2015-10-20|Method for processing an analog signal coming from a transmission channel, in particular a signal transmitted through powerline communication|
    CN201520975374.5U| CN205249269U|2015-03-27|2015-11-30|Receiver|
    CN201510860006.0A| CN106027439B|2015-03-27|2015-11-30|To the method and receiver handled from the channel analog signal for sending channel|
    CN201910461075.2A| CN110224721B|2015-03-27|2015-11-30|Method and receiver for processing an analog signal from a transmission channel|
    US14/984,966| US9729199B2|2015-03-27|2015-12-30|Method for processing an analog signal coming from a transmission channel, in particular a signal carried by power line communications|
    US15/629,257| US10050672B2|2015-03-27|2017-06-21|Method for processing an analog signal coming from a transmission channel, in particular a signal carried by power line communications|
    [返回顶部]