US20010001007A1 - Methods and device for estimating and correcting clipping in a discrete multi-tone communications system - Google Patents
Methods and device for estimating and correcting clipping in a discrete multi-tone communications system Download PDFInfo
- Publication number
- US20010001007A1 US20010001007A1 US09/729,008 US72900800A US2001001007A1 US 20010001007 A1 US20010001007 A1 US 20010001007A1 US 72900800 A US72900800 A US 72900800A US 2001001007 A1 US2001001007 A1 US 2001001007A1
- Authority
- US
- United States
- Prior art keywords
- signal
- signals
- subchannels
- clipping
- data
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Images
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L27/00—Modulated-carrier systems
- H04L27/26—Systems using multi-frequency codes
- H04L27/2601—Multicarrier modulation systems
- H04L27/2647—Arrangements specific to the receiver only
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L27/00—Modulated-carrier systems
- H04L27/26—Systems using multi-frequency codes
- H04L27/2601—Multicarrier modulation systems
- H04L27/2614—Peak power aspects
- H04L27/2623—Reduction thereof by clipping
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L5/00—Arrangements affording multiple use of the transmission path
- H04L5/003—Arrangements for allocating sub-channels of the transmission path
- H04L5/0044—Arrangements for allocating sub-channels of the transmission path allocation of payload
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L5/00—Arrangements affording multiple use of the transmission path
- H04L5/003—Arrangements for allocating sub-channels of the transmission path
- H04L5/0058—Allocation criteria
- H04L5/006—Quality of the received signal, e.g. BER, SNR, water filling
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L5/00—Arrangements affording multiple use of the transmission path
- H04L5/0001—Arrangements for dividing the transmission path
- H04L5/0003—Two-dimensional division
- H04L5/0005—Time-frequency
- H04L5/0007—Time-frequency the frequencies being orthogonal, e.g. OFDM(A), DMT
Definitions
- the invention relates generally to digital information transmission systems and specifically to techniques for controlling signal clipping over a discrete multi-tone channel such as an asymmetric digital subscriber line.
- Digital Subscriber Line (“DSL”) modems provide high-speed communication links over existing copper twisted-pair wiring connecting the telephone company central office to the residential customer. With one DSL modem in the CO and one in the home, a customer can connect to the backbone data network (e.g. the Internet) and achieve data throughput as high as 8 Mbits/second. Such a high data rate can facilitate new Internet applications, entertainment, education, etc.
- backbone data network e.g. the Internet
- ADSL Asymmetric Digital Subscriber Line
- ADSL Discrete Multi-Tone
- DMT Discrete Multi-Tone
- the line code is a multicarrier transmission method that partitions the available transmission bandwidth into many narrow bandwidth subchannels over which parallel data streams are transmitted. Because the different subchannels are affected differently by the transmission channel, the quality of the individual subchannels can be estimated and the data density of each subchannel can be adjusted to achieve maximum throughput.
- ADSL modems apply a large amount of digital and analog signal processing to achieve data throughput that can be very near the theoretical channel capacity. Because of the Gaussian nature of DMT signals, it is necessary to support very high peak-to-average ratio (“PAR”) values to prevent distortion and ensure high performance. However, to reduce modem complexity and cost, DMT modems are often designed to support lower PAR values. This savings in complexity often causes clipping of the DMT signal, which can reduce performance.
- Clipping of the signal, or voltage waveform occurs when large amplitude values are reduced to some maximum value that is less than the original value. This might occur when the signal contains amplitude values that exceed the maximum value supported by a transmitter circuit or algorithm, and hence are limited to the maximum allowed value.
- RS Reed-Solomon
- interleaving provides the greatest protection against clipping but requires a large amount of memory and causes a long delay and high latency that is unacceptable for certain applications.
- RS coding without interleaving removes latency problems but also provides less protection against clipping. It is desirable to altogether avoid implementation of RS decoding at the receiver for certain applications to reduce complexity and cost.
- the ADSL standard allows the RS encoding to be turned off by setting the number of added parity bytes to zero. If another method of mitigating clipping effects is employed, the RS decoding operation can be eliminated without sacrificing performance. The elimination of the RS decoding operation allows implementation of the ADSL standard on a wider variety of general purpose programmable platforms.
- the present invention comprises a method and device for dealing with clipping problems wherein the receiver compensates for standards compliant transmitted signals using a relatively low-complexity Analog Front End (“AFE”).
- AFE Analog Front End
- a clipping estimation and correction method that operates within the constraints of the T1.413 ADSL standard and eliminates the dependence on Reed-Solomon decoding and interleaving for adequate performance when the transmitter clips the transmitted signal.
- the method involves the steps of measuring the signal to noise ratio (“SNR”) in all operating subchannels of the transmission channel; transmitting the SNR measurements for the operating subchannels to a transmitting device on the channel; communicating data over the operating subchannels using the SNR measurements to affect the throughput of data transmitted over the channel; exploiting the lowered throughput of the transmitted data to reliably compute an estimate of the clipping that occurred in the transmitter; applying the estimate of the clip to construct a compensation signal; and adding the compensation signal at the receiver to correct the clip that occurred at the transmitter.
- SNR signal to noise ratio
- the method can also include the steps of initializing the transmission subchannels by forcing fewer bits than they are capable of carrying and distinguishing DMT frames that have been clipped in the transmitter to make a threshold comparison that determines when a clip has occurred. If a clip has occurred, correction is applied. Otherwise, the received data needs no correction and is passed on to the user without modification.
- the method comprises the steps of transmitting a signal on the DSL system; analyzing the signal to determine if it has been clipped during transmission; and computing an error compensation signal for the signal if it is determined it has been clipped during transmission.
- a method of estimating the amplitude and location of a clip involves the steps of initializing the subchannels to create a compensation signal and adding the compensation signal in the receiver to enhance performance during decoding of the clipped frames.
- the clipping estimation and correction methods of the present invention alleviate the need to compensate for transmitter clipping effects using forward error-correction coding.
- the methods of the present invention can be implemented on any modem platform, even one that does not support Reed-Solomon decoding.
- a general purpose programmable modem that implements the T1.413 standard without Reed-Solomon decoding capabilities.
- the modem supports T1.413 standard transmission/reception functions over a discrete multi-tone communications system and can be linked to an independent processing system such as a desktop or notebook computer.
- the modem comprises an interface having an analog side and a digital side with the analog side coupled to the communications system.
- the modem contains a signal processing device for transmitting and receiving signals to and from the communications system.
- the signal processing device has a system side and a user side with the system side coupled to the digital side of the interface.
- the interface is coupled to the user side of the signal processing device and forms a communications pathway between the independent processing system and the ADSL modem.
- the modem incorporates the clipping error estimation and correction functions of the present invention.
- the ADSL modem contains stored program segments capable of causing the signal processing circuit to analyze original signals received from the communications system to determine if they have been clipped during transmission; compute error compensation signals for signals that are clipped during transmission; and add computed error compensation signal to the original signals received to obtain corrected analog signals.
- FIG. 1 is a diagram illustrating a communications system wherein the invention can be practiced
- FIG. 2 shows the functions required for transmission of data from one ADSL modem to another
- FIG. 3 is a block diagram for a receiver architecture that implements the clipping estimation and correction methods according to one embodiment
- FIG. 4 is a process flow diagram illustrating a method for detecting the need for fixing a clip and estimating the clipping signal according to one embodiment
- FIG. 5 shows a frame of data generated by an ADSL DMT transmitter
- FIG. 6 shows the frame of data of FIG. 4 after clipping
- FIG. 7 shows the impulse response of the DSL channel used in an ADSL simulation
- FIG. 8 shows the subchannel SNR of a single received data frame for each of the 255 subchannels.
- FIG. 1 illustrates an example communications system 10 in which the invention can be practiced according to one embodiment.
- the communications system 10 includes a service provider 40 that is accessible by one or more subscribers 15 .
- the subscriber 15 has the transmission/reception and data processing equipment enabling access to the service provider 40 .
- a first processing system 12 is operably coupled to a first modem 16 via interface 14 .
- the interface 14 provides a communications pathway for unmodulated data transfers between the first modem 16 and the first processing system 12 .
- the first modem 16 could be internally fixed inside the first processing system 12 and coupled through a standard interface of the first processing system 12 .
- Example interface standards employed between the modem 16 and the first processing system 12 include EISA, PCMCIA, RS-232 and other industry accepted interface protocols.
- the first processing system 12 communicates digitally with the first modem 16 .
- Data from the first processing system 12 is transferred through the interface 14 and maintained in a buffering mechanism such as an internal memory space (ROM or RAM) or other similar configuration inside the first modem 16 .
- a universal asynchronous receiver transmitter (“UART”) or other similar device is used to handle data flow between the first processing system 12 and the modem 16 .
- the first processing system 12 can be a desktop or notebook computer, workstation or other similar computing device.
- the first modem 16 includes a signal processing device for converting digital data from the processing system 12 to an equivalent modulated analog waveform.
- the modem also includes an interface such as an analog front end (“AFE”) which couples the analog signals carrying the data to the communications link 18 to the central office 24 .
- AFE analog front end
- the transmit and receive functions of the modem 16 are operated by the processing system 12 .
- a digital-to-analog/analog-to-digital conversion circuit or other similar signal processing device can be used to transmit and receive analog signals via the communications link 18 .
- the communications link 18 can be analog twisted pair (“ATP”) wiring or other medium of the type often used in public switched telephone networks (“PSTN”).
- ATP analog twisted pair
- PSTN public switched telephone networks
- the digital data from the processing system 12 is converted to analog equivalent signals which are transmitted on the ATP wiring 18 using known modulation methods. Examples of such modulation methods include Quadrature Amplitude Modulation (“QAM”), Trellis Code Encoding (“TCE”) and Frequency Shift Keying (“FSK”) among others.
- modulation methods include Quadrature Amplitude Modulation (“QAM”), Trellis Code Encoding (“TCE”) and Frequency Shift Keying (“FSK”) among others.
- the service provider 40 receives the signals from the far end central office 28 the digital backbone connection 30 .
- the data is passed to the second processing system 36 at the service provider location 40 .
- the processing system 36 is employed by the service provider 40 to gain access to the central office 28 and to format, receive and transmit data to other entities such as subscriber 15 .
- Both processing systems 12 and 36 contain the necessary processing and storage capabilities and run suitable application programs for such functions.
- an ADSL modem on the CPE unit 18 side communicates with another ADSL modem 22 maintained by the central office 24 .
- multiple transmission channels can be maintained: a high speed downstream channel, a medium speed duplex channel, and a Plain Old Telephone Service (“POTS”) channel.
- POTS Plain Old Telephone Service
- the POTS channel can be split off from the ADSL modem by filters, thus guaranteeing uninterrupted POTS, even if ADSL fails.
- the high speed channel ranges from 1.5 to 6.1 Mbps, while duplex rates range from 16 to 640 kbps.
- Each channel can be submultiplexed to form multiple lower rate channels. Still other data rates may be employed as may be standardized in the industry.
- Downstream data rates depend on a number of factors, including the length of the copper line, its wire gauge, presence of bridged taps, and cross-coupled interference. Line attenuation increases with line length and frequency, and decreases as wire diameter increases.
- the communications system 10 supports both downstream and upstream communications.
- the modems 16 and 22 send probing tones via the analog line 18 .
- the central office 24 comprises the call switching and routing equipment necessary to complete a path between the subscriber 15 and the service provider 40 .
- the functionality of the subscriber 15 and the service provider 40 can be implemented using known methods and devices.
- the signaling protocols used by the various devices on the system 10 may include those supported by the International Standard Organization (“ISO”), the International Brass and Telephone Consultative Committee (“CCITT”) and the Electronics Industries Association (“EIA”) among other standards bodies.
- ISO International Standard Organization
- CITT International Telegraph and Telephone Consultative Committee
- EIA Electronics Industries Association
- data from the first processing system 12 is passed to the first modem 16 , modified appropriately with signal processing techniques to prepare the data for robust transmission, and converted using a digital to analog convertor or other similar conversion device.
- the modem 16 transmits the converted digital stream using a modulated analog signal carrier 20 over analog twisted pair 18 to the central office 24 .
- another DSL modem 22 decodes the data and ports it to the digital backbone network 26 .
- the digital backbone network 26 connects the central offices 24 and 28 .
- the central office 28 receives the digital data from the backbone 26 and port it on the digital link 30 the service provider's processing platform 36 .
- the near end central office 24 communicates with the far end central office 28 via the digital backbone network 26 which is often a high speed digital link providing a high data rate.
- the analog signals 18 are received by the second modem 22 and demodulated to obtain the original data sequence generated by the first processing system 12 .
- the second modem 22 passes the digital bit stream sequence to the digital backbone network 26 and on the far end central office 28 .
- processing systems 12 and 36 communicate with each other through communications facilities comprised of communications mediums 16 , 18 , 22 , 24 , 26 , 28 , 30 and 32 .
- the service provider 40 can be an Internet service provider or dedicated data service provider of the type available in industry.
- the modems 16 annd 22 comprise the data communication equipment (“DCE”) responsible for providing the required translation and interface between the digital and analog signals carried on the system 10 .
- DCE data communication equipment
- the modems 16 and 22 are designed to transmit and receive analog signals tuned to one or more analog carrier frequencies.
- the data from the processing systems 12 is superimposed upon one or more carrier frequencies and transmitted on the system 10 .
- the data signals is often referred to as the base band signal.
- the carrier signal 18 is changed back to the digital signal at the receiving modem 22 through the process of demodulation.
- the modems 16 and 22 provide the transmit and receive functions for the binary bits of digital data generated by the processing system 12 at the subscriber location 15 .
- FIG. 2 is a process flow diagram 50 illustrating the functions required for transmission of data from one Discrete Multi-Tone (“DMT”) modem to another.
- DMT Discrete Multi-Tone
- the functions are implemented on a pair of T1.413 compliant Asymmetric Digital Subscriber Loop (“ADSL”) modems.
- ADSL Asymmetric Digital Subscriber Loop
- each modem implements both transmit 52 and receive 100 functions.
- the methods of the present invention can be implemented on any modem platform, even one that does not support Reed-Solomon decoding.
- a general purpose programmable modem that implements the T1.413 standard is disclosed.
- the modem supports T1.413 standard transmission/reception functions over a discrete multi-tone communications system and can be linked to an independent processing system 12 such as a desktop or notebook computer.
- the modem comprises an interface having an analog side and a digital side with the analog side coupled to the communications system 10 .
- a signal processing device for transmitting and receiving signals to and from the communications system 10 where the signal processing device is coupled to the digital side of the interface.
- the user interface formings a communications pathway between the independent processing system 12 and the ADSL modem.
- the modem incorporates the clipping error estimation and correction functions of the present invention.
- the ADSL modem contains stored program segments capable of causing the signal processing circuit to analyze original signals received from the communications system to determine if they have been clipped during transmission and compute error compensation signals for signals that are clipped during transmission. The modem can then add the computed error compensation signal to the original signals received to obtain corrected analog signals.
- the transmit path 54 shown is in one direction although an actual bidirectional ADSL communications link between modems extends in both directions .
- Reed-Solomon (“RS”) forward error-correction encoding 56 can be applied to both fast bits 54 and slow bits 55 in the transmitter.
- the RS-encoded slow bits 57 are interleaved 60 to provide extra resistance to clipping and channel error. Thus, performance is improved at the expense of additional encoding delay.
- the RS-encoded fast bits 59 are combined with the RS-encoded and interleaved slow bits 61 in a frame of bits that is converted into a frame of two-dimensional transmission symbols 62 .
- the transmission signals represent information in a form that is conducive to robust transmission over an ADSL communications link.
- the Inverse Discrete Fourier Transform (“IDFT”) 64 is applied to the frame of transmission symbols 63 after appropriate conjugate mirroring of the frame according to the T1.413 standard.
- the conjugate symmetry enforced on the IDFT input 63 causes the output 65 to be completely real.
- the REAL ⁇ operator 66 indicates that the imaginary component of the IDFT output 65 is not used.
- a cyclic prefix is created from the last samples of the frame of real data and concatenated 68 to the beginning of the frame.
- the resulting frame of data is passed through a Digital-to-Analog Converter (“DAC”) 70 and a transmit filter 72 that provides the low-pass and high-pass filtering functions.
- DAC Digital-to-Analog Converter
- a line driver circuit 74 can be used to provide the current and/or voltage gain required to place the filtered signal onto the twisted pair telephone line with an ADSL format with the appropriate power level.
- the signal After transmission through the channel 76 , the signal passes through receive buffers and filters 78 and into the Analog-to-Digital Converter (“ADC”) 80 .
- the effective analog channel 82 seen by the transmission system is formed by the analog portion of the DAC 70 , the transmit filters 72 , the line driver 74 , the actual DSL channel 76 , the receive buffers and filters 78 , and the analog portion of the ADC 80 .
- a Time-domain Equalizer (“TEQ”) is applied 84 to the signal to shorten the time-domain impulse response of the entire analog channel.
- TEQ Time-domain Equalizer
- the combined analog channel 82 and TEQ 84 form an effective “shortened” channel whose impulse response has energy concentrated in a small number of samples. Preferably, significant energy is concentrated in the same number of samples as are present in the cyclic prefix 68 .
- a frame of data is formed and the cyclic prefix is removed 86 .
- the Discrete Fourier Transform (“DFT”) 88 can then be applied to the signal.
- DFT Discrete Fourier Transform
- FEQ Frequency-domain Equalization
- the frame of FEQ output symbols is decoded into bits 92 .
- the slow bits are passed through a de-interleaver 94 which spreads out transmission errors over many frames. Both fast and slow bits are passed through RS decoders 96 , 98 to correct transmission errors and/or errors caused by clipping in the transmitter.
- FIG. 2 shows the possible locations where clipping can occur in the transmitter (indicated by the “ ⁇ ”).
- the location ⁇ at which a clip occurs typically depends on the design of the transmitter.
- the method of estimating the clipping distortion and applying clipping compensation of the present invention works within the constraints of the T1.413 standard thus guaranteeing compatibility between a low-complexity modem implemented without RS decoding capabilities and the large possible range of T1.413 compliant modems that may be employed.
- the Signal-to-Noise (“SNR”) is measured in all the operating subchannels of the receiver.
- the SNR values are then communicated back to the transmitter so that the transmitter can properly allocate bits to each subchannel to maximize the transmission throughput.
- a small subset of the operating subchannels are forced to carry fewer bits than they are capable of supporting. This provides the receiver with the ability to estimate the position and amplitude of clipping errors. For example, during initialization the receiver can inform the transmitter that subchannels 33 , 34 , 35 , and 36 have a low SNR and should only carry two bits each. This results in a symbol constellation composed of four points (“4-QAM”) being used for subchannels 33 , 34 , 35 and 36 .
- the 4-QAM result applies to the case of T1.413 transmission where the Trellis encoding is not employed. Similar constraints hold true when Trellis encoding is applied, i.e. two bits per subchannel will still result in a low-density constellation.
- the receiver can reliably estimate that the transmitted symbol was the constellation point nearest the received symbol. A reliable estimate of the transmission error can then be computed as the difference between the received signal and the closest constellation point.
- the use of a reliable estimate of the transmission error in a given frame is provided within the T1.413 standard.
- the reliable estimate facilitates the methods for computing a clipping compensation signal at the receiver as described herein.
- FIG. 3 a block diagram of a receiver according to one embodiment is shown and denoted generally as 110 .
- the receiver is configured to perform the clipping estimation and correction functions.
- the received signal 112 from the channel is passed through the TEQ 84 .
- the cyclic prefix is removed 86 prior to the DFT function 88 being applied.
- the FEQ 90 corrects channel distortion as in a normal DMT receiver with a portion of the frame output from the FEQ being analyzed.
- the tones which were made to carry fewer bits are made available to the receiver for computing an error correction signal E[k]. If analysis of the incoming signal 112 indicates a need for clipping correction, then the correction signal (E[k]) is computed 114 and added 116 to the FEQ output 112 .
- the analysis of the incoming signal 112 can be accomplished in a variety of ways. For example, in one embodiment a buffer 120 is used to indicate that the correction signal E[k] is added 116 back to the frame from which the clipping parameters were estimated. The clipping-corrected frame of symbols are then passed through the symbol decoder 92 to produce the transmitted bits.
- FIG. 4 shows a preferred embodiment of the method 150 used for detecting the need for fixing a clip and then estimating the clipping signal.
- the clipping signal allows the receiver to construct a correction signal E[k] if needed.
- the FEQ output 112 is analyzed, and the SNR is computed at step 156 for the tones which were forced to carry less bits than they are capable of carrying.
- the SNR is compared to a stored threshold, step 158 . If the SNR is above the threshold, then no significant clipping occurred. Also, no clipping compensation should be applied and process flow is directed to step 160 .
- the compensation signal (E[k]) is set to zero for all values (k) as shown in step 160 . This results in no clipping compensation being applied. Alternatively, the step of adding the compensation signal can be avoided and there is no need to add zeros.
- the clipping error e[n] can be defined as the difference between the unclipped signal and the clipped signal in the transmitter.
- the frequency-domain error signal at the receiver can be used to compute the amplitude, step 164 , and phase, step 162 , of the time-domain clipping error.
- an estimate is made based on the premise that at least one clip occurred in the transmitter.
- a model for the clipping error e[n] can be devised as an impulse of amplitude ( ⁇ ) and delay (d) as shown below in Equation 1:
- Equation 2 The DFT of the clipping error is therefore computed as shown below in Equation 2:
- the absolute value of the frequency-domain error signal is equal to the amplitude of the clipping error alpha ( ⁇ ).
- the phase of the frequency-domain error signal can be used to compute the delay (d) as shown in step 164 .
- the estimated values of the delay (d) and alpha ( ⁇ ) specify the location and value of the time-domain clipping error [n].
- step 166 to enhance the quality of the clipping compensation signal, additional operations can be applied to verify and possibly correct the position estimated delay (d).
- an estimate of the clipped DMT signal at the transmitter can be reconstructed by applying an IDFT operation, step 168 , to the FEQ output 112 .
- the most probable position of the clipping error can be analyzed to see if the time-domain sample is near the amplitude rails.
- the amplitude rails comprise the positive and negative signal limits of the system 10 as shown in FIGS. 5 and 6.
- step 168 involves a small search of the samples on either side of the most probable clip location. If the sample values are found in the search that are closer to the rail value than the sample at the estimated clipping location, the estimate of the clipping location is changed to the location of the point nearest the rails.
- the IDFT can be computed to obtain the required information, e.g. the IDFT at the estimated clipping location and two points on either side of that can be computed for a total of five IDFT points.
- the clipping compensation signal E[k] can be computed at step 170 . While E[k] can be generated as a time-domain signal, some advantages are provided by directly computing the frequency-domain compensation signal. Thus, the compensation signal E[k] can be directly added to the FEQ output 112 and further decoding operations can then be applied. This approach is preferred in some applications since it results in low complexity of the receiver.
- FIG. 5 shows a frame of data 200 generated by an ADSL DMT transmitter.
- the dashed lines 202 , 204 represent the maximum amplitude levels supported in the transmitter.
- a sample point 206 that exceeds the maximum allowable signal level levels for the system 10 .
- FIG. 6 shows the same frame of data 200 after clipping wherein the new largest value 208 fitting within the rails.
- the signal 200 has been clipped 208 to an amplitude value that is contained within the limited values supported by the ADSL modem.
- the most probable location where the signal was clipped are in the IDFT operation 64 , the DAC 70 , the transmit filter 72 and the line driver 74 .
- FIG. 7 shows the response curve 210 of the DSL channel used in the simulation. Note that the impulse response 212 of the ADSL channel overlaps with the voice band which extends from 20 Hz to 3.4 KHz.
- FIG. 8 shows the subchannel SNR of a single received data frame for each of the 255 subchannels.
- the original SNR 220 is shown by the dashed line.
- the enhanced signal 222 is shown by the solid line. Because of the significant clip at the transmitter, the subchannel SNR is limited to about 36 dB.
- the SNR after enhancement of signal 222 results from application of the clipping estimation and correction method 150 according to the present invention. As before, overlap of the response signal from the ADSL channel is shown overlapping the voice band.
- the SNR at the receiver after enhancement with the clipping estimation and correction method 222 has been increased dramatically.
- the original and enhanced SNR are roughly equivalent.
- the original system SNR is very low.
- the optimal threshold to use in the algorithm is determined by the quality of the transmission channel which, in turn, is determined by line length and interference characteristics.
- the error signal can be estimated more accurately when the tones are analyzed to estimate the clip parameters having high SNR.
- the optimal threshold can be computed based on performance parameters required for a specific application and a particular modem.
- a correction method is used wherein a threshold is set so that the algorithm is not applied if the clip is judged to be small enough that the algorithm would not perform well.
- the clipping estimation and correction method 150 is applied to correct multiple clips with a slight modification. For example, to correct two clips, the model of the clip at the transmitter is changed to:
- Equation 3 allows a correction signal to be computed for two clips of different amplitudes and delays.
- the analysis of clipping error may be performed more carefully to properly estimate the parameters of the equation.
- more tones are required for adequate estimation of the parameters. At the minimum, two tones are used in the estimation procedure to determine the four parameters.
- the transmitter clipping occurs at the IDFT operation 64 , the DAC 70 , or the early stages of the transmit filter 72 since a clip that occurs at the line driver 74 is more difficult to correct since it is a nonlinear response that occurs between components of the overall analog channel 82 .
- a modem can be designed accordingly to guarantee the location of the clipping using known designs.
- the clip signal is modified to represent the analog channel with two components.
- the first component is the, part of the channel before the clip.
- the second component is the part of the channel after the clip.
- a clip that occurs at the line driver 74 can be represented as a delayed and shifted impulse subtracted from the signal after the first portion of the channel.
- the first portion of the analog channel 82 is constructed at the receiver.
- the FEQ 90 reverses the effect of which includes the first portion.
- the FEQ 90 output frame is multiplied by the frequency response of the first portion of the analog channel.
- the analysis and clipping compensation signal is computed as before.
- the resulting compensation signal E[k] can be multiplied by the inverse of the frequency response of the first portion of the analog channel 82 before being added to the FEQ output 112 . Essentially, this estimates the delayed and scaled impulse that occurs in the line driver 74 and computes a compensation signal E[k] that appears as though the clip has traveled through only the second half of the transmission channel 82 as opposed to the entire transmission channel 82 .
Abstract
A method for obtaining a reliable estimate of the transmitter clipping error compliant with T1.413 ADSL standard is disclosed. An architecture is disclosed that uses the clipping error estimate at the receiver to reconstruct a frequency-domain compensation signal. The method for computing the compensation signal is disclosed along with an asymmetric digital subscriber line modem supporting T1.413 standard transmission/reception functions over a discrete multi-tone communications system capable of estimating clipping errors and computing clipping compensation signals.
Description
- 1. The invention relates generally to digital information transmission systems and specifically to techniques for controlling signal clipping over a discrete multi-tone channel such as an asymmetric digital subscriber line.
- 2. The background is described in connection with an asymmetric digital subscriber line communications system, although the principles disclosed may be applied to other multi-tone or multi-carrier systems.
- 3. Digital Subscriber Line (“DSL”) modems provide high-speed communication links over existing copper twisted-pair wiring connecting the telephone company central office to the residential customer. With one DSL modem in the CO and one in the home, a customer can connect to the backbone data network (e.g. the Internet) and achieve data throughput as high as 8 Mbits/second. Such a high data rate can facilitate new Internet applications, entertainment, education, etc.
- 4. Asymmetric Digital Subscriber Line (“ADSL”) signaling was established as the ANSI T1.413 standard described in “T1.413-1995: Telecommunications—Asymmetric Digital Subscriber Line (ADSL) metallic interface”, 1995, incorporated herein by reference. Discrete Multi-Tone (“DMT”) modulation was chosen as the transmission method for ADSL. DMT is often called the line code and is a multicarrier transmission method that partitions the available transmission bandwidth into many narrow bandwidth subchannels over which parallel data streams are transmitted. Because the different subchannels are affected differently by the transmission channel, the quality of the individual subchannels can be estimated and the data density of each subchannel can be adjusted to achieve maximum throughput.
- 5. Although the performance of a DMT system can be extraordinary, the complexity of the standard has precluded timely deployment of ADSL modems. As such, designs for simplifying modem implementation within the constraints of the ADSL standard are of significant merit.
- 6. ADSL modems apply a large amount of digital and analog signal processing to achieve data throughput that can be very near the theoretical channel capacity. Because of the Gaussian nature of DMT signals, it is necessary to support very high peak-to-average ratio (“PAR”) values to prevent distortion and ensure high performance. However, to reduce modem complexity and cost, DMT modems are often designed to support lower PAR values. This savings in complexity often causes clipping of the DMT signal, which can reduce performance.
- 7. Clipping of the signal, or voltage waveform, occurs when large amplitude values are reduced to some maximum value that is less than the original value. This might occur when the signal contains amplitude values that exceed the maximum value supported by a transmitter circuit or algorithm, and hence are limited to the maximum allowed value.
- 8. In ADSL modems, some of the degradation caused by clipping is mitigated by forward error correction in the form of Reed-Solomon (“RS”) coding, or RS coding combined with interleaving. RS coding combined with interleaving provides the greatest protection against clipping but requires a large amount of memory and causes a long delay and high latency that is unacceptable for certain applications. RS coding without interleaving removes latency problems but also provides less protection against clipping. It is desirable to altogether avoid implementation of RS decoding at the receiver for certain applications to reduce complexity and cost.
- 9. The ADSL standard allows the RS encoding to be turned off by setting the number of added parity bytes to zero. If another method of mitigating clipping effects is employed, the RS decoding operation can be eliminated without sacrificing performance. The elimination of the RS decoding operation allows implementation of the ADSL standard on a wider variety of general purpose programmable platforms.
- 10. Prior methods have been proposed for solving the DMT clipping problem. They typically involve application of complicated signal processing algorithms in the transmitter. In general, a DMT transmitter might offset the impairments generated in the Analog Front End (“AFE”) by applying techniques to prevent clips in the AFE. However, most of these techniques require special coordination of actions by the receiver that does not fit within the confines of the ADSL standard.
- 11. As such, a technique is needed that fits within the ADSL standard and can be implemented at the receiver to allow performance gains with any ADSL standard modem.
- 12. The present invention comprises a method and device for dealing with clipping problems wherein the receiver compensates for standards compliant transmitted signals using a relatively low-complexity Analog Front End (“AFE”).
- 13. According to one embodiment, a clipping estimation and correction method is disclosed that operates within the constraints of the T1.413 ADSL standard and eliminates the dependence on Reed-Solomon decoding and interleaving for adequate performance when the transmitter clips the transmitted signal. The method involves the steps of measuring the signal to noise ratio (“SNR”) in all operating subchannels of the transmission channel; transmitting the SNR measurements for the operating subchannels to a transmitting device on the channel; communicating data over the operating subchannels using the SNR measurements to affect the throughput of data transmitted over the channel; exploiting the lowered throughput of the transmitted data to reliably compute an estimate of the clipping that occurred in the transmitter; applying the estimate of the clip to construct a compensation signal; and adding the compensation signal at the receiver to correct the clip that occurred at the transmitter.
- 14. The method can also include the steps of initializing the transmission subchannels by forcing fewer bits than they are capable of carrying and distinguishing DMT frames that have been clipped in the transmitter to make a threshold comparison that determines when a clip has occurred. If a clip has occurred, correction is applied. Otherwise, the received data needs no correction and is passed on to the user without modification.
- 15. Also disclosed is a method of compensating for clipping of signals transmitted over a digital subscriber line (“DSL”) system wherein the signals are modulated using a multi-carrier transmission signaling technique over multiple subchannels of the DSL system and the signal to noise ratio (“SNR”) of the subchannels vary from high to low. The method comprises the steps of transmitting a signal on the DSL system; analyzing the signal to determine if it has been clipped during transmission; and computing an error compensation signal for the signal if it is determined it has been clipped during transmission.
- 16. In another embodiment, a method of estimating the amplitude and location of a clip is disclosed. The method involves the steps of initializing the subchannels to create a compensation signal and adding the compensation signal in the receiver to enhance performance during decoding of the clipped frames.
- 17. The clipping estimation and correction methods of the present invention alleviate the need to compensate for transmitter clipping effects using forward error-correction coding. Thus, the methods of the present invention can be implemented on any modem platform, even one that does not support Reed-Solomon decoding.
- 18. According to another embodiment, a general purpose programmable modem that implements the T1.413 standard without Reed-Solomon decoding capabilities is disclosed. The modem supports T1.413 standard transmission/reception functions over a discrete multi-tone communications system and can be linked to an independent processing system such as a desktop or notebook computer. The modem comprises an interface having an analog side and a digital side with the analog side coupled to the communications system. The modem contains a signal processing device for transmitting and receiving signals to and from the communications system. The signal processing device has a system side and a user side with the system side coupled to the digital side of the interface. The interface is coupled to the user side of the signal processing device and forms a communications pathway between the independent processing system and the ADSL modem.
- 19. The modem incorporates the clipping error estimation and correction functions of the present invention. In particular, according to one embodiment, the ADSL modem contains stored program segments capable of causing the signal processing circuit to analyze original signals received from the communications system to determine if they have been clipped during transmission; compute error compensation signals for signals that are clipped during transmission; and add computed error compensation signal to the original signals received to obtain corrected analog signals.
- 20. Other aspects of the invention including specific implementations are understood by reference to the following detailed description taken in conjunction with the appended drawings in which:
- 21.FIG. 1 is a diagram illustrating a communications system wherein the invention can be practiced;
- 22.FIG. 2 shows the functions required for transmission of data from one ADSL modem to another;
- 23.FIG. 3 is a block diagram for a receiver architecture that implements the clipping estimation and correction methods according to one embodiment;
- 24.FIG. 4 is a process flow diagram illustrating a method for detecting the need for fixing a clip and estimating the clipping signal according to one embodiment;
- 25.FIG. 5 shows a frame of data generated by an ADSL DMT transmitter;
- 26.FIG. 6 shows the frame of data of FIG. 4 after clipping;
- 27.FIG. 7 shows the impulse response of the DSL channel used in an ADSL simulation; and
- 28.FIG. 8 shows the subchannel SNR of a single received data frame for each of the 255 subchannels.
- 29. References in the figures correspond to like numerals in the detailed description unless otherwise indicated.
- 30.FIG. 1 illustrates an
example communications system 10 in which the invention can be practiced according to one embodiment. Thecommunications system 10 includes aservice provider 40 that is accessible by one ormore subscribers 15. Generally, thesubscriber 15 has the transmission/reception and data processing equipment enabling access to theservice provider 40. - 31. As shown, a
first processing system 12 is operably coupled to afirst modem 16 viainterface 14. Theinterface 14 provides a communications pathway for unmodulated data transfers between thefirst modem 16 and thefirst processing system 12. Alternatively, thefirst modem 16 could be internally fixed inside thefirst processing system 12 and coupled through a standard interface of thefirst processing system 12. Example interface standards employed between themodem 16 and thefirst processing system 12 include EISA, PCMCIA, RS-232 and other industry accepted interface protocols. - 32. In operation the
first processing system 12 communicates digitally with thefirst modem 16. Data from thefirst processing system 12 is transferred through theinterface 14 and maintained in a buffering mechanism such as an internal memory space (ROM or RAM) or other similar configuration inside thefirst modem 16. In many applications, a universal asynchronous receiver transmitter (“UART”) or other similar device is used to handle data flow between thefirst processing system 12 and themodem 16. - 33. The
first processing system 12 can be a desktop or notebook computer, workstation or other similar computing device. Thefirst modem 16 includes a signal processing device for converting digital data from theprocessing system 12 to an equivalent modulated analog waveform. The modem also includes an interface such as an analog front end (“AFE”) which couples the analog signals carrying the data to the communications link 18 to thecentral office 24. - 34. Typically, the transmit and receive functions of the
modem 16 are operated by theprocessing system 12. A digital-to-analog/analog-to-digital conversion circuit or other similar signal processing device can be used to transmit and receive analog signals via the communications link 18. The communications link 18 can be analog twisted pair (“ATP”) wiring or other medium of the type often used in public switched telephone networks (“PSTN”). - 35. The digital data from the
processing system 12 is converted to analog equivalent signals which are transmitted on theATP wiring 18 using known modulation methods. Examples of such modulation methods include Quadrature Amplitude Modulation (“QAM”), Trellis Code Encoding (“TCE”) and Frequency Shift Keying (“FSK”) among others. Likewise, theservice provider 40 receives the signals from the far endcentral office 28 thedigital backbone connection 30. The data is passed to thesecond processing system 36 at theservice provider location 40. Theprocessing system 36 is employed by theservice provider 40 to gain access to thecentral office 28 and to format, receive and transmit data to other entities such assubscriber 15. - 36. Both
processing systems - 37. As described herein, the invention has particular application in a DMT-compliant signaling system with a pair of ADSL modems communicating over a
network 50. In one embodiment, an ADSL modem on theCPE unit 18 side communicates with anotherADSL modem 22 maintained by thecentral office 24. With such a configuration, multiple transmission channels can be maintained: a high speed downstream channel, a medium speed duplex channel, and a Plain Old Telephone Service (“POTS”) channel. The POTS channel can be split off from the ADSL modem by filters, thus guaranteeing uninterrupted POTS, even if ADSL fails. The high speed channel ranges from 1.5 to 6.1 Mbps, while duplex rates range from 16 to 640 kbps. Each channel can be submultiplexed to form multiple lower rate channels. Still other data rates may be employed as may be standardized in the industry. - 38. Downstream data rates depend on a number of factors, including the length of the copper line, its wire gauge, presence of bridged taps, and cross-coupled interference. Line attenuation increases with line length and frequency, and decreases as wire diameter increases.
- 39. Preferably, the
communications system 10 supports both downstream and upstream communications. Typically, when asubscriber 15 wishes to place a call, themodems analog line 18. Thecentral office 24 comprises the call switching and routing equipment necessary to complete a path between thesubscriber 15 and theservice provider 40. - 40. The functionality of the
subscriber 15 and theservice provider 40 can be implemented using known methods and devices. For example, the signaling protocols used by the various devices on thesystem 10 may include those supported by the International Standard Organization (“ISO”), the International Telegraph and Telephone Consultative Committee (“CCITT”) and the Electronics Industries Association (“EIA”) among other standards bodies. - 41. Once a call is established, data from the
first processing system 12 is passed to thefirst modem 16, modified appropriately with signal processing techniques to prepare the data for robust transmission, and converted using a digital to analog convertor or other similar conversion device. Next themodem 16 transmits the converted digital stream using a modulatedanalog signal carrier 20 over analogtwisted pair 18 to thecentral office 24. At the central office, anotherDSL modem 22 decodes the data and ports it to thedigital backbone network 26. Thedigital backbone network 26 connects thecentral offices central office 28 receives the digital data from thebackbone 26 and port it on thedigital link 30 the service provider'sprocessing platform 36. - 42. As shown, the near end
central office 24 communicates with the far endcentral office 28 via thedigital backbone network 26 which is often a high speed digital link providing a high data rate. The analog signals 18 are received by thesecond modem 22 and demodulated to obtain the original data sequence generated by thefirst processing system 12. Thesecond modem 22 passes the digital bit stream sequence to thedigital backbone network 26 and on the far endcentral office 28. Thus,processing systems communications mediums service provider 40 can be an Internet service provider or dedicated data service provider of the type available in industry. Themodems 16 annd 22 comprise the data communication equipment (“DCE”) responsible for providing the required translation and interface between the digital and analog signals carried on thesystem 10. - 43. Preferably, the
modems processing systems 12 is superimposed upon one or more carrier frequencies and transmitted on thesystem 10. The data signals is often referred to as the base band signal. Thecarrier signal 18 is changed back to the digital signal at the receivingmodem 22 through the process of demodulation. In essence themodems processing system 12 at thesubscriber location 15. - 44.FIG. 2 is a process flow diagram 50 illustrating the functions required for transmission of data from one Discrete Multi-Tone (“DMT”) modem to another. Preferably, the functions are implemented on a pair of T1.413 compliant Asymmetric Digital Subscriber Loop (“ADSL”) modems. Typically, each modem implements both transmit 52 and receive 100 functions. Thus, the methods of the present invention can be implemented on any modem platform, even one that does not support Reed-Solomon decoding.
- 45. For example, according to another embodiment, a general purpose programmable modem that implements the T1.413 standard is disclosed. The modem supports T1.413 standard transmission/reception functions over a discrete multi-tone communications system and can be linked to an
independent processing system 12 such as a desktop or notebook computer. Preferably, the modem comprises an interface having an analog side and a digital side with the analog side coupled to thecommunications system 10. A signal processing device for transmitting and receiving signals to and from thecommunications system 10 where the signal processing device is coupled to the digital side of the interface. The user interface formings a communications pathway between theindependent processing system 12 and the ADSL modem. - 46. The modem incorporates the clipping error estimation and correction functions of the present invention. In particular, according to one embodiment, the ADSL modem contains stored program segments capable of causing the signal processing circuit to analyze original signals received from the communications system to determine if they have been clipped during transmission and compute error compensation signals for signals that are clipped during transmission. The modem can then add the computed error compensation signal to the original signals received to obtain corrected analog signals.
- 47. With reference to FIG. 2, the transmit
path 54 shown is in one direction although an actual bidirectional ADSL communications link between modems extends in both directions . Reed-Solomon (“RS”) forward error-correction encoding 56 can be applied to bothfast bits 54 andslow bits 55 in the transmitter. Preferably, the RS-encodedslow bits 57 are interleaved 60 to provide extra resistance to clipping and channel error. Thus, performance is improved at the expense of additional encoding delay. - 48. Next, the RS-encoded
fast bits 59 are combined with the RS-encoded and interleavedslow bits 61 in a frame of bits that is converted into a frame of two-dimensional transmission symbols 62. Preferably, the transmission signals represent information in a form that is conducive to robust transmission over an ADSL communications link. Next, the Inverse Discrete Fourier Transform (“IDFT”) 64 is applied to the frame oftransmission symbols 63 after appropriate conjugate mirroring of the frame according to the T1.413 standard. - 49. The conjugate symmetry enforced on the
IDFT input 63 causes theoutput 65 to be completely real. As shown, the REAL{}operator 66 indicates that the imaginary component of theIDFT output 65 is not used. Next a cyclic prefix is created from the last samples of the frame of real data and concatenated 68 to the beginning of the frame. Thus, the resulting frame of data is passed through a Digital-to-Analog Converter (“DAC”) 70 and a transmitfilter 72 that provides the low-pass and high-pass filtering functions. The filtering functions are required to ensure the transmitted signal does not exceed the T1.413 power spectral density mass. - 50. A
line driver circuit 74 can be used to provide the current and/or voltage gain required to place the filtered signal onto the twisted pair telephone line with an ADSL format with the appropriate power level. After transmission through thechannel 76, the signal passes through receive buffers and filters 78 and into the Analog-to-Digital Converter (“ADC”) 80. Theeffective analog channel 82 seen by the transmission system is formed by the analog portion of theDAC 70, the transmitfilters 72, theline driver 74, theactual DSL channel 76, the receive buffers and filters 78, and the analog portion of theADC 80. - 51. In one embodiment, a Time-domain Equalizer (“TEQ”) is applied 84 to the signal to shorten the time-domain impulse response of the entire analog channel. The combined
analog channel 82 andTEQ 84 form an effective “shortened” channel whose impulse response has energy concentrated in a small number of samples. Preferably, significant energy is concentrated in the same number of samples as are present in thecyclic prefix 68. - 52. A frame of data is formed and the cyclic prefix is removed 86. The Discrete Fourier Transform (“DFT”) 88 can then be applied to the signal. As shown the transformed signal is processed with a Frequency-domain Equalization (“FEQ”)
function 90 which compensates for channel distortion. Next, the frame of FEQ output symbols is decoded intobits 92. The slow bits are passed through a de-interleaver 94 which spreads out transmission errors over many frames. Both fast and slow bits are passed throughRS decoders - 53.FIG. 2 shows the possible locations where clipping can occur in the transmitter (indicated by the “★”). The location ★ at which a clip occurs typically depends on the design of the transmitter. Preferably, the method of estimating the clipping distortion and applying clipping compensation of the present invention works within the constraints of the T1.413 standard thus guaranteeing compatibility between a low-complexity modem implemented without RS decoding capabilities and the large possible range of T1.413 compliant modems that may be employed.
- 54. In operation, during initialization and training, the Signal-to-Noise (“SNR”) is measured in all the operating subchannels of the receiver. The SNR values are then communicated back to the transmitter so that the transmitter can properly allocate bits to each subchannel to maximize the transmission throughput.
- 55. In one embodiment, a small subset of the operating subchannels are forced to carry fewer bits than they are capable of supporting. This provides the receiver with the ability to estimate the position and amplitude of clipping errors. For example, during initialization the receiver can inform the transmitter that subchannels 33, 34, 35, and 36 have a low SNR and should only carry two bits each. This results in a symbol constellation composed of four points (“4-QAM”) being used for subchannels 33, 34, 35 and 36. The 4-QAM result applies to the case of T1.413 transmission where the Trellis encoding is not employed. Similar constraints hold true when Trellis encoding is applied, i.e. two bits per subchannel will still result in a low-density constellation.
- 56. In addition, due to low data density, it is highly likely that the received symbol will be nearest in Euclidean distance to the transmitted symbol at the receiver. Preferably, the receiver can reliably estimate that the transmitted symbol was the constellation point nearest the received symbol. A reliable estimate of the transmission error can then be computed as the difference between the received signal and the closest constellation point.
- 57. The use of a reliable estimate of the transmission error in a given frame is provided within the T1.413 standard. The reliable estimate facilitates the methods for computing a clipping compensation signal at the receiver as described herein.
- 58. Turning to FIG. 3, a block diagram of a receiver according to one embodiment is shown and denoted generally as 110. In general, the receiver is configured to perform the clipping estimation and correction functions.
- 59. The received
signal 112 from the channel is passed through theTEQ 84. Next, the cyclic prefix is removed 86 prior to theDFT function 88 being applied. TheFEQ 90 corrects channel distortion as in a normal DMT receiver with a portion of the frame output from the FEQ being analyzed. The tones which were made to carry fewer bits are made available to the receiver for computing an error correction signal E[k]. If analysis of theincoming signal 112 indicates a need for clipping correction, then the correction signal (E[k]) is computed 114 and added 116 to theFEQ output 112. - 60. The analysis of the
incoming signal 112 can be accomplished in a variety of ways. For example, in one embodiment abuffer 120 is used to indicate that the correction signal E[k] is added 116 back to the frame from which the clipping parameters were estimated. The clipping-corrected frame of symbols are then passed through thesymbol decoder 92 to produce the transmitted bits. - 61.FIG. 4 shows a preferred embodiment of the
method 150 used for detecting the need for fixing a clip and then estimating the clipping signal. The clipping signal allows the receiver to construct a correction signal E[k] if needed. First, theFEQ output 112 is analyzed, and the SNR is computed atstep 156 for the tones which were forced to carry less bits than they are capable of carrying. At this point, the SNR is compared to a stored threshold,step 158. If the SNR is above the threshold, then no significant clipping occurred. Also, no clipping compensation should be applied and process flow is directed to step 160. In one embodiment, the compensation signal (E[k]) is set to zero for all values (k) as shown instep 160. This results in no clipping compensation being applied. Alternatively, the step of adding the compensation signal can be avoided and there is no need to add zeros. - 62. If the SNR is below the threshold, then a significant clip has occurred and process flow is directed to either step 162 or
step 164. The clipping error e[n] can be defined as the difference between the unclipped signal and the clipped signal in the transmitter. As such, the frequency-domain error signal at the receiver can be used to compute the amplitude,step 164, and phase,step 162, of the time-domain clipping error. In one embodiment, an estimate is made based on the premise that at least one clip occurred in the transmitter. Thus, a model for the clipping error e[n] can be devised as an impulse of amplitude (α) and delay (d) as shown below in Equation 1: - e[n]=α {overscore (o)}(n−d)
Equation 1 - 63. The DFT of the clipping error is therefore computed as shown below in Equation 2:
- E[k]=α exp (−j2IIkd)/
N Equation 2 - 64. In
equation 2, k is the sample index of the time-domain DMT signal and N is the number of DFT points. According to one embodiment of the ADSL standard, N=512, k ranges from 0 to 255, and d can assume any value from 0 to 511. - 65. The absolute value of the frequency-domain error signal is equal to the amplitude of the clipping error alpha (α). The phase of the frequency-domain error signal can be used to compute the delay (d) as shown in
step 164. The estimated values of the delay (d) and alpha (α) specify the location and value of the time-domain clipping error [n]. As shown instep 166, to enhance the quality of the clipping compensation signal, additional operations can be applied to verify and possibly correct the position estimated delay (d). - 66. If channel distortion has been properly compensated by the receiver, then an estimate of the clipped DMT signal at the transmitter can be reconstructed by applying an IDFT operation,
step 168, to theFEQ output 112. At this point, the most probable position of the clipping error can be analyzed to see if the time-domain sample is near the amplitude rails. The amplitude rails comprise the positive and negative signal limits of thesystem 10 as shown in FIGS. 5 and 6. - 67. If the time-domain sample is within the amplitude rail of she system, then the current position estimate is probably correct. If not, then it is highly likely that one of the surrounding samples is near the amplitude rails and therefore the most likely position of the clipping error is at one of the surrounding samples. An ad-hoc technique may be applied to improve performance of the location estimation function as is known by those of ordinary skill.
- 68. In one embodiment,
step 168 involves a small search of the samples on either side of the most probable clip location. If the sample values are found in the search that are closer to the rail value than the sample at the estimated clipping location, the estimate of the clipping location is changed to the location of the point nearest the rails. - 69. Preferably, only a few points on either side of the estimated clipping location in the search are considered. This limits the need for computing a full IDFT at
step 168. A few points of the IDFT can be computed to obtain the required information, e.g. the IDFT at the estimated clipping location and two points on either side of that can be computed for a total of five IDFT points. - 70. Based on the best estimate of alpha (α) and (d), the clipping compensation signal E[k] can be computed at
step 170. While E[k] can be generated as a time-domain signal, some advantages are provided by directly computing the frequency-domain compensation signal. Thus, the compensation signal E[k] can be directly added to theFEQ output 112 and further decoding operations can then be applied. This approach is preferred in some applications since it results in low complexity of the receiver. - 71. In other embodiments, it could prove advantageous to generate the compensation signal E[k] in the time domain and add it to the time domain signal before performing another DFT operation.
- 72. The clipping estimation and
correction method 150 can be verified in the context of an ADSL transmission system operating over a standard 24 AWG twisted pair to verify operation and performance. FIG. 5 shows a frame ofdata 200 generated by an ADSL DMT transmitter. The dashedlines sample point 206 that exceeds the maximum allowable signal level levels for thesystem 10. - 73.FIG. 6 shows the same frame of
data 200 after clipping wherein the newlargest value 208 fitting within the rails. Thesignal 200 has been clipped 208 to an amplitude value that is contained within the limited values supported by the ADSL modem. As explained above, the most probable location where the signal was clipped are in theIDFT operation 64, theDAC 70, the transmitfilter 72 and theline driver 74. - 74.FIG. 7 shows the
response curve 210 of the DSL channel used in the simulation. Note that theimpulse response 212 of the ADSL channel overlaps with the voice band which extends from 20 Hz to 3.4 KHz. - 75.FIG. 8 shows the subchannel SNR of a single received data frame for each of the 255 subchannels. The
original SNR 220 is shown by the dashed line. Theenhanced signal 222 is shown by the solid line. Because of the significant clip at the transmitter, the subchannel SNR is limited to about 36 dB. The SNR after enhancement ofsignal 222 results from application of the clipping estimation andcorrection method 150 according to the present invention. As before, overlap of the response signal from the ADSL channel is shown overlapping the voice band. - 76. As shown, the SNR at the receiver after enhancement with the clipping estimation and
correction method 222 has been increased dramatically. In the high-frequency regions where the channel noise is greater than the clipping noise, the original and enhanced SNR are roughly equivalent. In general, when large clips occur at the transmitter, a more dramatic improvement is observed because the original system SNR is very low. - 77. Preferably, the optimal threshold to use in the algorithm is determined by the quality of the transmission channel which, in turn, is determined by line length and interference characteristics. Typically, the error signal can be estimated more accurately when the tones are analyzed to estimate the clip parameters having high SNR. In one embodiment, the optimal threshold can be computed based on performance parameters required for a specific application and a particular modem.
- 78. In general, a heuristic approach for selecting the delay (d) and alpha (α) based on observations of the transmission channel are highly effective for determining a near-optimal threshold setting. When the error magnitude computed at the receiver is small, the phase computation is difficult and the position estimate of the clip will be inaccurate. However, if the error magnitude is large, accurate phase computation is reliable, and hence clip position estimation is reliable. Thus, large clips which would cause the most harm are easiest to correct, while small clips that would do the least damage are more difficult to correct.
- 79. According to one embodiment, a correction method is used wherein a threshold is set so that the algorithm is not applied if the clip is judged to be small enough that the algorithm would not perform well.
- 80. The probability of having two or more independent clips in a transmitted frame is extremely small. However, the probability is non-zero and hence should be given some consideration. According to one embodiment, the clipping estimation and
correction method 150 is applied to correct multiple clips with a slight modification. For example, to correct two clips, the model of the clip at the transmitter is changed to: - e[n]α {overscore (o)} [n−d 1 ]+β {overscore (o)} [n−d 2] Equation 3
- 81. Equation 3 allows a correction signal to be computed for two clips of different amplitudes and delays. The analysis of clipping error may be performed more carefully to properly estimate the parameters of the equation. Typically, more tones are required for adequate estimation of the parameters. At the minimum, two tones are used in the estimation procedure to determine the four parameters.
- 82. Preferably, the transmitter clipping occurs at the
IDFT operation 64, theDAC 70, or the early stages of the transmitfilter 72 since a clip that occurs at theline driver 74 is more difficult to correct since it is a nonlinear response that occurs between components of theoverall analog channel 82. A modem can be designed accordingly to guarantee the location of the clipping using known designs. - 83. In one embodiment, the clip signal is modified to represent the analog channel with two components. The first component is the, part of the channel before the clip. The second component is the part of the channel after the clip. Using the modified clip signal, a clip that occurs at the
line driver 74 can be represented as a delayed and shifted impulse subtracted from the signal after the first portion of the channel. - 84. According to one embodiment, the first portion of the
analog channel 82 is constructed at the receiver. Next, theFEQ 90 reverses the effect of which includes the first portion. Here, theFEQ 90 output frame is multiplied by the frequency response of the first portion of the analog channel. At this point, the analysis and clipping compensation signal is computed as before. - 85. It is noted that the resulting compensation signal E[k] can be multiplied by the inverse of the frequency response of the first portion of the
analog channel 82 before being added to theFEQ output 112. Essentially, this estimates the delayed and scaled impulse that occurs in theline driver 74 and computes a compensation signal E[k] that appears as though the clip has traveled through only the second half of thetransmission channel 82 as opposed to theentire transmission channel 82. - 86. While the invention has been described in conjunction with preferred embodiments, it should be understood that modifications will become apparent to those of ordinary skill in the art and that such modifications are intended to be included within the scope of the invention and the following claims.
Claims (36)
1. A method of communicating data over a Discrete Multi-Tone (“DMT”) communications system comprising the steps of:
measuring the signal to noise ratio (“SNR”) in all operating subchannels of the system;
transmitting SNR measurements for the operating subchannels to a transmitting device on said system; and
transmitting data over each of the operating subchannels using the SNR measurements to affect the throughput of data transmitted on the system.
2. The method according to claim I further comprising the steps of:
measuring the thoughput of transmitted data over the operating subchannels;
using throughput measurements to compute an estimate of clipping that occurred in the transmitter;
applying the estimate of clipping to construct a compensation signal; and
adding the compensation signal at the receiver to correct for clipping that occurred at the transmitter.
3. The method according to wherein the step of transmitting data over each of said operating subchannels is performed by adjusting the amount of data transmitted over individual operating subchannels to get a better estimate of the clipping parameters.
claim 1
4. The method according to wherein the step of transmitting data over each of said operating subchannels is performed by transmitting more data in subchannels with a high SNR.
claim 1
5. The method according to further comprising the step of designating operating subchannels with a low SNR.
claim 1
6. The method according to further comprising the step of using said operating subchannels with a low SNR to compute an error signal representative of the amount of distortion likely to occur in each of said operating subchannels.
claim 5
7. The method according to further comprising the step of limiting the amount of data transmitted over said operating subchannels with a low SNR to two bits each.
claim 5
8. The method according to further comprising the step of designating operating subchannels with a high SNR.
claim 1
9. The method according to claim I further comprising the step of estimating the amount of transmission error that is likely to occur in each of said operating subchannels.
10. The method according to further comprising the step of constructing an error compensation signal to compensate for transmission errors that are likely to occur in each of said operating subchannels.
claim 1
11. The method according to wherein the step of communicating data over each of said operating subchannels comprises the further steps of:
claim 1
separating said data into a plurality of transmission bits; and
adjusting the allocation of said plurality of transmission bits over said operating subchannels so that more transmission bits are transmitted in subchannels having a high SNR.
12. A method of compensating for clipping of signals transmitted over a digital subscriber line (“DSL”) system, the signals modulated using a multi-carrier transmission signaling technique over multiple subchannels of the DSL system, the signal to noise ratio (“SNR”) of said multiple subchannels varying from high to low, the method comprising the steps of:
transmitting a signal on the DSL system;
analyzing the signal to determine if it has been clipped during transmission; and
computing an error compensation signal for the signal if it is determined it has been clipped during transmission.
13. The method according to wherein the step of transmitting a signal over the DSL system comprises the steps of:
claim 12
separating the signal into a plurality of signal components; and
modulating each of the plurality of components over the multiple subchannels so that specific subchannels with low SNR values carry less signal components.
14. The method according to wherein the step of analyzing the transmitted signal to determine if it has been clipped comprises the steps of:
claim 12
analyzing the output from the specific channels designated to carry fewer bits than they are capable of carrying; and
computing an error signal based on the amount of signal distortion occurring over said specific channels.
15. The method according to wherein the specific subchannels used to compute the error signal are assigned 4-QAM data.
claim 13
16. The method according to further comprising the step of adding the error compensation signal to all signals arriving at a receiver on said DSL system.
claim 12
17. The method according to wherein the step of computing an error compensation signal is performed by estimating the amount of clipping that may have occurred at a transmitter on the DSL system.
claim 12
18. The method according to wherein the step of computing an error compensation signal is performed by computing the amplitude and phase of the clip impulse for signals clipped during transmission.
claim 12
19. The method according to wherein the step of computing an error compensation signal is performed by computing the amplitude and phase of clips occurring in a single frame of data.
claim 12
20. The method according to wherein the step of computing an error compensation signal is performed by assuming that at least one clipping error e[n] occurred in a transmitter of the DSL system.
claim 12
21. The method according to wherein said clipping error e[n] is modeled as an impulse of amplitude (α) and delay (d) governed by the relationship e[n]=α {overscore (o)} (n−d).
claim 20
22. The method according to wherein the step of analyzing the signal to determine if it has been clipped during transmission is performed by the further steps of:
claim 21
forming an initial estimate of the position of any possible clipping impulses within a frame of data;
is sampling signal values within the frame on either side of the initial estimate; and
determining if the sampled signal values are closer to the rails of the DSL system than the estimate.
23. The method according to wherein the initial estimate of the position of the clip within the frame of data is changed to a sample location where the sampled signal value is closer to the rail than the initial estimate.
claim 22
24. The method according to wherein only a few points on either side of the initial estimate are considered.
claim 22
25. A device for communicating information over a discrete multi-tone communications system comprising:
an analog front end coupling said device to said communications system;
a signal processing circuit driving said analog front end; and
an interface capable of transmitting digital data from an independent processing system to the signal processing circuit, the signal processing circuit configured to convert the digital data to analog signal waveforms and transmit the waveforms over the communications system via the analog front end.
26. The device according to further including a storage device containing a plurality of program instructions that control the functionality of said signal processing circuit, said storage device operably linked to said independent processing system.
claim 25
27. The device according to wherein said program instructions form routines that cause the signal processing circuit to perform the following functions:
claim 25
transmit signals over the communications system;
receive signals transmitted over the communications system;
analyze received signals to determine if they have been clipped during transmission; and
compute error compensation signals for signals that are clipped during transmission.
28. The device according to further including a microprocessor coupled to said signal processing circuit, the microprocessor configured to read said program instructions and cause said signal processing device to perform said functions.
claim 27
29. The device according to wherein said signal processing circuit contains a digital-to-analog conversion circuit for converting digital signals to modulated analog equivalents suitable for transmission on said communications system.
claim 25
30. The device according to wherein said modulated analog equivalents are T1.413 compliant signals.
claim 29
31. An asymmetric digital subscriber line (“ADSL”) modem supporting T1.413 standard transmission/reception functions over a discrete multi-tone communications system, the ADSL modem communicable linked to an independent processing system and comprising:
an interface having an analog side and a digital side, the analog side coupled to the communications system;
a signal processing device for transmitting and receiving signals to and from the communications system, the signal processing device having a system side and a user side, the system side coupled to the digital side of the interface; and
a user interface coupled to the user side of the signal processing device and forming a communications pathway between the independent processing system and the ADSL modem.
32. The ADSL modem according to wherein said signal processing device contains a conversion circuit for converting between T1.413 compliant digital and analog signals.
claim 31
33. The ADSL modem according to further comprising a plurality of stored program segments operably coupled to said signal processing circuit for controlling the functions thereof.
claim 31
34. The ADSL modem according to wherein said stored program segments are capable of causing said signal processing circuit to measure the signal to noise ratio of transmitted signals over said communications signal.
claim 33
35. The ADSL modem according to wherein said stored program segments are capable of causing said signal processing circuit to perform the following functions:
claim 33
analyze original signals received from the communications system to determine if they have been clipped during transmission;
compute error compensation signals for original signals that are clipped during transmission; and
add computed error compensation signal to the original signals received to obtain corrected analog signals.
36. The ADSL modem according to wherein said stored program segments are further capable of causing said signal processing circuit to convert analog signals to their digital data equivalent, said analog signal comprising both original analog signals and corrected analog signals.
claim 35
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
US09/729,008 US6381268B2 (en) | 1998-05-12 | 2000-12-04 | Methods and device for estimating and correcting clipping in a discrete multi-tone communications system |
Applications Claiming Priority (2)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
US09/076,418 US6363109B1 (en) | 1998-02-03 | 1998-05-12 | Methods and device for estimating and correcting clipping in a discrete multi-tone communications system |
US09/729,008 US6381268B2 (en) | 1998-05-12 | 2000-12-04 | Methods and device for estimating and correcting clipping in a discrete multi-tone communications system |
Related Parent Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
US09/076,418 Division US6363109B1 (en) | 1998-02-03 | 1998-05-12 | Methods and device for estimating and correcting clipping in a discrete multi-tone communications system |
Publications (2)
Publication Number | Publication Date |
---|---|
US20010001007A1 true US20010001007A1 (en) | 2001-05-10 |
US6381268B2 US6381268B2 (en) | 2002-04-30 |
Family
ID=22131887
Family Applications (3)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
US09/076,418 Expired - Lifetime US6363109B1 (en) | 1998-02-03 | 1998-05-12 | Methods and device for estimating and correcting clipping in a discrete multi-tone communications system |
US09/729,008 Expired - Lifetime US6381268B2 (en) | 1998-05-12 | 2000-12-04 | Methods and device for estimating and correcting clipping in a discrete multi-tone communications system |
US09/729,464 Expired - Lifetime US6363110B1 (en) | 1998-05-12 | 2000-12-04 | Methods and device for estimating and correcting clipping in a discrete multi-tone communications system |
Family Applications Before (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
US09/076,418 Expired - Lifetime US6363109B1 (en) | 1998-02-03 | 1998-05-12 | Methods and device for estimating and correcting clipping in a discrete multi-tone communications system |
Family Applications After (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
US09/729,464 Expired - Lifetime US6363110B1 (en) | 1998-05-12 | 2000-12-04 | Methods and device for estimating and correcting clipping in a discrete multi-tone communications system |
Country Status (2)
Country | Link |
---|---|
US (3) | US6363109B1 (en) |
EP (1) | EP0957615A3 (en) |
Cited By (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US7016401B1 (en) | 1999-12-22 | 2006-03-21 | Motorola, Inc | Method and apparatus for automated correlation of digital modulation impairment |
US20070217543A1 (en) * | 2006-03-17 | 2007-09-20 | Fujitsu Limited | Peak suppression method, peak suppression apparatus and wireless transmission apparatus |
US20080292323A1 (en) * | 2007-05-24 | 2008-11-27 | Applied Optoelectronics, Inc. | Systems and methods for reducing clipping in multichannel modulated optical systems |
US20100104041A1 (en) * | 2007-02-28 | 2010-04-29 | Osaka University | Transmission method, transmission system, transmission device and reception device |
Families Citing this family (38)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
FR2758030B1 (en) * | 1996-12-31 | 1999-03-26 | Sgs Thomson Microelectronics | METHOD AND DEVICE FOR SHAPING A CLIPPING NOISE OF A MULTI-PORTABLE MODULATION |
US6654431B1 (en) * | 1999-09-15 | 2003-11-25 | Telcordia Technologies, Inc. | Multicarrier personal access communication system |
US6879640B1 (en) | 1999-10-20 | 2005-04-12 | Broadcom Corporation | Method, apparatus and system for high-speed transmission on fiber optic channel |
AU1222701A (en) * | 1999-10-20 | 2001-04-30 | Broadcom Corporation | Method, apparatus and system for high-speed transmission on fiber optic channel |
EP2293459A3 (en) | 2000-01-07 | 2014-04-16 | TQ Delta, LLC | Systems and methods for characterizing transmission lines in a multi-carrier DSL environment |
US6690719B1 (en) * | 2000-02-14 | 2004-02-10 | Itran Communications Ltd. | Host to modem interface |
US7496924B2 (en) * | 2000-03-27 | 2009-02-24 | Siemens Communications, Inc. | Dispatching messages among registered software modules in telecommunications system including telephony internet server coupled between packet network and PBX |
US6732281B1 (en) * | 2000-05-31 | 2004-05-04 | 3Com Corporation | System and method for loading data onto carriers |
US6963603B1 (en) | 2000-06-06 | 2005-11-08 | Ikanos Communication, Inc. | Method and apparatus for pre-compensation of an XDSL modem |
US6741235B1 (en) * | 2000-06-13 | 2004-05-25 | Michael Goren | Rapid entry of data and information on a reduced size input area |
US6931079B1 (en) * | 2000-06-21 | 2005-08-16 | Broadcom Corporation | Power reduction |
US7131038B2 (en) * | 2000-10-12 | 2006-10-31 | 3Com Corporation | Performance evaluation of multicarrier channels |
US7103096B2 (en) * | 2000-10-12 | 2006-09-05 | 3Com Corporation | Performance evaluation of multicarrier channels with forward error correction and automatic retransmission request |
US6732323B1 (en) * | 2000-10-12 | 2004-05-04 | 3Com Corporation | Method of selecting initialization parameters for multi-channel data communication with forward error correction |
US7672381B1 (en) * | 2000-10-17 | 2010-03-02 | Motorola, Inc. | Method of multiple-carrier communication within a noncontiguous wideband spectrum and apparatus therefor |
US7283583B2 (en) * | 2001-01-11 | 2007-10-16 | Tioga Technologies, Inc. | Adaptive rate transmission with dual noise margins |
US6954505B2 (en) | 2001-03-29 | 2005-10-11 | Texas Instruments Incorporated | Discrete multitone modulation with reduced peak-to-average ratio using unloaded subchannels |
US6718493B1 (en) | 2001-05-17 | 2004-04-06 | 3Com Corporation | Method and apparatus for selection of ARQ parameters and estimation of improved communications |
JP4040403B2 (en) | 2001-11-27 | 2008-01-30 | ソニー株式会社 | Information processing apparatus and method, recording medium, and program |
US7959395B2 (en) * | 2002-07-22 | 2011-06-14 | Brooks Automation, Inc. | Substrate processing apparatus |
US7988398B2 (en) | 2002-07-22 | 2011-08-02 | Brooks Automation, Inc. | Linear substrate transport apparatus |
US7280621B1 (en) | 2003-03-31 | 2007-10-09 | 3Com Corporation | Preamble detector method and device for OFDM systems |
US7548579B1 (en) | 2003-03-31 | 2009-06-16 | 3Com Corporation | Symbol spreading method and device for OFDM systems |
US20060039455A1 (en) * | 2004-03-19 | 2006-02-23 | Quamtum Advanced Technologies, Inc. | Composite waveform for transmitting an increased amount of data over an analog medium, and devices and methods for producing, encoding, transmitting and decoding same |
US7502336B2 (en) * | 2004-06-30 | 2009-03-10 | 2Wire, Inc. | Multi-carrier communication using adaptive tone-pruning |
EP1797690A1 (en) * | 2004-10-06 | 2007-06-20 | Nokia Corporation | Receiver-site restoration of clipped signal peaks |
CA2582957C (en) * | 2004-10-11 | 2013-09-03 | 2Wire Inc. | Periodic impulse noise mitigation in a dsl system |
US20060083321A1 (en) * | 2004-10-15 | 2006-04-20 | Hossein Sedarat | Multi-carrier communication bit-loading in presence of radio-frequency interferers |
US7953163B2 (en) * | 2004-11-30 | 2011-05-31 | Broadcom Corporation | Block linear equalization in a multicarrier communication system |
US7852950B2 (en) * | 2005-02-25 | 2010-12-14 | Broadcom Corporation | Methods and apparatuses for canceling correlated noise in a multi-carrier communication system |
US9374257B2 (en) * | 2005-03-18 | 2016-06-21 | Broadcom Corporation | Methods and apparatuses of measuring impulse noise parameters in multi-carrier communication systems |
US7924956B2 (en) * | 2005-03-31 | 2011-04-12 | Intel Corporation | System and method for compensation of non-linear transmitter distortion |
US7813439B2 (en) * | 2006-02-06 | 2010-10-12 | Broadcom Corporation | Various methods and apparatuses for impulse noise detection |
GB0620158D0 (en) * | 2006-10-11 | 2006-11-22 | Nokia Corp | Method and apparatus for reducing peak-to-average power ratio of a signal |
KR101322835B1 (en) * | 2007-08-03 | 2013-10-25 | 포항공과대학교 산학협력단 | Apparatus and method for iterative reconstructuring of amplitude clipped signal |
US8724464B2 (en) * | 2007-12-17 | 2014-05-13 | Broadcom Corporation | Method and system for near continuous data rate limit adjustment via a plurality of link variables in an energy efficient network |
US8605837B2 (en) * | 2008-10-10 | 2013-12-10 | Broadcom Corporation | Adaptive frequency-domain reference noise canceller for multicarrier communications systems |
US8602706B2 (en) * | 2009-08-17 | 2013-12-10 | Brooks Automation, Inc. | Substrate processing apparatus |
Family Cites Families (11)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US4679227A (en) * | 1985-05-20 | 1987-07-07 | Telebit Corporation | Ensemble modem structure for imperfect transmission media |
US5623513A (en) * | 1993-12-13 | 1997-04-22 | Amati Communications Corporation | Mitigating clipping and quantization effects in digital transmission systems |
US5519731A (en) * | 1994-04-14 | 1996-05-21 | Amati Communications Corporation | ADSL compatible discrete multi-tone apparatus for mitigation of T1 noise |
US6137839A (en) * | 1996-05-09 | 2000-10-24 | Texas Instruments Incorporated | Variable scaling of 16-bit fixed point fast fourier forward and inverse transforms to improve precision for implementation of discrete multitone for asymmetric digital subscriber loops |
FR2758030B1 (en) * | 1996-12-31 | 1999-03-26 | Sgs Thomson Microelectronics | METHOD AND DEVICE FOR SHAPING A CLIPPING NOISE OF A MULTI-PORTABLE MODULATION |
US6208637B1 (en) * | 1997-04-14 | 2001-03-27 | Next Level Communications, L.L.P. | Method and apparatus for the generation of analog telephone signals in digital subscriber line access systems |
JP3033708B2 (en) * | 1997-04-21 | 2000-04-17 | 埼玉日本電気株式会社 | Peak clipping circuit and quadrature modulation transmitter having the peak clipping circuit |
US6240129B1 (en) * | 1997-07-10 | 2001-05-29 | Alcatel | Method and windowing unit to reduce leakage, fourier transformer and DMT modem wherein the unit is used |
US6219378B1 (en) * | 1997-09-17 | 2001-04-17 | Texas Instruments Incorporated | Digital subscriber line modem initialization |
US6081502A (en) * | 1997-09-18 | 2000-06-27 | Orckit Communications Ltd. | Method and apparatus for reducing probability of clipping |
US6047025A (en) * | 1998-02-13 | 2000-04-04 | Motorola, Inc. | Method and apparatus for equallization in an asymmetric digital aubscriber line communications system |
-
1998
- 1998-05-12 US US09/076,418 patent/US6363109B1/en not_active Expired - Lifetime
-
1999
- 1999-05-11 EP EP99201455A patent/EP0957615A3/en not_active Withdrawn
-
2000
- 2000-12-04 US US09/729,008 patent/US6381268B2/en not_active Expired - Lifetime
- 2000-12-04 US US09/729,464 patent/US6363110B1/en not_active Expired - Lifetime
Cited By (8)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US7016401B1 (en) | 1999-12-22 | 2006-03-21 | Motorola, Inc | Method and apparatus for automated correlation of digital modulation impairment |
US7154942B2 (en) | 1999-12-22 | 2006-12-26 | General Instrument Corporation | Method and apparatus for automated correlation of digital modulation impairment |
US20070217543A1 (en) * | 2006-03-17 | 2007-09-20 | Fujitsu Limited | Peak suppression method, peak suppression apparatus and wireless transmission apparatus |
US7839949B2 (en) * | 2006-03-17 | 2010-11-23 | Fujitsu Limited | Peak suppression method, peak suppression apparatus and wireless transmission apparatus |
US20100104041A1 (en) * | 2007-02-28 | 2010-04-29 | Osaka University | Transmission method, transmission system, transmission device and reception device |
US8594224B2 (en) * | 2007-02-28 | 2013-11-26 | Osaka University | Transmission method, transmission system, transmission device and reception device |
US20080292323A1 (en) * | 2007-05-24 | 2008-11-27 | Applied Optoelectronics, Inc. | Systems and methods for reducing clipping in multichannel modulated optical systems |
US8165475B2 (en) * | 2007-05-24 | 2012-04-24 | Applied Optoelectronics | Systems and methods for reducing clipping in multichannel modulated optical systems |
Also Published As
Publication number | Publication date |
---|---|
EP0957615A2 (en) | 1999-11-17 |
US20010000220A1 (en) | 2001-04-12 |
US6363109B1 (en) | 2002-03-26 |
US6363110B1 (en) | 2002-03-26 |
US6381268B2 (en) | 2002-04-30 |
EP0957615A3 (en) | 2002-11-06 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
US6381268B2 (en) | Methods and device for estimating and correcting clipping in a discrete multi-tone communications system | |
US6549512B2 (en) | MDSL DMT architecture | |
US6219378B1 (en) | Digital subscriber line modem initialization | |
KR100356383B1 (en) | Method of mitigating the effects of clipping or quantization in the d/a converter of the transmit path of an echo canceller | |
US6052411A (en) | Idle mode for digital subscriber line | |
US8265215B1 (en) | Method and system for determining symbol boundary timing in a multicarrier data transmission system | |
FI115685B (en) | A method for transmitting a pattern in a multi-carrier modulation transmission system | |
KR100421212B1 (en) | Point-to-multipoint transmission method of multicarrier system for continuous transmission of high-speed multicarrier data signals in digital subscriber line | |
US7075999B2 (en) | Multicarrier modulation with data dependent frequency-domain redundancy | |
US7352805B2 (en) | Systems and methods for adaptive VDSL with variable sampling frequency and time-domain equalizer | |
US7787550B2 (en) | Combined frame alignment and timing recovery in digital subscriber line (DSL) communications systems | |
EP1610515A1 (en) | Method and apparatus for timing synchronization in a multicarrier system | |
GB2303032A (en) | Adaptive bit swapping between channels of a discrete multitone system | |
US20020080867A1 (en) | Robust signaling techniques in multicarrier systems | |
US6512789B1 (en) | Partial equalization for digital communication systems | |
US7542513B2 (en) | Communication data using wideband communications | |
US20050276339A1 (en) | Training sequence for symbol boundary detection in a multicarrier data transmission system | |
US6366555B1 (en) | Method and device for controlling signal clipping in a discrete multi-tone communications system | |
US6603811B1 (en) | Low complexity frequency domain equalizer having fast re-lock | |
US7397860B1 (en) | Fractional local peak detection and mitigation for PAR reduction | |
US6483869B1 (en) | Frequency decimated DMT modulation modem | |
US8498326B1 (en) | Method and apparatus for forward error correction | |
KR100512172B1 (en) | Asymmetric digital subscriber line system | |
JP2003124901A (en) | Method for initially setting communication system having receiver and transmitter of different bandwidth | |
US20050163154A1 (en) | Communication method and communication system |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
STCF | Information on status: patent grant |
Free format text: PATENTED CASE |
|
AS | Assignment |
Owner name: RENESAS TECHNOLOGY CORP., JAPAN Free format text: ASSIGNMENT OF ASSIGNORS INTEREST;ASSIGNOR:MITSUBISHI DENKI KABUSHIKI KAISHA;REEL/FRAME:015185/0122 Effective date: 20030908 |
|
FPAY | Fee payment |
Year of fee payment: 4 |
|
FPAY | Fee payment |
Year of fee payment: 8 |
|
FPAY | Fee payment |
Year of fee payment: 12 |