JP2004064561A - Dmt modem and its timing reproducing method - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a DMT MODEM whose maximum transmission quantity is large and its timing reproducing method. <P>SOLUTION: This DMT(discrete multi-tone) MODEM 600 for receiving a signal transmitted from a transmitter is configured to select a signal used at the time of synchronizing a clock 506 of the DMT MODEM with a clock of the transmitter based on the rate of a signal received from the transmitter and noise. <P>COPYRIGHT: (C)2004,JPO

Description

[0001]
TECHNICAL FIELD OF THE INVENTION
The present invention relates to a DMT modem and a timing recovery method thereof, and more particularly, to a DMT modem capable of increasing a maximum transmission amount and a timing recovery method thereof.
[0002]
[Prior art]
As a technique for providing a digital subscriber line transmission system using an existing telephone line as a high-speed data communication line, xDSL (Digital Subscriber Line) is known. xDSL is a transmission method using a telephone line and is one of modulation and demodulation techniques. The xDSL is roughly divided into an upstream transmission speed from a subscriber's house (hereinafter, referred to as a subscriber side) to an accommodation station (hereinafter, referred to as a station side) and a downstream transmission speed from the station side to the subscriber side. Are symmetric and asymmetric.
[0003]
Asymmetric xDSL includes ADSL (Asymmetric DSL). Recommendations include (1) International Telecommunication Union Telecommunication Standardization Sector (ITU-T). 992.1 (G. dmt), (2) ITU-T G. 992.2 (G. lite), (3) ANSI (American National Standards Institute) standard T1.413. Both employ a DMT (Discrete Multiple Tone) modulation method as a modulation method. The number of carriers in (1) and (3) is 256, and the number of carriers in (2) is 128.
[0004]
FIG. 9 shows the principle of the DMT modulation method. As shown in FIGS. 9 and 10, the DMT decomposes the transmission data 5 into a group of consecutive N bits in a serial-to-parallel buffer (SP buffer) 10. Multiple carriers f at modulation rate ₁ , F ₂ , ... f _M 2 for ^N This is a method for performing quadrature amplitude phase modulation (QAM) of a value. M carrier waves f ₁ , F ₂ , ... f _M Are QAM-modulated in each group, and the modulation outputs are combined (inverse Fourier transform 30), the data signal speed is as follows.
Data signal rate = N × 4000 × M (bps).
[0005]
In the ADSL full standard, a large number of carriers having a frequency interval of 4 kHz and a frequency of 25 kHz or more are used. From the above M and N, the maximum transmission speed is as follows.
The number M of usable carriers differs depending on the content of the above recommendation, and the same G. In dmt, M and N are different depending on the target region (North America (Annex A), Europe (Annex B), and Japan (Annex C)).
For example, in the case of ANSI T1.413, the following is performed.
Number of uplink usable carriers (M): 25 (# 6 to 15, 17 to 31) ... # 16 is a pilot.
Number of downlink usable carriers (M): 249 (# 6-63, 65-255) ... # 64 is a pilot.
Maximum bitmap number (N): 15 bits (up and down).
Uplink maximum transmission speed: 15 (bit) × 4 (kHz) × 25 (line) = 1500 (kbps).
Maximum downlink transmission speed: 15 (bit) × 4 (kHz) × 249 (line) = 14940 (kbps).
The maximum transmission speed of the uplink and downlink is as described above, but the optimal carrier and the number of bitmaps are selected according to the noise of the used line, crosstalk, etc. it can.
[0006]
Since the line quality is determined by determining the signal-to-noise ratio (SNR) from the received waveform, the modem on the receiving side determines the number of bitmaps for each carrier (see the received bitmap 150 in FIG. 10). For this reason, prior to communication, the modems notify each other of the number of bitmaps used by each other (see the transmission bitmap 60 in FIG. 10).
[0007]
As shown in FIG. 10, the demodulation is performed by the decoder 130 for each carrier obtained by Fourier transforming the received waveform by the FFT 110. The full-duplex communication is performed by an echo canceller method, a frequency division multiplex method for changing a frequency band in transmission and reception, a time division multiplex method for changing time allocation, and the like.
[0008]
Further, the DMT modulation method will be described with reference to FIG. 9 and 10, the corresponding components are denoted by the same reference numerals and described.
[0009]
FIG. 10 is a block diagram showing a configuration of the ADSL transceiver of ANSI T1.413. In FIG. 10, only the transmitting unit on the station side and the receiving unit on the subscriber side are shown, and the receiving unit on the station side and the transmitting unit on the subscriber side are omitted from the drawing. Has substantially the same configuration as the illustrated subscriber-side receiving unit, and the subscriber-side transmitting unit has substantially the same configuration as the illustrated station-side transmitting unit. Here, only the modulation and demodulation in the downlink direction from the station to the subscriber will be described, but the modulation and demodulation in the uplink direction from the subscriber to the station is performed in the same manner as in the downlink direction.
[0010]
As shown in FIG. 10, the transmitting unit on the station side includes a serial-to-parallel buffer (SP buffer) 10 for converting serial transmission data into parallel, an encoder 20, and a 512-point inverse fast Fourier transformer ( IFFT) 30, a cyclic prefix adding unit 40, a parallel to serial buffer (PS buffer) 41, a D / A converter 50, and a transmission bitmap 60.
[0011]
The receiving section on the subscriber side includes an A / D converter 80 for converting an analog signal from the subscriber line 70 into a digital signal, a time domain equalizer (TEQ) 90, and a serial to parallel buffer (S- A P buffer 101, a 512-point fast Fourier transformer (FFT) 110, a frequency domain equalizer (FEQ) 120, a decoder 130, and a parallel-to-serial buffer (PS buffer) 140 are provided. .
[0012]
Next, the operation will be described with reference to FIGS.
[0013]
First, the transmission data 5 is input to the ADSL transceiver on the station side, and stored in the SP buffer 10 for one symbol time (1/4 kHz). The stored data is divided by the number of transmission bits per carrier predetermined in the transmission bitmap 60, and 256 N-bit data 6 ₁ , 6 ₂ , ... 6 _M Is output to the encoder 20. Here, M = 256 is the number of carriers when the inverse Fourier transform is performed by the IFFT 30. Also, 256 N-bit data 6 ₁ , 6 ₂ , ... 6 _M Corresponds to one symbol.
[0014]
In the encoder 20, the input bit sequence 6 ₁ , 6 ₂ , ... 6 _M Signal points for quadrature amplitude modulation (256 QAM signals 7 ₁ , 7 ₂ , ... 7 _M ) And outputs it to the IFFT 30. The IFFT 30 performs an inverse fast Fourier transform to perform quadrature amplitude modulation on each signal point, and outputs the result to the PS buffer 41. At this time, if the output of 512 points of the IFFT 30 is set to 0 to 511, the cyclic / prefix adding section 40 adds 32 points (samples) of 480 to 511 as the cyclic prefix to the head of the DMT symbol.
[0015]
The output of the PS buffer 41 is sent to the D / A converter 50, where it is converted to an analog signal at a sampling frequency of 2.208 MHz and transmitted to the subscriber side via the metallic line 70. Here, the sampling frequency corresponds to the case of 512 outputs of the IFFT 30, and 2.208 MHz = (4.3125 × 512) KHz. Note that carrier # 0 is direct current (DC), carrier # 1 is 4.3125 kHz, and carrier #M is (4.3125 × M) kHz.
[0016]
In the receiving unit in the ADSL transceiver on the subscriber side, the digital signal is converted into a 2.208 MHz digital signal by the A / D converter 80 and output to the TEQ 90. After processing by the TEQ 90 so that inter symbol interference (ISI) falls within the cyclic prefix of 32 samples, the data is stored in the SP buffer 101 for one DMT symbol. In the SP buffer 101, serial-to-parallel conversion is performed at 512 points. At this time, the cyclic prefix is deleted by the cyclic prefix deleting unit 100.
[0017]
The output of the SP buffer 101 is input to the FFT 110. In the FFT 110, fast Fourier transform is performed, and signal points (255 QAM signals 8 ₁ , 8 ₂ , ... 8 _M ) Is generated (demodulated). Thereafter, the demodulated signal points are compensated by the FEQ 120 for each carrier having a different frequency, for the influence on the amplitude and phase received through the metallic line 70, and the compensated 255 QAM signals 8 ' ₁ , 8 ' ₂ … 8 ' _M Is decoded by the decoder 130 in accordance with the reception bitmap 150 holding the same value as the transmission bitmap 60. The decoded data is stored in the PS buffer 140 and becomes the received data as a bit string.
[0018]
According to ANSI T1.413, 255 QAM signals 8 ' ₁ , 8 ' ₂ … 8 ' _M Of which, tone 64 is used to transmit a "pilot tone" that enables synchronization of the clocks of the central office (CO: office) and customer premises equipment (CPE: subscriber) modems.
[0019]
The pilot tone is often transmitted by the CO modem (master), and the pilot tone is used to synchronize the clock of the CPE modem (slave) with the clock of the CO modem. The CPE modem continuously checks the received pilot signal to control the clock of the CPE modem using conventional pilot tracking techniques.
[0020]
FIG. 11 is a block diagram showing a configuration of a CPE modem described in Japanese Patent Application Laid-Open No. 2001-285269. 11, the same components as those in FIG. 10 are denoted by the same reference numerals, and the description thereof will be omitted.
[0021]
As shown in FIG. 11, the CPE modem 500 includes a reception filter 75, an A / D converter 80, a sample clock 506, a TEQ 90, an SP buffer 101, an FFT 110, an FEQ 120, and a gain adjustment unit. 125, a decoder 130, a phase error detector (clock frequency adjuster) 145, and a PLL unit 165. The PLL section 165 has a loop filter 146 and a VCO 147.
[0022]
Receive filter 75 includes a splitter that filters out bidirectional two-wire signals that are not included in a typical DSL transmission band. For example, low frequency POST signals are blocked by a splitter so as not to interfere with high frequency ADSL signals, as is well known in the art.
[0023]
The FFT 110 converts the digital samples input from the time domain equalizer 90 into the frequency domain, and separates the digital samples into all sub-channels (# 1 to 255). One of the sub-channels is the pilot tone transmitted on tone 64 (# 64).
[0024]
This pilot tone is defined as follows in the section of Pilot in 6.11.1.2 of ANSI T1.413-1998. Subcarrier 64 (f = 276 kHz) is used as a pilot, ie, b ₆₄ = 0 and g ₆₄ = 1. The data modulated by the pilot subcarrier has a constant value {0, 0}, and generates a constellation point of {+, +} (see FIG. 5B).
[0025]
The gain adjustment unit 125 adjusts the gain of the signal (# 1 to # 255) output from the FEQ 120 and outputs the signal to the decoder 130.
[0026]
The phase error detector 145 receives, from the FEQ 120, a pilot tone (# 64) included for each symbol. Phase error detector 145 measures the phase error between symbols based on current and previously received pilot tones.
[0027]
The pilot tone is a signal subjected to quadrature amplitude modulation (QAM) as described above, and is shown as a complex number (that is, a real component and an imaginary component) (the horizontal axis in a diagram showing a constellation such as FIG. 5B). X indicates the real part and Y axis indicates the imaginary part).
[0028]
Thus, the pilot tones are shown as vectors when plotted on the XY plane. This pilot vector rotates over time when the receiver clock is not synchronized with the remote transmitter clock (see FIG. 5 (b)). The phase error detector 145 gives a phase difference (error) between pilot tone symbols to the PLL unit 165 as a phase error.
[0029]
The sampling clock 506 is set as a clock of the A / D converter 80. When a phase error is input from the phase error detector 145, the PLL unit 165 adjusts the sampling clock 506.
[0030]
Further, in the multi-carrier transmission system of the DMT system, as disclosed in US Pat. No. 5,479,447, in an ADSL modem on the subscriber side, the SNR (Signal to Signal) of each of a plurality of carriers is used. Ratio (signal-to-noise ratio) is measured (step S21 in FIG. 12), and the bit allocation to each carrier is obtained according to the SNR measured for each carrier. That is, as described above, the modem on the receiving side determines the number of bitmaps for each carrier (see the received bitmap 150 in FIG. 10).
[0031]
The station-side device and the subscriber-side device measure the ratio of the modulated signal to the noise (SNR) of the line during training for communication, and determine the number of bits to be transmitted on each modulated carrier. For a carrier having a large SNR, a larger number of transmission bits is allocated, and where the SNR is small, a smaller number of transmission bits is allocated. As a result, the receiving side creates a bitmap indicating the number of transmission bits corresponding to the carrier number from the measured SNR. The receiving side notifies the transmitting side of this bitmap during training, so that the transmitting and receiving sides can perform modulation and demodulation using the same bitmap during steady data communication.
[0032]
As described above, the bit allocation (the number of bitmaps) to each of the plurality of carriers is obtained from the SNR of each of the plurality of carriers. Is determined (step S22 in FIG. 12).
[0033]
That is, although not shown in FIG. 12, the timing of the sampling clock 506 is reproduced using the pilot tone # 64. At this time, in FIG. 11, among signals (# 1 to 255) corresponding to one symbol output from FEQ 120, a signal corresponding to pilot tone # 64 is supplied to phase error detector 145. As a result, timing reproduction is performed using pilot tone # 64.
[0034]
As described above, while the timing reproduction using the pilot tone # 64 is being performed, as shown in step S21, the SNR of each carrier is measured. Next, in step S22, the maximum transmission amount R1 is calculated based on the measurement result of the SNR of each carrier in step S21.
[0035]
By the way, as shown in FIGS. 13A and 13B, when the SNR of pilot tone # 64 is poor, timing reproduction (sampling clock synchronization) using pilot tone # 64 is not performed well. Sampling jitter occurs.
[0036]
FIG. 13B shows a state in which, when the QAM signals (one value) of the pilot tones # 64 of a plurality of symbols are overlaid and plotted, the plot points are not stable (concentrated) at one point but have variations. I have.
[0037]
The state in which the sampling jitter is occurring is a state in which the receiving side has not been able to sample at the same period as the transmitting side.
[0038]
As a result of the occurrence of the sampling jitter, even if the original data carrier (the carrier other than the pilot tone # 64) does not have a low SNR, as shown in FIG. In a QAM signal (four values), the constellation (signal arrangement) is not stabilized in the phase direction, and the SNR deteriorates. When the SNR of the data carrier is bad, the maximum transmission amount R1 decreases.
[0039]
[Problems to be solved by the invention]
It is desired that the maximum transmission amount is large.
It is desired that sampling timing reproduction can be reliably performed.
It is desired that the sampling clock be synchronized between the transmitting side and the receiving side.
It is desired that generation of sampling jitter can be suppressed.
[0040]
An object of the present invention is to provide a DMT modem having a large maximum transmission amount and a timing recovery method thereof.
It is another object of the present invention to provide a DMT modem and a timing recovery method capable of reliably recovering sampling timing.
Still another object of the present invention is to provide a DMT modem that can synchronize a sampling clock between a transmitting side and a receiving side, and a method for reproducing the timing thereof.
Still another object of the present invention is to provide a DMT modem that can suppress occurrence of sampling jitter and a timing recovery method thereof.
[0041]
[Means for Solving the Problems]
Hereinafter, [Means for Solving the Problems] will be described using the numbers and symbols used in [Embodiments of the Invention]. These numbers and symbols are added to clarify the correspondence between the description of [Claims] and the description of [Embodiments of the Invention]. It should not be used to interpret the technical scope of the described invention.
[0042]
The DMT modem according to the present invention is a DMT (Discrete Multi-Tone) modem (600) for receiving a signal transmitted from a transmitter, the DMT modem being based on a signal-to-noise ratio of a signal received from the transmitter. The signal used to synchronize the clock (506) of the modem (600) with the clock of the transmitter is selected.
[0043]
The DMT modem of the present invention is a DMT (Discrete Multi-Tone) modem (600) for receiving a pilot tone (# 64) transmitted from a transmitter, wherein the DMT modem (600) is configured to receive the received pilot tone (# 64). When the clock (506) of the DMT modem (600) is synchronized with the clock of the transmitter using a tone (# 64), the reception data transmitted from the transmitter and received by the DMT modem (600) is transmitted. When the clock (506) of the DMT modem (600) is synchronized with the clock of the transmitter using the first signal-to-noise ratio and the timing recovery signal (C) instead of the pilot tone (# 64) And a second signal-to-noise ratio of the received data transmitted from the transmitter and received by the DMT modem (600). Then, the signal used when synchronizing the clock (506) of the DMT modem (600) with the clock of the transmitter is determined as the pilot tone (# 64) or the timing recovery signal (C). Here, the timing reproduction signal (C) corresponds to the subcarrier (C).
[0044]
The DMT modem according to the present invention is a DMT (Discrete Multi-Tone) modem (600) for receiving a pilot tone (# 64) transmitted from a transmitter, wherein the DMT modem (600) includes the pilot tone (#). 64) or based on the phase difference between the timing recovery signals (C), the DMT modem clock is synchronized with the transmitter clock, and the pilot tone (# 64) or the timing recovery signal (C) is synchronized. And a phase error detector (155) for detecting the phase difference, and a signal-to-noise ratio measurement of each carrier for one symbol of the signal transmitted from the transmitter. The maximum transmission amount (R1, R2) to the DMT modem is determined, and the key used for timing reproduction is determined based on the maximum transmission amount (R1, R2). A carrier selection unit (157) for selecting a carrier, a carrier identification signal indicating the selected carrier, and a signal output to the phase error detector (155) based on the carrier identification signal. (# 64) or a selector (159) for determining the timing reproduction signal (C).
[0045]
The timing recovery method for a DMT modem according to the present invention comprises the steps of: (a) using a pilot tone (# 64) transmitted from a transmitter and received by a receiver (600), using a clock (506) of the receiver (600); And (b) a first signal of received data transmitted from the transmitter and received by the receiver (600) when (a) is being performed. Calculating a noise-to-noise ratio (S11); and (c) a signal-to-noise ratio of the signal corresponding to the pilot tone (# 64) included in the first signal-to-noise ratio data and the pilot tone (#). (D) determining a timing reproduction signal (C) in place of the pilot tone (# 64) based on the signal-to-noise ratio of a signal corresponding to a signal other than (d); Synchronizing the clock (506) of the receiver (600) with the clock of the transmitter using the signal for reproducing (C); and (e) performing the operation (d) when the operation (d) is performed. (S16) determining a second signal-to-noise ratio of the received data transmitted from the transmitter and received by the receiver (600); and (f) comparing the second signal-to-noise ratio with the second signal-to-noise ratio. Synchronizing the clock (506) of the receiver (600) with the clock of the transmitter using the timing recovery signal (C) even during steady data communication when the signal-to-noise ratio is good. And Further, when the first signal-to-noise ratio is better than the second signal-to-noise ratio, the clock (506) of the receiver (600) is transmitted using the pilot tone (# 64). Synchronizing with the machine clock.
[0046]
(G) receiving a pilot tone (# 64) transmitted from a transmitter by a receiver (600), and (h) receiving a pilot tone (# 64) transmitted from a transmitter by the receiver (600). Synchronizing the clock (506) of the receiver (600) with the clock of the transmitter using the received pilot tone (# 64); and (i) when the (h) is performed. (S11) determining a first signal-to-noise ratio of received data transmitted from the transmitter and received by the receiver (600); and (j) based on the first signal-to-noise ratio. (S12) determining a first maximum transmission amount (R1) when data is transmitted from the transmitter to the receiver (600); and (k) calculating the first signal-to-noise ratio data. The par included For timing reproduction instead of the pilot tone (# 64), based on the signal-to-noise ratio of the signal corresponding to the lot tone (# 64) and the signal-to-noise ratio of the signal corresponding to other than the pilot tone (# 64). Determining a signal (C) (S14); and (l) synchronizing a clock (506) of the receiver (600) with a clock of the transmitter using the timing recovery signal (C). (M) obtaining a second signal-to-noise ratio of received data transmitted from the transmitter and received by the receiver (600) when the (l) is being performed (S16); (N) determining a second maximum transmission amount (R2) when data is transmitted from the transmitter to the receiver (600) based on the second signal-to-noise ratio (S17); , (O) When the second maximum transmission amount (R2) is larger than the first maximum transmission amount (R1), the clock of the receiver (600) is used by using the timing reproduction signal (C). Synchronizing (506) with the clock of the transmitter. Further, when the first maximum transmission amount (R1) is larger than the second maximum transmission amount (R2), the clock (506) of the receiver (600) is used by using the pilot tone (# 64). Is synchronized with the clock of the transmitter.
[0047]
In the timing recovery method for a DMT modem according to the present invention, in (n), two bits are allocated to the timing recovery signal (C). Here, in the above (n), the reason why two bits are arranged in the timing reproduction signal (C) is that when the second signal-to-noise ratio is obtained in the above (m), the two bits are arranged. Since the measurement is performed using the random signal described above, there is no change in jitter between the measurement of the second signal-to-noise ratio and the data transmission of (n) (referred to as shottime).
[0048]
In the timing recovery method for a DMT modem according to the present invention, in (c) or (k), the timing recovery signal (C) is determined in consideration of a low carrier frequency.
[0049]
In the timing recovery method for a DMT modem according to the present invention, (p) when the code error rate when steady data communication is performed is high, the steps (a) to (f) or the step (g) is performed. ) To (o) are performed again.
[0050]
BEST MODE FOR CARRYING OUT THE INVENTION
Hereinafter, an embodiment of a timing recovery method for a discrete multi-tone (DMT) modem according to the present invention will be described with reference to the accompanying drawings. In the present embodiment, the same components as those in the related art are denoted by the same reference numerals, and detailed description thereof will be omitted.
[0051]
In the present embodiment, by selecting a carrier used for sampling timing reproduction after measuring the SNR, it is possible to increase the maximum transmission amount as compared with the related art.
[0052]
FIG. 1 is a block diagram showing the configuration of the present embodiment. In FIG. 1, the same components as those in FIG. 11 showing the conventional configuration are denoted by the same reference numerals. As shown in FIG. 1, the DMT (CPE) modem 600 according to the present embodiment includes an SNR measurement and carrier selection unit 157 and a selector 159.
[0053]
The SNR measurement and carrier selection unit 157 measures the SNR of each carrier of the signal (# 1 to 255) corresponding to one symbol output from the gain adjustment unit 125, obtains the maximum transmission amounts R1 and R2, and obtains the maximum transmission amounts R1 and R2. A carrier to be used for timing recovery is selected based on the transmission amounts R1 and R2. The SNR measurement and carrier selection unit 157 outputs the number of the selected carrier (# 64 or #C) to the selector 159.
[0054]
The selector 159 inputs the number (carrier number) of the carrier selected by the SNR measurement and carrier selection unit 157, and outputs the signals (# 1 to 255) output from the FEQ 120 from the SNR measurement and carrier selection unit 157. A signal corresponding to the input carrier number is supplied to the phase error detector 155.
[0055]
The phase error detector 155 measures the phase error between the current and previous (continuously multiple) received signals based on the signals corresponding to the selected carrier number. The signal corresponding to the selected carrier number is a signal subjected to quadrature amplitude modulation (QAM) as described above, and is represented as a complex number (that is, a real component and an imaginary component).
[0056]
Thus, the signal corresponding to the selected carrier (number) is shown as a vector when plotted on the XY plane. When the receiver clock is not synchronized with the remote transmitter clock, the vector of signals corresponding to the selected carrier number will rotate over time. The phase error detector 155 gives a phase difference (shift, error) between the symbols of the signal corresponding to the selected carrier to the PLL unit 165 as a phase error.
[0057]
The sampling clock 506 is set as a clock of the A / D converter 80. When a phase error is input from the phase error detector 155, the PLL unit 165 adjusts the sampling clock 506.
[0058]
Next, a flow of reproducing the sampling timing of the ADSL modem according to the present embodiment will be described with reference to FIG. The flow in FIG. 2 may be executed using either software or hardware. This flow corresponds to FIG. 12 showing a conventional flow.
[0059]
First, although not shown in FIG. 2, the timing of the sampling clock 506 is reproduced using the pilot tone # 64, as in the related art. At this time, the selector 159 of FIG. 1 supplies a signal corresponding to the pilot tone # 64 to the phase error detector 155 among the signals (# 1 to 255) output from the FEQ 120. As a result, timing reproduction is performed using pilot tone # 64.
[0060]
As described above, while the timing reproduction using the pilot tone # 64 is being performed, the SNR of each carrier is measured by the SNR measurement and carrier selection unit 157 as shown in step S11.
[0061]
Next, in step S12, the SNR measurement and carrier selection unit 157 calculates the maximum transmission amount R1 based on the measurement result of the SNR of each carrier in step S11.
[0062]
Next, as shown in step S13, based on the SNR of each carrier measured in step S11, the SNR measurement and carrier selection unit 157 changes the SNR of pilot tone # 64 to the SNR of carriers other than pilot tone # 64. It is determined whether the comparison is bad. In this case, specifically, it is determined whether or not the SNR of pilot tone # 64 is worse than many SNRs of carriers other than pilot tone # 64.
[0063]
As a result of step S13, when the SNR measurement and carrier selection unit 157 determines that the SNR of pilot tone # 64 is worse than the SNR of carriers other than pilot tone # 64 (S13-Y), the process proceeds to step S14. move on.
[0064]
On the other hand, as a result of step S13, when the SNR measurement and carrier selection unit 157 determines that the SNR of pilot tone # 64 is not worse than the SNR of carriers other than pilot tone # 64 (S13-N) Ends the flow of FIG. 2 and keeps the carrier used for timing recovery as pilot carrier # 64.
[0065]
In step S14, the SNR measurement and carrier selection unit 157 determines a carrier number candidate #C used for timing recovery. In step S14, the SNR measurement and carrier selection unit 157 selects a carrier number candidate C from carriers having a better SNR than # 64 (pilot).
[0066]
Next, in step S15, the SNR measurement and carrier selection unit 157 changes the carrier used for timing recovery to C (the carrier corresponding to the carrier number C). That is, the SNR measurement and carrier selection unit 157 outputs the carrier number C determined in step S14 to the selector 159. The selector 159 supplies the signal (carrier C) corresponding to the carrier number C to the phase error detector 155 among the signals (# 1 to 255) corresponding to one symbol output from the FEQ 120. As a result, timing reproduction using the carrier C is performed.
[0067]
Next, while the timing reproduction using the carrier C is being performed as described above, in step S16, the SNR measurement and carrier selection unit 157 measures the SNR of each carrier.
[0068]
Next, in step S17, the SNR measurement and carrier selection unit 157 calculates the maximum transmission amount R2 based on the measurement result of the SNR of each carrier in step S16. However, the carrier C has a 2-bit arrangement (four values indicated by reference numerals (d) and (e) in FIG. 3).
[0069]
Here, the reason why the number of bits arranged on the carrier C for timing reproduction is set to 2 bits is as follows. According to the ANSI standard, 0 bits or 2 bits or more must be allocated to carriers other than # 64. Are arranged on the same circumference, it is easy to measure an interval between two points indicating a phase shift, and it is also possible to secure a large interval and to easily detect a phase shift (FIG. 6 (b)). )reference). Also, since the SNR measurement is performed using a random signal having a 2-bit arrangement, there is no change in jitter between the SNR measurement and the data transmission (referred to as shottime).
[0070]
Next, in step S18, the SNR measurement and carrier selection unit 157 compares the maximum transmission amount R1 obtained in step S12 with the maximum transmission amount R2 obtained in step S17.
[0071]
When the result of step S18 is that the maximum transmission amount R2> the maximum transmission amount R1 (S18-Y), the SNR measurement and carrier selection unit 157 ends the flow of FIG. To decide.
On the other hand, when the maximum transmission amount R2 is not larger than the maximum transmission amount R1 (S18-N), the SNR measurement and carrier selection unit 157 proceeds to step S19.
[0072]
In step S19, the SNR measurement and carrier selection unit 157 returns the carrier used for timing recovery from carrier C to pilot carrier # 64. That is, SNR measurement and carrier selection section 157 outputs carrier number # 64 of the pilot carrier to selector 159. The selector 159 supplies a signal (pilot carrier # 64) corresponding to the carrier number # 64 to the phase error detector 155 among signals (# 1 to 255) corresponding to one symbol output from the FEQ 120. Thereby, timing reproduction using pilot carrier # 64 is performed.
[0073]
As described above, conventionally, the carrier used for timing recovery is fixed to pilot carrier # 64, but in the present embodiment, the carrier number used for timing recovery is variable (# 64 or C). The effect obtained from this will be described below.
[0074]
FIGS. 3A to 3E are diagrams illustrating the effect of the present embodiment.
The waveform indicated by reference symbol B in FIG. 3A is the same as the waveform in FIG. 13A, and FIGS. 3B and 3C are FIGS. 13B and 13C, respectively. Is the same as The waveforms indicated by reference symbol A in FIG. 3A, and FIGS. 3D and 3E correspond to the case where the present embodiment is used.
[0075]
As shown in FIGS. 3B and 3C, when the SNR of the pilot tone # 64 is poor, the SNR deteriorates due to sampling jitter, and the maximum transmission amount R1 decreases.
[0076]
FIG. 3D shows a state in which, in the present embodiment, when a carrier C having a good SNR is used for timing reproduction, the QAM signals (the above 2 bits: 4 values) of the carrier C for a plurality of symbols are plotted. And each plot point is stable at one point and there is no variation. As shown in FIG. 3D, when a carrier C having a good SNR is used for timing reproduction, no sampling jitter occurs, and as a result, as shown in FIG. The signal arrangement of the QAM signal is stable, the SNR is good, and the waveform A in FIG. 3A is obtained. The waveform A indicates that the SNR of the data carrier is improved over the waveform B indicating the SNR of the data carrier when the timing is reproduced using the pilot tone # 64.
[0077]
According to the present embodiment, when the SNR of a carrier used for sampling timing reproduction is lower than other carriers due to external noise or the like, a carrier (data carrier) having a good SNR is deteriorated by sampling jitter (FIG. a) Waveform B). In this case, by changing the carrier used for sampling timing reproduction to a carrier having a good SNR, the sampling jitter is reduced and the SNR is improved (see waveform A in FIG. 3A). Thus, according to the present embodiment, an effect that the maximum transmission amount increases is obtained.
[0078]
FIG. 4 shows the results of the experiment. When the carrier number used for timing recovery is # 64 (pilot tone), the SNR deteriorates as shown by the broken line, and the maximum transmission amount R1 is 8764 kbps, whereas the carrier used for timing recovery is 8764 kbps. When the carrier with the carrier number 96 was used as C, the SNR was improved as shown by the solid line, and the maximum transmission amount R2 was 9124 kbps.
[0079]
In step S14, if the carrier used for timing reproduction is selected to have the highest SNR, the maximum transmission amount R2 calculated in step S17 is not always the maximum. A carrier having a higher SNR can allocate more bits at the time of bit allocation. Nevertheless, if a carrier having a high SNR is used for timing recovery, only two bits are allocated as described above. Therefore, it is desirable to select a carrier used for timing reproduction in consideration of this point.
[0080]
Further, in the above, as shown in FIG. 2, based on a comparison between the maximum transmission amount R1 obtained in step S12 and the maximum transmission amount R2 obtained in step S17, the carrier used for timing reproduction is # 64 or C (Steps S18 and S19), but instead of this, the carrier used for timing reproduction is determined to be # 64 or C based on a comparison between the SNR obtained in step S11 and the SNR obtained in step S16. be able to.
[0081]
That is, in the present invention, steps S12 and S17 are not essential. In the present invention, either carrier # 64 or carrier C is suitable for timing reproduction only by comparing steps S11 and S16, which reflect the magnitude of sampling jitter when carrier # 64 and carrier C are used for timing reproduction, respectively. Can be determined.
[0082]
In this case, instead of the flow of FIG. 2, S11 → S13 → S14 → S15 → S16 → (Step of comparing the SNR determined in S11 with the SNR determined in S16) → (SNR determined in S16 is determined in S11 When the SNR is not better than the SNR, the carrier is returned to # 64 for timing reproduction, and the flow is terminated when the SNR obtained in S16 is better than the SNR obtained in S11).
[0083]
Next, a second embodiment will be described.
[0084]
As shown in step S14 of FIG. 2, in the first embodiment, the SNR measurement and carrier selection unit 157 determines the candidate #C of the carrier number to be used for the timing reproduction more than the # 64 (pilot). He was choosing from carriers with good SNR. On the other hand, in the second embodiment, in the step of determining the candidate #C of the carrier number to be used for the timing reproduction in the step S14, not only the SNR but also the carrier frequency is low (the carrier number is small). In consideration of the above, the candidate #C is determined.
[0085]
Hereinafter, the reason for considering the low carrier frequency in the second embodiment will be described.
[0086]
As will be described later, when a QAM signal (1 value) of pilot tone # 64, which is input to phase error detector 155 and used for reproduction of sampling timing, has a phase shift of ± π (rad). However, if a QAM signal of carrier C other than carrier number # 64 is used, only a phase shift of less than 1 / 4π (rad) can be detected.
[0087]
Therefore, when the QAM signal of the carrier C other than the carrier number # 64 is used, the timing is reproduced as compared with the case where the QAM signal (1 value) of the carrier (pilot carrier) of the carrier number # 64 is used. The possible range, that is, the range in which synchronization can be achieved is small.
[0088]
Hereinafter, the above problem, that is, the phase error characteristic in the phase error detector 155 will be described with reference to FIGS. 5A and 5B and FIGS. 6A and 6B.
[0089]
FIG. 5B shows a QAM signal (1 value, code 201) when the QAM signal (1 value) of the pilot tone # 64 is supplied to the phase error detector 155 by the selector 159. (A) shows a phase error characteristic when the QAM signal (1 value) of the pilot tone # 64 is used.
[0090]
FIG. 6B shows QAM signals (quaternary, codes 301 to 304) when the selector 159 supplies a QAM signal of the carrier C other than the pilot tone # 64 to the phase error detector 155. FIG. 6A shows a phase error characteristic when a QAM signal (4 values) other than the pilot tone # 64 is used.
[0091]
As shown in FIG. 5B, the constellation of the point 201 in the case where there is no phase error exists at the position of 1 / 4π (rad) (reference position). For / 4π (rad), the phase error amount is zero. In FIGS. 5B and 5A, crosses (X) indicate a case where there is a 1 / 4π phase error with respect to the reference position. As shown in FIGS. 5B and 5A, when the QAM signal (1 value) of pilot tone # 64 is used in phase error detector 155, the difference between the reference position and the reference position is less than ± π. A phase error can be detected. That is, since the phase error of 2π with respect to the reference position is the same as the reference position, it cannot be determined whether there is a phase shift of 2π or the phase shift is 0.
[0092]
As shown in FIG. 6B, the constellations of points 301 to 304 in the case where there is no phase error indicate the positions of １ ／ π, ／ π, ／ π, and ／ π (rad), respectively. 6 (a), the phase error amount is 0 for 1 / 4π, 3 / 4π, 5 / 4π, and 7 / 4π (rad). As shown in FIGS. 6B and 6A, when a QAM signal (four values) other than the pilot tone # 64 is used by the phase error detector 155, 1 / 4π with respect to the reference position. It is possible to detect a phase error of less than That is, in FIG. 6B, it is not possible to distinguish between a state in which the point 301 has moved 1 / 4π in the positive direction and a state in which the point 302 has moved 1 / 4π in the negative direction.
[0093]
As shown in FIGS. 5 (a), 5 (b), 6 (a) and 6 (b), the signal used in phase error detector 155 is a QAM signal (4) other than pilot tone # 64. Value), the range in which synchronization can be achieved is narrower than in the case of a QAM signal (1 value) of pilot tone # 64.
[0094]
From the above, when the QAM signal of the carrier C other than the carrier number # 64 is used, the range in which the timing can be reproduced becomes a disadvantage, and it is desirable to compensate for this disadvantage.
In this case, as the QAM signal of the carrier C with a higher carrier frequency (larger carrier number) is used, the amount of rotation (rotation angle) of the phase per sample becomes larger, and conversely, the carrier C with a lower carrier frequency becomes larger. As the QAM signal is used, the amount of phase rotation (rotation angle) per sample becomes smaller.
[0095]
Here, the above points will be described.
As shown in FIG. 7, “sample” refers to a sampling cycle (= reciprocal of the sampling frequency) of the AD converter, and “one sample” refers to a time corresponding to one sampling cycle. In the present embodiment, the sampling period is 1 / 2.208 MHz.
FIG. 8 shows the 0 to 30 data of the 512-sample time domain data (# 0 to 511) input to the FFT when the carrier signal is only # 64. In the case of the frequency of # 64, the sin wave oscillates 64 times in 512 samples, so that eight samples correspond to one cycle of the sin wave.
In FIG. 8, the solid line is an analog signal having a phase of π / 4 (rad) when viewed in the frequency domain. The symbol Δ indicates a digital signal when the data indicated by the solid line is sampled. In addition, the mark △ indicates that the phase in the time domain is shifted by one sample (# 64 is 8 samples and one period (2π), 2π / 8 = π / 4) with respect to the mark △. It is a digital signal at the time. Looking at this in the frequency range after FFT. The signal point is located at 0 (rad).
When viewed on the complex plane in the frequency domain, the phase per sample in the time domain is (Nπ) / 256 depending on the carrier number N. In the case of # 64, (64π) / 256 = π / 4.
From this, as the QAM signal of the carrier C having a higher carrier frequency (larger carrier number) is used, the amount of phase rotation (rotation angle) per sample is larger, and the QAM signal of the carrier C having a lower carrier frequency is increased. It can be seen that the phase rotation amount (rotation angle) per sample is smaller as it is used.
[0096]
Therefore, in the second embodiment, when the candidate #C of the carrier number used for the timing reproduction is determined in step S14, the condition in the first embodiment that the SNR is better than # 64 (pilot) is satisfied. In addition, the carrier number C is determined also on condition that the carrier frequency is low. As described above, by selecting the carrier C having a low carrier frequency, the small range in which the timing when the QAM signal of the carrier C other than the carrier number # 64 is used can be reproduced is compensated.
[0097]
Further, as described above with reference to FIGS. 5A, 5B, 6A, and 6B, the phase error characteristic indicates that the signal used in the phase error detector 155 is , QAM signal (4 values) other than pilot tone # 64 or QAM signal (1 value) of pilot tone # 64. Therefore, the phase error detector 155 inputs a carrier number used for timing recovery from the SNR measurement and carrier selection unit 157, and sets conditions in the phase error detector 155 according to the input value (this condition setting). Is necessary not only in the second embodiment but also in the first and third embodiments.)
[0098]
Next, a third embodiment will be described.
[0099]
In the first or second embodiment, the timing is reproduced using the carrier C other than the pilot tone (# 64) or # 64, and synchronization is achieved, but the initialization sequence (FIG. 43 of ANSI T1.413-1998). In the third embodiment, when there are many errors in the code error rate of the received data in the steady data communication performed after the above, a step of restarting the initialization sequence is performed.
[0100]
That is, in spite of the fact that the timing is reproduced and the synchronization is established, the error in the code error rate of the received data is large, as shown in FIG. 5B and FIG. It is conceivable that the motor rotates from the reference position to a point that cannot be detected as a phase error, and is erroneously determined to be synchronized at that position and locked.
[0101]
Specifically, as shown in FIG. 5B, when the pilot tone (# 64) is used, and when the phase error from the reference position is 2π (rad), the phase error is 0 (rad). rad), it may be erroneously determined that synchronization is achieved at that position, and the case may be locked.
As shown in FIG. 6B, when a carrier C other than the pilot tone (# 64) is used, and when the phase error from the reference position is π / 2 (rad), the phase error Is 0 (rad), it may be erroneously determined that synchronization is achieved at that position and locked.
[0102]
In the third embodiment, when there are many errors in the code error rate of received data despite the fact that synchronization is achieved by performing timing reproduction, the initialization sequence is immediately (actively) restarted to perform timing reproduction. Can be quickly eliminated.
[0103]
As described above, when the signal used in phase error detector 155 is a QAM signal (1 value) of pilot tone # 64, synchronization is established when there is a 2π (rad) phase error. On the other hand, when a QAM signal (four-valued) other than pilot tone # 64 is used, it is determined that synchronization is achieved when there is a phase error of π / 2 (rad). Misjudgment is made. From this, it is considered that the erroneous determination is more likely to occur when a QAM signal (four values) other than the pilot tone # 64 is used. Therefore, the third embodiment is more effective when timing reproduction is performed using a QAM signal (four values) other than pilot tone # 64.
[0104]
【The invention's effect】
According to the present invention, the maximum transmission amount can be increased.
[Brief description of the drawings]
FIG. 1 is a block diagram showing a configuration of an ADSL modem in a first embodiment of a DMT modem timing reproduction method according to the present invention.
FIG. 2 is a flowchart showing the operation of the ADSL modem in the first embodiment of the DMT modem timing recovery method according to the present invention.
FIG. 3 is a diagram showing the effect of the first embodiment of the timing recovery method for a DMT modem according to the present invention.
FIG. 4 is a graph showing experimental results showing the effects of the first embodiment of the DMT modem timing recovery method according to the present invention.
FIG. 5 is a diagram for explaining a phase error characteristic in a second embodiment of the DMT modem timing recovery method according to the present invention.
FIG. 6 is another diagram for explaining a phase error characteristic in the second embodiment of the timing recovery method of the DMT modem according to the present invention.
FIG. 7 is a diagram for explaining a sample in the second embodiment of the timing recovery method of the DMT modem according to the present invention.
FIG. 8 is another diagram for explaining the relationship between the carrier number of the carrier used for timing recovery and the amount of phase rotation per sample in the second embodiment of the timing recovery method of the DMT modem according to the present invention. It is.
FIG. 9 is a block diagram showing the principle of the DMT modulation method.
FIG. 10 is a block diagram illustrating a configuration of an ADSL transceiver.
FIG. 11 is a block diagram showing an example of a conventional CPE modem.
FIG. 12 is a flowchart illustrating an operation for calculating a maximum transmission amount in a conventional DMT-based multicarrier transmission system.
FIG. 13 is a diagram illustrating a problem when the SNR of pilot tone # 64 is poor in a conventional DMT-based multicarrier transmission system.
[Explanation of symbols]
5 Transmission data
10 SP buffer
20 Encoder
30 Inverse Fourier transform unit (IFFT)
40 Cyclic prefix addition section
41 PS buffer
50 D / A converter
60 Transmission bitmap
70 subscriber line
75 Receive Filter
80 A / D converter
90 TEQ
100 Cyclic prefix deletion section
101 SP buffer
110 FET
120 FEQ
125 gain adjustment unit
130 decoder
140 PS buffer
145 phase error detector
146 Loop filter
147 VCO
150 Receive bitmap
155 phase error detector
157 SNR measurement and carrier selection unit
159 selector
165 PLL section
500 CPE modem
506 sample clock
600 DMT modem

Claims (8)

A DMT (Discrete Multi-Tone) modem for receiving a signal transmitted from a transmitter,
A DMT modem for selecting a signal to be used when synchronizing the clock of the DMT modem to the clock of the transmitter based on a signal to noise ratio of a signal received from the transmitter. A DMT (Discrete Multi-Tone) modem for receiving a pilot tone transmitted from a transmitter,
The DMT modem uses the received pilot tone to synchronize the clock of the DMT modem with the clock of the transmitter, and transmits a first one of received data transmitted from the transmitter and received by the DMT modem. Received data transmitted from the transmitter and received by the DMT modem when the clock of the DMT modem is synchronized with the clock of the transmitter using a signal-to-noise ratio and a timing recovery signal instead of the pilot tone. A DMT modem that determines a signal to be used when synchronizing a clock of the DMT modem with a clock of the transmitter to the pilot tone or the signal for timing recovery based on the second signal-to-noise ratio. A DMT (Discrete Multi-Tone) modem for receiving a pilot tone transmitted from a transmitter,
The DMT modem synchronizes a clock of the DMT modem with a clock of the transmitter based on a phase difference between the pilot tones or a signal for timing recovery,
A phase error detector that receives the pilot tone or the timing reproduction signal and detects the phase difference;
A maximum transmission amount from the transmitter to the DMT modem is determined based on a measurement result of a signal-to-noise ratio of each carrier for one symbol of a signal transmitted from the transmitter, and a timing is determined based on the maximum transmission amount. A carrier selection unit for selecting a carrier to be used for reproduction,
A DMT which receives a carrier identification signal indicating the selected carrier, and determines a signal to be output to the phase error detector based on the carrier identification signal, as the pilot tone or the timing reproduction signal. modem. (A) using a pilot tone transmitted from a transmitter and received at a receiver to synchronize the clock of the receiver with the clock of the transmitter;
(B) determining a first signal-to-noise ratio of received data transmitted from the transmitter and received by the receiver when (a) is being performed;
(C) the pilot based on a signal-to-noise ratio of a signal corresponding to the pilot tone and a signal-to-noise ratio of a signal other than the pilot tone included in the first signal-to-noise ratio data; Determining a timing reproduction signal instead of a tone;
(D) synchronizing the clock of the receiver with the clock of the transmitter using the timing recovery signal;
(E) obtaining a second signal-to-noise ratio of received data transmitted from the transmitter and received by the receiver when the step (d) is performed;
(F) when the second signal-to-noise ratio is better than the first signal-to-noise ratio, use the timing recovery signal to transmit the clock of the receiver even during steady data communication. Synchronizing with a clock of a personal computer, a method of reproducing timing of a discrete multi-tone (DMT) modem. (G) receiving at the receiver a pilot tone transmitted from the transmitter;
(H) synchronizing the clock of the receiver with the clock of the transmitter using a pilot tone received at the receiver;
(I) obtaining a first signal-to-noise ratio of received data transmitted from the transmitter and received by the receiver when the (h) is performed;
(J) determining a first maximum transmission amount when data is transmitted from the transmitter to the receiver based on the first signal-to-noise ratio;
(K) based on a signal-to-noise ratio of a signal corresponding to the pilot tone included in the first signal-to-noise ratio data and a signal-to-noise ratio of a signal other than the pilot tone. Determining a timing reproduction signal instead of a tone;
(L) using the timing recovery signal to synchronize the clock of the receiver with the clock of the transmitter;
(M) determining a second signal-to-noise ratio of received data transmitted from the transmitter and received by the receiver when (l) is performed;
(N) determining a second maximum transmission amount when data is transmitted from the transmitter to the receiver based on the second signal-to-noise ratio;
(O) synchronizing the clock of the receiver with the clock of the transmitter using the timing recovery signal when the second maximum transmission amount is larger than the first maximum transmission amount; A timing recovery method for a DMT (Discrete Multi-Tone) modem, comprising: The timing recovery method for a DMT modem according to claim 5,
In the above (n), a timing recovery method for a DMT modem, wherein 2 bits are arranged in the timing recovery signal. The timing recovery method for a DMT modem according to any one of claims 4 to 6,
In the above (c) or (k), a timing recovery method for a DMT modem in which the timing recovery signal is determined in consideration of a low carrier frequency. The timing recovery method for a DMT modem according to any one of claims 4 to 7,
Furthermore,
(P) When the code error rate when the steady data communication is performed is high, the step (a) to the step (f) or the step (g) to the step (o) is performed again. Reproduction method of the provided DMT modem.

Images