JPH01194622A - Data speed setting system - Google Patents

Abstract

PURPOSE:To always transmit data even when much noise or distortion exist on a transmission path by setting a data speed according to the size of the equalizing residual quantity (error signal) of an automatic equalizer at the time of training completion. CONSTITUTION:When a training signal is sent from a modem 15, the training signal through the transmission path is received by a modem 16, tap updating is carried out by an automatic equalizer 2, and the training is carried out. A rate sequence is generated in a rate sequence generating circuit 10 so that a rate sequence may be set at the signal at the higher data speed, when the output signal in a averaging circuit 6 is smaller than a threshold value TH 8 at the time of the training completion, and the rate sequence may be set at the signal at the lower data speed in a contrary case. the generated rate sequence is constituted as the training signal by a training signal generating circuit 11, and sent to the transmission path through a modulating part 12. The sent training signal is received by the modem 15, and in the same way, the training of an automatic equalizer 2 is carried out.

Description

[Detailed Description of the Invention] [Industrial Application Field] The present invention relates to a modulation/demodulation device (hereinafter referred to as a modem) that includes data rate information in a training signal for training an automatic equalizer. This relates to a data rate setting method.

[Conventional technology]

Conventionally, in a modem having an automatic equalizer, before data transmission, the automatic equalizer in the receiving section of the own station is set up by receiving a training signal from the modem of the other station. This training means creating the opposite characteristics of distortion, such as amplitude distortion and phase distortion, in the transmission path between the other station and the own station, and equalizing the signal that is distorted by passing through the transmission path.

That is, in FIG. 2, the received demodulated signal supplied to the demodulated signal input terminal 1 is passed through the automatic equalizer 2 to the error calculation circuit 3.
is input. Here, the error between the reference generated from the reference generation circuit 4 and the output of the automatic equalizer 2 is calculated. Then, the tap coefficient updating circuit 5 updates the internal tap coefficient value C of the automatic transmission line 2 from the demodulated signal and the error signal to form an inverse characteristic of the transmission path. The algorithm for this tap update is shown in "Digital Signal Processing" (edited by the Institute of Electronics, Information and Communication Engineers, pp. 235-242), and a typical example is as follows.

Here, each symbol is as follows.

X(k-j)' By continuing to update the demodulated signal taps, the error signal ek gradually becomes smaller and becomes ek#O as shown by symbol a in FIG. 5, and equalization is completed.

As shown in FIG. 3, before data transmission, for example, a training signal is sent from the modem 15 of the local station, and is received by the modem 16 of the opposite station through a transmission path, thereby setting up an automatic equalizer. Also, the partner station modem 16
Then, upon receiving this training signal, the training signal is sent out, and automatic setup of the local station modem 15 is performed in the same way, and from then on, the computer or terminal device (hereinafter referred to as the terminal device) 14 is set up. 1
7 allows data transmission.

However, if the noise in the transmission path is large or the distortion is large, the error signal e output from the error calculation circuit 3
No matter how many taps are updated, m will not become ek''o, as indicated by the symbol in FIG. 5. This means that it is not completely equalized, and the larger the error signal ek, the more
The bit error rate during data transmission increases. If the error signal ek is larger than a certain value, there may be too many bit errors and data transmission may become impossible.

In FIG. 2, 9 is an SR3 signal input terminal, 10 is a rate sequence generation circuit, 11 is a training signal generation circuit,
12 is a modulation section.

[Problem to be solved by the invention]

However, such conventional methods have the problem that data transmission becomes impossible when noise and distortion in the transmission path are large. The following method can be used to remedy this problem. If there is an error in the data between the terminal device 1417, after retransmitting the data several times according to the communication control procedure, if the error still occurs in the data, the terminal device 14.17 and the modem 15.16 SR3 between
(Signaling Rate 5 select)
By lowering the data transmission rate of modem 1516 with the signal, data transmission is resumed with a lower bit error rate.

For example, the bit error rate can be reduced by lowering the data transmission speed.
CITT) V.33 compliant modems have a data rate of 14.
There are 4kb/s and 12kb/s, 14, 4kb
/s is a 12B value orthogonal amplitude modulation method, and 12kb/S is a 14. This method has a large margin against transmission noise and distortion compared to 4 kb/s. Therefore, with the same amount of noise and distortion, the bit error rate is lower in the case of 12 kb/s than in the case of 14.4 kb/s. However, this method has problems in that it takes time to change the data transmission rate, and the control method is complicated.

[Means to solve the problem]

In order to solve this problem, the present invention reduces the rate sequence in the training signal to a lower data transmission rate unless the value of the error signal becomes less than a certain value at the end of training of the automatic equalizer. The modem that has been configured and has received this training signal similarly sends out the rate sequence in the training signal that it sends out as low data rate information.

[Effect]

Since the data transmission rate is determined when the training signal is transmitted and received before data transmission, no extra time is required after the start of communication.

〔Example〕

FIG. 1 is a block diagram showing one embodiment of the present invention.

Before data transmission, training signals are sent and received between modems. For example, a case where a training signal is sent from modem 15 will be explained. The training signal that has passed through the transmission path is received by the modem 16, and taps are automatically updated in the second mode to perform training. At this time, the error signal ek output from the error calculation circuit 3 is input to the tap coefficient updating circuit 5, and is also input to the averaging circuit 6 to be averaged. At the end of the training of the automatic equalizer 2, the comparison circuit 7 compares and determines the averaging circuit 6 and the threshold value (TH) 8. As a result, if the output signal of the averaging circuit 6 is smaller than the threshold value (TH) 8, the rate sequence is set to a signal with a high data rate, and vice versa, the rate sequence is set to a signal with a low data rate. A rate sequence is generated in a sequence generation circuit 10.

In this case, the SR3 signal is ignored. Here, the training signal is, for example, CCIT
T) According to the recommendation of V.33, as shown in Figure 4,
It is composed of four types of signals up to (1), and data rate information is included in the rate sequence (3). The generated rate sequence is configured as a training signal by the training signal generation circuit 11, and sent to the transmission path through the modulation section 12. This transmitted training signal is received by the modem 15 and similarly trains the automatic equalizer 2.

The modem 15 sets the data rate according to the rate sequence of the received training signal. (Setting the modem's own data rate according to the rate sequence of the training signal is done in a known manner).

If this data rate matches the data rate in the training signal previously sent by the modem 15, a rate sequence matching the data rate of that rate sequence is generated, and the training signal containing it is transmitted to the modem 15.
Send to.

Modem 16 receives this training signal and sets it to the data rate indicated by the rate sequence therein.

After that, data transmission begins. By doing this, the data rate of the modem 15.16 is set without going through the communication control procedure of the terminal device 14.17.

Here, if the data rate of the rate sequence of the modem 16 does not match the data in the training signal received from the modem 15, the data after making it match is sent from the modem 16 to the modem 15. As a result, both sides are set to a data rate suitable for the line conditions before starting communication.

〔Effect of the invention〕

As explained above, in this invention, the data rate is set depending on the level of the equalization residual amount (error signal) of the automatic equalizer at the end of training, so there is a lot of noise and distortion in the transmission path. This has the effect that data transmission will not become impossible even in the case of

[Brief explanation of the drawing]

FIG. 1 is a block diagram showing an embodiment of the present invention. FIG. 2 is a block diagram showing an example of a conventional device. FIG. 3 is a modem system configuration diagram. FIG. 4 is a training signal configuration diagram. is a graph showing the state of the error signal during automatic similitude training. 1... Demodulated signal input terminal, 2... Automatic equalizer,
3...Error calculation circuit, 4...Reference circuit, 5...Tap update circuit, 6...Averaging circuit,
7... Comparison circuit, 8... Lelevhold value, 9
...SR8 signal input terminal, 10...rate sequence generation circuit 11...training signal generation circuit, 12...modulation section.

Claims (1)

[Claims] In a modulation/demodulation device having a built-in automatic equalizer and including data transmission rate information in a training signal for training the automatic equalizer, an error signal of an output of the automatic equalizer at the end of training. A training signal containing data transmission rate information configured according to the magnitude of the training signal is transmitted, and the modem that receives this training signal determines the data rate based on the data transmission rate information in the received training signal, and the data rate is the training signal. A data rate setting method characterized in that data rates of mutual modulation and demodulation devices are set by transmitting a training signal that includes data rate information that is the same as data rate information of a received training signal in a case different from that before the signal was received.