JPS6347306B2 - - Google Patents

Description

[Detailed description of the invention]

(a) Technical Field of the Invention The present invention relates to a training detection method for improving the instantaneous interruption tolerance characteristics when a receiving carrier interruption occurs during training of a data modem that has a function of compensating for the characteristics of a transmission path. It is. (b) Background of the technology On the receiving side of data communication, when a momentary interruption of the receiving carrier occurs, it is possible to improve momentary interruption tolerance so that valid data can continue to be received without invalidating the data after the momentary interruption has been recovered. It is requested. (c) Prior art and problems In data transmission using a telephone line, the data signal is A specific training signal is transmitted prior to the start of transmission. The format of this training signal is shown in the table below (according to CCITT document COM.SP.A-No. 148). Here, "S" means a segment, and "total" means the entire training signal.

[Table] Segment 1 of the training signal does not contain any signal, but the transition point with segment 2 is used to detect the start of data reception. Segment 2 of the training signal consists of repetitions of two signal elements. As shown in FIG. 1, the first signal element A has a relative amplitude of 3 and defines an absolute phase reference of 180 degrees. The second signal element B changes with the data rate, (1-j) at 7200 bit/sec, 9600 bit/sec
When , there are 3 (1-j) vectors. Segment 2 is ABAB during the 128 symbol interval...
Repeat... Training signal segment 3
consists of two signal elements that vary according to an equalizer conditioning pattern, also called a PN code. The first signal element C has a relative amplitude of 3 and an absolute phase of 0°, and the second signal element D varies with the data rate, at 7200 bit/sec (-1 + j),
At 9600 bit/sec, there are 3 (-1+j) vectors. The equalizer conditioning pattern is a pseudo-random sequence, and when this sequence is 0, C occurs, and when this sequence is 1, D occurs. Furthermore, the signal of segment 4 is used to pull in the descrampler circuit. This segment 2 signal includes a demodulated carrier. That is, vectors A and B are vectors
It can be considered as the sum of A' and CAR and the sum of vector B' and CAR, and segment 2 is A, B, A, B
When iterating, the invariant component is CAR, and the alternating component is
A' and B', the sum of alternating components A' and B' is zero, and the constant component CAR is the carrier signal. Similarly, when calculating the invariant component and therefore the carrier component for segment 3, vectors C and D
are vector A′ and CAR′ and vector B′ and
Since it is the sum of CAR′, the carrier component is CAR′, which is the carrier component CAR of segment 2.
The phase is shifted by 180°. When pulling in the equalizer (automatic equalizer), the transmitting side sends out the signal of segment 3, and the receiving side also generates the same signal, and for example, compares the two to find the signal necessary for waveform correction. To carry out such an operation, it is necessary to detect the start point of segment 3 and to start generating the above-mentioned signal on the receiving side. Conventionally, when a failure occurs due to the influence of a transmission path and a momentary interruption of the receiving carrier occurs, the following two methods have been considered as methods for improving resistance to momentary interruption. That is, (a) a method in which training detection is not performed, and (b) a method in which training detection is performed have been implemented. (b) A method that does not perform training detection is a method that deals with this by extending the OFF time of the character detector. This conventional method has ideal instantaneous interruption tolerance for short instantaneous interruptions, but it is ineffective against long interruptions because of constraints on the time in the NET area. Furthermore, (b) if the training detection method is implemented, it will be effective against long instantaneous interruptions, but an offset will occur against short instantaneous interruptions. FIG. 2 is a diagram showing the characteristic relationship between the instantaneous interruption time and the error occurrence time in the conventional methods (a) and (b). (d) Object of the Invention The object of the present invention is to eliminate the above-mentioned drawbacks, and to provide a training detection method for a data modem that minimizes data that becomes invalid due to instantaneous interruptions and improves instantaneous interruption tolerance characteristics. The purpose is to (e) Configuration of the invention: a first delay circuit that delays a received signal by one signal element via a waveform shaping filter; and an adder circuit that adds an input signal that does not pass through the first delay circuit;
a second delay circuit that delays the output of the adder circuit by 180 degrees; the output of the second delay circuit is input to a multiplier circuit; and the output of the second adder circuit is input to the second delay circuit;
input to the multiplication circuit without passing through the delay circuit,
This is achieved by providing a training detection method for a data modem, characterized in that the output effect of the multiplier circuit is input to sign determination via a tapped delay element to detect whether it is a training signal or data. (f) Examples of the invention Examples of the invention will be described in detail below with reference to the drawings. FIG. 3 is a sequence diagram of training signals used in the present invention. The contents of each of the training signal segments SEG1 to SEG4 are as explained in FIG. 1. FIG. 4 is a circuit diagram showing an embodiment of a data modem training detection method to which the present invention is applied. In the drawing, ROF is a waveform shaping filter that shapes the waveform of the input signal, T is a delay element, + is an addition circuit, × is a multiplication circuit, and CJU is a sign determination that determines whether it is training or data based on the output from the tapped delay line. Each circuit is shown below. The signal input via the communication line is first input to the waveform shaping filter ROF, noise components etc. are removed, and only the desired reception signal is received. Signal elements A and B in segment SEG2 shown in Figure 3
When inputted, the signal is passed through the waveform shaping filter ROF, delayed by one signal element at the delay element _TA , and the signal A+B is extracted at the adder circuit +. A at point a
As shown in the figure, the +B vector component is only a negative imaginary component. Next, the A+B signal is a delay element
When delayed by 180 degrees at T _B and complex conjugated, the A+B vector component at point b becomes only a plus imaginary component, as shown in the figure. When the A+B vector component converted to the plus imaginary component and the minus imaginary A+B vector component before conversion are multiplied by the multiplier, the A+B vector component is changed from an imaginary number to a real number, and the output is the training signal. When the real axis component is expressed in terms of polarity, the result of +++++++++++ is obtained. This output result is input to the sign determination circuit CTU from each tap of the tapped delay elements T _C to T _G. In the sign judgment circuit CJU, when the real axis component of the output result from the tapped delay element is expressed by polarity, if it is ++++++++++++, it is judged that the received signal is training, and the real axis component of the output result from the tapped delay element is expressed by polarity. If it is random, +-++-+-++, it is determined to be data. That is, training or data identification can be performed using the code sequence of the received signal. If a failure occurs during data transmission and a momentary interruption occurs, using the training detection circuit according to the present invention can easily detect whether the received signal is data or training signal. For example, since valid data can be continuously received without invalidating data after the momentary interruption has been recovered, resistance to momentary interruption can be improved. (g) Effects of the Invention As explained in detail above, according to the training detection method for data modems of the present invention, by adding a training detection circuit, training signals can be easily detected. This has a great effect in that valid data can be continuously received without invalidating the data after the momentary data interruption has been recovered, and the resistance to momentary interruption can be significantly improved.

[Brief explanation of the drawing]

Figure 1 is an eye pattern diagram of the training signal.
Fig. 2 is a characteristic diagram of a conventional method for improving momentary interruption tolerance when an instantaneous interruption occurs in a receiving carrier, Fig. 3 is a time chart of a training signal, and Fig. 4 is a diagram of a training detection method for a data modem according to the present invention. FIG. 2 is a circuit configuration diagram of an example. In the drawings, A, B, C, D, and Z are signal elements, ROF is a waveform shaping filter, T _A to T _G are delay elements, + is an addition circuit, × is a multiplication circuit, and CJU is a sign determination circuit.

Claims (1)

[Claims] 1. A first delay circuit that delays a received signal by one signal element via a waveform shaping filter, an adder circuit that adds an input signal that does not pass through the first delay circuit, and a delay circuit that delays the output of the adder circuit by 180 degrees. a second delay circuit; the output of the second delay circuit is input to a multiplication circuit; while the output of the second addition circuit is input to the multiplication circuit without passing through the second delay circuit; A training detection method for a data modem, characterized in that the output result of the multiplication circuit is input to a sign determination circuit via a delay element with a tap to detect whether it is a training signal or data.