JPH01140813A - Duo-binary encoder - Google Patents

Abstract

PURPOSE:To control the extent of a power spectrum by reading data which synchronizes and corresponds with a sample clock and integrating a binary/ ternary conversion processing and a band restriction processing. CONSTITUTION:A precoder 17 adds binary digital data and data delayed by 1/fb. A digital filter 18 reads sample data from a ROM18c based on one bit serial data outputted from the precoder 17 and the sampling clock SPc and latches in a latch circuit 18d. A transmission characteristic is cos(pif/fb). Here, fb is a transmission bit rate. When a role of factor (k) is set to '0', the characteristic of the digital filter 18 becomes cos(pif/2fb) in 0<=fb/2, and '0' in f>=fb/2, whereby the extent of the power spectrum can be controlled.

Description

[Detailed Description of the Invention] [Object of the Invention] (Industrial Application Field) The present invention relates to a duobinary encoder that performs band limitation by converting binary digital data into ternary data, and particularly relates to a duobinary encoder that performs band limitation by converting binary digital data into ternary data. This invention relates to a method in which value/three-value conversion processing and band limiting processing are performed at once using a digital filter.

(Prior Art) As is well known, in a digital data transmission system, it is necessary to limit the transmission band in order to improve transmission efficiency and prevent unnecessary interference from occurring. For example, when transmitting 1-bit serial baseband data,
A technique that equalizes the transmitted power spectrum to a cosine roll-off characteristic is commonly used.

In this case, the amplitude characteristic of the equal-other filter corresponds to the square root of the cosine roll-off characteristic, assuming that bit 1 of the input data is an impulse. Therefore, the equal-other filter is point symmetric with respect to the point (fb/2.1/2) within the frequency f range where 0≦f≦fb, where fb is the bit rate and meets Nyquist's condition regarding code amount interference. By passing the signal through a filter having the same characteristics again on the receiving side, it becomes possible to satisfy the Nyquist condition and obtain a good eye pattern without code amount interference.

In other words, the eye pattern of the transmitted waveform, which has a power spectrum with cosine roll-off characteristics, is not converged, and the amplitude is much larger than the amplitude of the convergence point (discrimination point) of the eye pattern obtained by passing it through the same filter again. It becomes. For example, when the roll-off factor is "1", the maximum amplitude of the transmitted waveform is (8/π-1) times the discrimination point amplitude after reception convergence, or approximately 1.55 times (approximately 3.8 dB). There is.

In addition, when the roll-off factor is close to 0, the characteristics of the transmitting and other filters alone approach the ideal square wave low-pass characteristics, so the eye pattern of the transmitting waveform converges at the discrimination point. However, the amplitude at areas other than the discrimination point becomes extremely large. For example, when the roll-off factor is less than or equal to rO, IJ, it becomes r2.42 J times or more, and when the roll-off factor is "0", it becomes r6.67 J times.

In this way, the eye pattern of the transmitted waveform equalized to the power spectrum with cosine roll-off characteristics is non-convergent at the discrimination point and the output amplitude increases, so it is necessary to reduce the roll-off factor and tighten the band limit. The more the need arises, the greater the increase in amplitude becomes. Therefore, the equalization of the rolloff factor to a small cosine rolloff spectrum is
It lacks practicality, and as a practical roll-off factor, r
The lower limit is approximately O.3J (amplitude is 1.72 times).

The cause of the increase in amplitude due to equalization to a cosine roll-off spectrum is the inverse Fourier transform of the amplitude characteristic, that is, the long-term decay motion of the impulse response. For this reason, equalization to, for example, 0% roll-off characteristics is impractical, and implementation using a digital filter is also impossible because its coefficient tap length becomes long.

On the other hand, methods have been considered to suppress the transmission energy and make it less susceptible to the influence of amplitude increase due to band limitation by performing ternary transform encoding on binary baseband data. For example, partial response (1, O, - 1) shape and (1,
1) The shape etc. are known. In particular, the (1,1) type is known as a duobinary code, and there is almost no increase in amplitude when limiting the band, and the eye opening ratio in the amplitude direction and the time direction hardly deteriorates.

FIG. 5 shows such a conventional duobinary encoder. That is, the input terminal 11 has 2
Digital data of the value is supplied serially, 1 bit at bit rate fb. This digital data is sent to the adder 12
Precoder 12 consisting of a and 1/fb delay circuits 12b
Then, 1/fb delayed data and MOD (modulo) 2
After being added, the adder 13a and the 1/fb delay circuit 13b
and 1/2 limiter 13c (1.1) type binary/
The data is supplied to a ternary conversion circuit 13 and converted into ternary data.

Then, the 3 value outputted from this binary/ternary value conversion circuit 13 is
As shown in FIG. 6, the value data string has a bit rate fb
A low-pass filter (
The signal is band-limited by LPF (hereinafter referred to as LPF) 14 and guided to output terminal 15 .

The duobinary encoder shown in FIG.
When the passband of the ternary data string output from the value conversion circuit I3 is completely limited to 1/2 or less of the bit rate fb by the L P F 14 having ideal square wave characteristics, the seventh As shown in the figure, the eye pattern shows very little deterioration.

Figure 8 shows the characteristic A of the transmitted power spectrum in this case and the characteristic B of the transmitted power spectrum when normal binary data is equalized to 100% cosine roll-off characteristic.
This shows that. In other words, the total energy of the output signal after duobinary conversion and ideal low-frequency limiting is:
Completely 1/2 of the total energy of the binary transmission signal equalized to a 100% cosine roll-off power spectrum
It has become.

Therefore, even when a high frequency signal is modulated and transmitted using baseband data, there is no unnecessary spread of the transmission spectrum, and interference with adjacent transmission channels can be prevented.

In this way, the duobinary code uses L P F 14 with ideal square wave characteristics to adjust the existence band of the transmitted power spectrum and the energy of the transmitted signal without deteriorating the eye pattern. It can be reduced to 1/2 of that of a normal binary system.

However, L P with ideal square wave characteristics
Achieving F14 has the problem that it is nearly impossible to sufficiently suppress the broadening of the transmitted power spectrum.

(Problems to be Solved by the Invention) As described above, in the conventional duobinary encoding digital data transmission method, the transmission spectrum is reduced to 1/2 of the transmission bit rate fb without deteriorating the eye pattern. Although it has the potential to limit bandwidth,
Since an LPF with ideal square wave characteristics cannot be realized, there is a problem in that the transmission spectrum cannot be completely limited.

Therefore, the present invention was made in consideration of the above circumstances, and an object thereof is to provide an extremely good duobinary encoder that can completely limit the transmission spectrum and sufficiently suppress the spread of the output power spectrum. shall be.

[Structure of the Invention] (Means for Solving the Problems) That is, the duobinary encoder according to the present invention performs sampling from a memory in which sample data of a final output waveform is stored, based on binary digital data. It reads the corresponding data in synchronization with the clock.
It is equipped with a digital filter having transmission characteristics that combine binary/ternary conversion processing and band limiting processing.

(Function) According to the above configuration, since a digital filter having a transmission characteristic that combines binary/three-value conversion processing and band limiting processing is provided, an LPF having ideal square wave characteristics can be used. This makes it possible to limit the transmission spectrum and sufficiently suppress the spread of the transmitted power spectrum without using it.

(Example) Hereinafter, an example of the present invention will be described in detail with reference to the drawings. In FIG. 1, binary digital data is supplied to the input terminal 1B in a 1-bit serial manner at a bit rate fb. This digital data is sent to the adder 17
Precoder 17 consisting of a and 1/fb delay circuits 17b
After being added to the 1/fb delayed data by MOD2, the data is supplied to the digital filter 18.

This digital filter 18 uses a sampling clock S
Divide Pc into 1/2 to generate data clock 1/
In synchronization with the data clock output from the 2 frequency divider 18a and the 1/2 frequency divider 18a, the 1-bit serial data string output from the precoder 17 is latched sequentially, and the n-bit serial data string is outputted from the precoder 17. A register 18b that outputs to /kuralel,
The outputs of this register 18b are addresses A n-1 to A other than the highest address An.

is determined, and the above sampling clock S
The highest address An is determined by Pc, and the sample data read from the read-only memory (hereinafter referred to as ROM) 18c, which stores the sample data of the final transmission waveform, is read from the ROM 18c.
It is composed of a latch circuit 18d that latches in synchronization with Pc.

The digital filter 18 includes a precoder 17
The sample data is read from the ROM 18c based on the 1-bit serial data and sampling clock SPc output from the ROM 18c, and the cos as shown in Fig. 2 is a system that integrates binary/ternary conversion processing and band limit processing. It has a transmission characteristic of (πf/fb).

In this way, the ternary data output from the digital filter 18 is converted into analog data by the D/A (digital/analog) conversion circuit 19, and then 0≦f≦
The signal is band-limited by an LPF 20 having a pass frequency band of fb/2, and then guided to an output terminal 21.

Here, the (1,1) type binary/ternary conversion circuit in duobinary encoding has an amplitude transmission frequency characteristic of cos(πf/fb), where fb is the transmission bit rate. In addition, the characteristics of the LPF for limiting the transmission band are "1" when 0≦f≦(1-k)fb/2, and when (1-k)fb/2<f< (1+k)fb/2, ( 121+sin (π2k) 1-2f fb
), and when f≧(1+k)fb/2, it becomes “0”.

Here, k is a roll-off factor, and the upper limit of the transmission spectrum is limited to (fb/2) to fb corresponding to k-0 to k-1. In this case, as described above, if the roll-off factor k is small, it will be difficult to realize it even if a digital filter is used, so the value that can be put to practical use is between k-0,3 and 1.

In this embodiment, since the digital filter 18 having characteristics that have both the characteristics of the binary/ternary conversion circuit and the characteristics of the LPF for limiting the transmission band is used, the minimum band limit of k-0 (0% roll-off equalization) is possible.

Therefore, the characteristic of the digital filter 18 that combines binary/ternary conversion processing and band limiting processing is cos(πf/2fb) when 0≦r<fb/2, and "0" when f≧fb/2. Then, the impulse response 2 cos(
πfb t) /πfl -(2fb t) 2) The convergence of 2) is fast as shown in FIG. 3, and for example, a highly accurate digital filter 18 can be realized with filter coefficients within ±12 time slots of the bi-sotorate. I can do it.

In this regard, the conventional duobinary system shown in FIG.
In order to obtain the same characteristics as in the above embodiment in the encoder, it is necessary to manufacture a filter (ideal square wave LPF) having the characteristics of the k-0 transmission band limiting LPF described above. Since the impulse response 5 in (πfb t) /πfb t lasts for a long time as shown in FIG. 4, it is impossible to realize it with a filter coefficient of a finite time throttle length.

Therefore, according to the configuration of the above embodiment, the duobinary-encoded transmission spectrum is reduced to 1/1 of the bit rate.
can be completely limited to two bands. Therefore, even when a high frequency signal is modulated and transmitted using baseband data, there is no unnecessary spread of the transmission spectrum, and interference with adjacent transmission channels can be prevented.

It should be noted that the present invention is not limited to the above-mentioned embodiments, and can be implemented with various modifications without departing from the gist thereof.

[Effects of the Invention] Therefore, as detailed above, the present invention provides an extremely good duobinary encoder that can completely limit the transmission spectrum and sufficiently suppress the spread of the output power spectrum. be able to.

[Brief explanation of the drawing]

FIG. 1 is a block configuration diagram showing an embodiment of a duobinary encoder according to the present invention, FIG. 2 is a characteristic curve diagram showing the characteristics of a digital filter of the same embodiment, and FIG. 3 is a digital filter of the same embodiment. FIG. 4 is a characteristic curve diagram showing the impulse response of the LPF necessary for a conventional duobinary encoder to have characteristics similar to those of the same embodiment. FIG. 5 is a characteristic curve diagram showing the impulse response of the conventional duobinary encoder. A block configuration diagram showing a duobinary encoder, the mb diagram is an LP diagram used in the conventional duobinary encoder.
Figure 7 is a diagram showing the eye pattern when the characteristics of the same LPF are ideal, Figure 8 is a diagram showing the characteristics of the same LPF.
The characteristics of the output power spectrum by duobinary encoding when the characteristics of F are ideal and the normal binary data are 1
FIG. 3 is a characteristic curve diagram showing the characteristics of a transmission power spectrum when equalized to a 00% cosine roll-off characteristic. l]...Input terminal, 12...Precoder, 13...
・Binary/three-value conversion circuit, 14...LPFS15...
Output terminal, 16... Input terminal, 17... Precoder, 18... Digital filter, I9... D/A conversion circuit, 20... LPF, 21... Output terminal. Applicant's agent Patent attorney Takehiko Suzue Figure 3 Figure 7 Figure g

Claims (1)

[Claims] In a duobinary encoder that performs band limitation by converting binary digital data into ternary data, the above two
Based on the digital data of the value, the corresponding data is read out in synchronization with the sampling clock from the memory where the sample data of the final transmission waveform is stored.
A duobinary encoder comprising a digital filter having a transmission characteristic that combines three-value conversion processing and the band-limiting processing.