Apparatus and Method for Digitally Generating a Modified Duobinary Signal

1. Field of the Invention

This invention relates to binary data transmission systems wherein the bit rate of the binary data exceeds that at which intersymbol interference would occur. More particularly, the invention relates to method and apparatus for digitally converting a binary signal into a modified duobinary signal.

2. Description of the Prior Art

Duobinary systems are explained, for example in an article appearing in IEEE Transactions on Communications and Electronics, vol. 82, May 1963, pp. 214 - 218, as well as a variety of generally available publications. In particular, the duobinary system is disclosed and claimed in U.S. Pat. No. 3,238,299 entitled, "High-Speed Data Transmission System" by the present inventor. Subsequent publications by the present inventor have disclosed a number of variations in the basic duobinary concepts. The duobinary technique, as originally developed, contemplated a baseband three-level, correlative level-coded signal in which the amplitude of the wave increased with decreasing frequency, reaching a maximum value at a frequency which was essentially zero. Important among the additional publications of interest is the article entitled, "Correlative Digital Communication Techniques", appearing in IEEE Transactions on Communications Technology, vol. 13, June 1965, pp. 203 - 208. In addition, a number of correlative level-coded techniques are described in U.S. Pat. No. 3,388,330 entitled, "Partial Response Multilevel Data System". More particularly, a technique for converting a binary signal into a modified duobinary signal and the particularities of the characteristics of the modified duobinary signal are described in considerable detail in U.S. Pat. No. 3,457,510 entitled, "Modified Duobinary Data Transmission" by the present inventor. U.S. Pat. No. 3,457,510 is incorporated herein by reference.

It is important to note first that the conversion filter of the subject U.S. Pat. No. 3,457,510 comprises a bandpass filter having zero transmission at zero frequency and at 2f.sub.1 Hz. The maximum transmission occurs at the frequency f.sub.1 Hz. Further, the conversion of the correlative two-level signal is accomplished by filtering or shaping of the two-level signal. In contrast, the correlative level-coded modified duobinary waveform of the present invention is accomplished solely by means of digital techniques.

The invention contemplates an apparatus and method for digitally processing a binary pulse train to obtain a modified duobinary signal. The binary signal is first digitally transformed into an altered binary pulse train having each bit correlated with the second preceding bit. Next, the transformed binary pulse train is digitally converted into a three-level digital waveform having extreme levels representing a first signal level of said binary pulse train and a center level representing the second signal level of said binary pulse train.

FIG. 1 is a generalized block diagram showing the basic elements of the invention.

FIG. 2 is a detailed schematic diagram showing a preferred embodiment of the invention for digital generation of the modified duobinary waveform from a binary input signal.

FIG. 3 is a waveform diagram which shows the processing of the binary signal as it travels through the various steps in the digital generation process to obtain the modified duobinary waveform.

With respect to FIGS. 1 and 2, the letters in the circles designate the waveforms and refer to waveforms having the same respective letter as given in FIG. 3. The generalized block diagram of FIG. 1 illustrates the digital process wherein a binary signal is changed into a modified duobinary signal. A binary input such as illustrated at A, FIG. 3, is applied at input 2 and is transformed in digital transformation processor 4 into an altered signal which is also binary in form but is transformed so that the present bit is correlated with the second bit back from any particular bit in question. The altered binary signal appears on output lead 6 and is shown as waveform C in FIG. 3. The altered complementary signal appears on lead 8 and is shown at D in FIG. 3. The altered and complementary altered signals are applied to the input of digital logic conversion processor 10 where they are converted into digital modified duobinary signals on output lead 16 as is shown in waveform K of FIG. 3. Clock timing is applied by clock 12 via lead 14 to the digital transformation processor 4 and digital logic conversion processor 10. The clock rate is equal to the binary data rate.

A preferred embodiment of the invention is shown in FIG. 2 in which the binary input signal is applied via lead 2 to a J-K masterslave flip-flop combination. The binary input signal from lead 2 is applied simultaneously to both the J and K inputs of flip-flop 20, and a partially alternate binary signal is obtained from the Q-output of the flip-flop on lead 22. The partially altered signal which is shown as waveform B, FIG. 3, is applied simultaneously to the J and K inputs of flip-flop 24. The transformed signal, which is the altered binary form of the input binary signal applied to input lead 2 to the transformation processor 4, is outputted on lead 6 from the Q-output of flip-flop 24. The altered complementary waveform is outputted from Q-output of flip-flop 24 via lead 8. The altered and complementary altered forms of the binary signal are shown as waveforms C and D, respectively, in FIG. 3. These are applied via leads 6 and 8 to the digital logic conversion processor 10.

The altered and complementary altered binary signals are next each delayed by two digits in the master-slave J-K flip-flop, which consists of flip-flops 30 and 40. The altered waveform is applied via lead 26 to the J-input of flip-flop 30, while the altered complementary binary signal is applied via lead 28 to the K-input of flip-flop 30. Referring to FIG. 3, it may be seen that the altered waveform which appears at lead 36 from the Q-output of flip-flop 30 is delayed by one bit interval, and the altered complementary output appearing on lead 38 from the Q-output of flip-flop 30 is delayed by one bit as shown in waveform F, FIG. 3. The Q and Q-outputs are applied, respectively, to the J and K inputs of flip-flop 40, and the delayed altered and complementary delayed altered binary signals, each signal being delayed by two digits, appear at the Q and Q-outputs of flip-flop 40 on leads 42 and 44, respectively.

The altered and complementary delayed altered binary signals are applied to the inputs of NAND-gate 46. The complementary altered and the delayed altered binary signals are applied to the inputs of gate 48. The outputs of gates 46 and 48 are shown as waveforms I and J, respectively, of FIG. 3, and these are combined through the transistor circuits associated with transistors 54 and 60 and via transformer 64 to obtain a digitally generated modified duobinary output signal which appears on leads 16A and 16B. A digitally generated modified duobinary signal is as shown at waveform K in FIG. 3. Referring to the U.S. Pat. No. 3,457,510, incorporated herein by reference, it may be seen that waveform K is the digital equivalent of waveform F, FIG. 1, of the referenced patent. It will be seen that in both cases the extreme levels of the wave represent or correspond to the first level or MARK condition of the input binary waveform A. The intermediate, middle or center level of waveform F of the reference or K of the instant inverter represents the second level or SPACE condition of binary input signal A. Note further that the modified duobinary signal of F or K follows a predetermined set of rules. These rules may be readily realized by grouping all of the successive MARKS in pairs and assigning the pair number to each MARK as illustrated in FIG. 3. Successive MARKS are indicated by numerals 1 or 2, with a repetition of this numbering for the next pair of MARKS. A MARK bearing number 1 in a pair of two successive MARKS will be seen to always have the opposite polarity relative to the previous MARK which, of course, carries the number 2. The polarity of the MARK identified by number 2 relative to the previous MARK bearing number 1 is governed by a set of odd and even rules as in a straight duobinary system and method. More specifically, if the number of intervening spaces between a pair of MARKS numbered 1 and 2 is even, then the polarities of these MARKS are the same. If the number of intervening spaces between a pair of MARKS numbered 1 and 2 is odd, then the polarities of these two MARKS are opposite. A correlation of properties of the waveform K produced in accordance with the present invention permits the ready detection of errors in received and transmitted data.

An understanding of the apparatus and method of digitally generating a modified duobinary signal may also be obtained by relating the circuit structure to a mathematical analysis of the problem. Note that in the following discussion the waveform symbols A through K are used to designate the character of the input of various points within the circuit as shown in FIG. 3. This was done to simplify the mathematical presentation. Let us consider the output stage consisting of two NAND gates 46 and 48; resistors 50, 51, 56, and 57; capacitors 52 and 58; diodes 59 and 61; transistors 54 and 60; and transformer 64. Input G is delayed by two unit intervals from input C. Hence

where .DELTA. indicates one unit delay T seconds long and 1/T is the speed in bits per second. Then .DELTA..sup.2 indicates two units delay. Input D is merely the inverse of C and may be denoted as

input H is then .DELTA..sup.2 D, or in terms of C, from (2)

consequently, four possible conditions may exist at the inputs to the NAND-gates and they are expressed in terms of C, i.e., C, C, .DELTA..sup.2 C and .DELTA..sup.2 C. The process of conversion is in general (C - .DELTA..sup.2 C) where the minus indicates conventional algebraic subtraction. The four possible results of such a subtraction are shown in Table I, remembering that C and .DELTA..sup.2 C can assume only binary values, namely "1" or "0".

Let us assume that C and .DELTA..sup.2 C (or G) are both "0". When C = G = 0, then H = D = 1. A NAND-gate with inputs C = 0 and H = 1 would normally have a positive pulse output. However, the diode 59 in parallel with resistor 51 prevents this pulse from occurring and clamps it to value +V. Hence, transistor 54 stays in the cutoff condition, and there is no output pulse at K. Similarly for the NAND-gate 48. It follows that for C = .DELTA..sup.2 C = 0 in Table I there is no output (indicated by "0"), and the center level of the modified duobinary signal appears at K, corresponding to binary "0" at the input A. When C = G = 1 in Table I, H = D = 0. Again, both NAND-gates would normally have a positive output pulse, but this does not occur because of the action of diodes 59 and 61. The diodes clamp their respective transistors to the cutoff voltage +V. Therefore, both transistors remain cut off. It follows that there is no output pulse at point K, and again a center level of the modified duobinary signal is indicated. When C = 1 and G = 0, D = 0 and H = 1, gate 46 has a negative pulse output, passed by capacitor 52. For the duration of this negative output, transistor 54 is saturated and there is pulse output at K, say positive, corresponding to the top level of the modified duobinary signal and binary "1" at A. Gate 48 would normally have a positive output, but this is clamped by diode 61 so that gate 48 has no output. Similarly for C = 0 and G = 1, D = 1 and H = 0. Here only gate 48 has negative pulse output which saturates transistor 60, resulting in a negative pulse at K. This corresponds to the bottom level of the modified duobinary signal and binary "1" at A. All the above conditions are consistent with the entries in Table I.

One of the other major advantages of a level-coded process and system is that redundant digits are not required. It will be realized that the insertion of a digital digit in any binary pulse train reduces the amount of data that can be transmitted over any particular system.