Method and Apparatus for in-Band Signalling in Data Transmission

Data communication systems make use of various channels such as telephone, radio or optical links to transmit digital data from one point to another. In order to establish route and conduct certain testing and problem diagnosis, certain control information has to be communicated between various terminals in the system. Terminal equipment at the receiving end of the channel must be prepared and notified to accept the data by means of control words, and the transmitting end of the link must be notified when the receiving station is ready to receive data and to signal it.

In the prior art, such control signals were transmitted over a separate link from that which was used to transmit the data, or was sent at a specified time known to both the transmitter and receiver so as to distinguish it from data. Such communication has often included voice transmission wherein the operators at both ends of the link would confirm that the link had been established and that the equipment was in readiness for the transmission. Such techniques require additional bandwidth in the form of a separate channel or time allocated within the data channel to distinguish control information from data. In synchronous links, differences in communications time caused by different signal paths between the two links require that sufficient time be allocated before data is transmitted to insure that the receiver has properly received the control signals. In either case, it is required to establish or maintain a separate control channel in order to regulate the transmission of data. Additional bandwidth in the form of a separate channel or in allocated times on the primary channel are referred to as overhead connected with the transmission of the data.

This invention generally relates to digital data communications systems. More particularly, the present invention relates to communications systems wherein control signals are required as well as data between the transmit and receive sites. The present invention describes a means of communicating both the control signals and the data over the same channel.

Briefly described, the present invention involves a communication system in which multibit words representing either data or control signals are transmitted. Identification for the control words is employed so that no restriction is required on the bit sequences that may be transmitted as data including data signals identical with control words.

More particularly, the in-band signalling system, in accordance with the invention, comprises a means of encoding the control bits in a different format from that used for the data bits. In the system, code translators are employed to recode the data in such a way that it cannot be interpreted as control signals. A monitor at the transmitter is employed to examine the encoded data prior to its transmission to delete from the transmitted sequence any bit pattern that could be erroneously interpreted as control signals. In such cases where deletion is required, the original data stream is retransmitted in the normal manner since the deletion process is interpreted as a data error. In this manner, any data bit pattern retransmitted will be encoded in such a way that it will have low probability of being encoded again with the same bit pattern. The method may be employed in either simplex or duplex transmission. In the former case, the receiver notifies the transmitter that the intended message was not received and causes the message to be retransmitted. In one way transmission systems, error detection at both the transmitter and receiver cause both ends of the link to recognize that retransmission of the data will take place.

The above features of this invention may be more clearly understood by the following detailed description and by reference to the drawings in which:

FIG. 1 is a block diagram of a system in accordance with this invention;

FIG. 2 is a block diagram of a suitable code translator or scrambler for use in connection with this invention; and

FIG. 3 is a block diagram of a suitable control word detector for use in accordance with this invention;

FIG. 4 is a pictorial display of a typical control word sequence used in connection with this invention; and

FIG. 5 is a block diagram of a suitable data decoder or descrambler used in accordance with this invention.

Now referring to FIG. 1, showing a transmitter section and a receiver section incorporating this invention, the transmitter section includes an input lead 10 from a data source plus a plurality of input lines 11 for control information. Data on line 10 is introduced into a code translator 12 which applies a linear algorithm to all data. The code translator is, for example, a multi-stage, e.g. 3-11 repeating bit sequence generator of the type disclosed in FIG. 2, or alternately an 8202 encoder, as disclosed in Signetics Applications Manual, Sections 3-16, copyright 1973, Signetics Corporation of Sunnyvale, California, 94086, U.S.A. Suffice it to say the code translator 12 modifies the incoming data in a predictable manner. The control lines carrying such signals as Request to Send, Local Loop Test, Remote Loop Test, or any other control signals required for normal operation of a data transmission system for establishing and maintaining communications. These signals are converted into bit sequence formats in a control word formatter 13 which are simply an encoder. Any control signal on its leads 11 is transformed into its distinct bit code and outputed on lead 15 as the B input to a switch 16 or other combining device.

The output of the translator 12 is introduced into a delay circuit 20 and simultaneously into a detector 21 which is programmed to detect any bit combinations constituting control words which might, by coincidence, appear in the modified data train of bits. The detector 21 corresponds to the decoder portion of FIG. 3 which looks for control format bit codes which the data may inadvertently resemble.

The illegal code detector 21 produces an output on lead 22 whenever a control code combination in the transformed data is detected and no output on the absence of such illegal code. Lead 22 is the enabling input to a code modifier or error injector 23 which is connected between the delay circuit 20 and the A input of switch 16. The simplest form of modifier 23 is an inverter which inverts one bit in the sequence. The change of a single bit, as by inversion, often is sufficient at the receiver to be detected as an error. This is true particularly where a parity check is applied to the data as received.

The output of the singly or doubly modified data on lead A is combined selectively with actual control words in switch 16 and introduced into the transmission channel which may be wire, optical or radio as the case may be.

The switch 16, in the preferred embodiment, is normal in the B position as shown whereby it is conditioned to send control words over the transmission channel in order to perform all of the necessary steps for establishing a connection with the receiving station, is in condition to receive data. These functions are often termed the "handshaking" operations of the data communications system. Switch 16 is operated by any suitable means including a start code contained in the data or arriving over control lines, or for that manner, manually.

At the receiving station the initial element is a buffer storage or delay circuit 30 having sufficient capacity to store a full control word including any identifying preamble or ending. In one embodiment, a control is made up of a 31 bit code followed by an 8 bit control code followed by the compliment of the 8 bit control code. In such a case, the delay circuit 30 has two sections, a 31 bit section and an 8 bit section with parallel output leads 31 to a latch circuit 32. Connected in parallel at the input to the receiving station with the delay circuit 30 is a control format decoder 33 which examines all data and control words coming in on the line and when ever a control sequence appears in the train of bits, enables the latch 32 over lead 34. The latch circuit 32 then passes the control code to the controlled circuits of the receiving station via its output leads 35. One of the output leads of the receiving station latch circuit 32, namely lead 35A, carries an output "Carrier On" which indicates that the system next will transmit data. The output of lead 35A is amplified and applied as the enabling input to an AND gate 36 in the received data channel.

The serial output of the delay circuit 30 is lead 40 which carries all received data and control words to a descrambler 41 which performs the inverse transformation on the incoming train of bits as occurred in the code translator 12 of the transmitter. The same decoders as disclosed above may be used for this purpose. The output of the descrambler 41 which is descrambled data and control words, is introduced as the second input to the AND gate 36. The gate 36, of course, is disabled where control words are present, so only passes data which, as it arrives at the transmitting station, or data which, when modified in the code translator 12, coincidentally appears as a control word but which has been modified or corrupted by the modifier 13. The latter, when passed on the utilization device, is easily detected as an error in accordance with well known error detecting techniques presently used almost universally in computers.

When this invention is used, both the data and control words are transmitted on the same channel and there are no restrictions on the format of the data. With the translation of all data in a cyclic fashion, the probability that the data matches a control word is at least as small as the probability of the code words appearing in the raw data. Where the cyclical code is selected to minimize code word generation, this probability is reduced even further.

As an example, a typical code sequence is an s bit preamble followed by t control bits followed by t inverted control bits. In typical applications, s equals a 31 bit sequence and t is 8. This is illustrated in FIG. 4 where the preamble comprises a pair of (1)s followed by alternating (0)s and (1)s for a total of 31 bits. The preamble is followed by an 8 bit control word and then by the 8 bit control word inverted. Other sequences may be used but we have found that an odd numbered bit preamble of sufficient length, e.g. in excess of 25 bits, is recommended. Using a 31 bit preamble the probability of the code appearing in random data is 2.sup.-.sup.31 or approximately one half chance per billion. When scrambled in the code translator which employs a seven stage shift register in which data is EXCLUSIVELY ORed with itself in a predetermined, fixed relationship, there, the probability of a code preamble appearing in the translated data is 2.sup.-.sup.31.

Even with the small but finite probability of scrambled data producing a code preamble, this data is not introduced into the system since it is further modified or corrupted in the control word preamble detector of FIG. 3. This data simulating control words will not pass through the system which can expect an error in the data to appear in 2.sup.31 bits or an error rate of 10.sup.-.sup.9 -- a rate far lower than normal data channel error rates of 10.sup.-.sup.5 to 10.sup.-.sup.8.

Where the system employing this invention requires or allows retransmission of detected errors, the probability of a second or subsequent repeat error existing is 2.sup.-.sup.7. This is due to the fact that the encoding technique employs an odd number of bits in the psuedo random sequence, e.g. 2.sup.7 - 1 or 127, and repeating or cyclical coding of the same data will not occur and data or retransmission will precess with respect to the bit sequence in the scrambler or translator. Where the cyclical code is selected to minimize code word generation, this probability is reduced even further.

Also, employing the above techniques, control word detection is enhanced by the procedure of sending a control word followed by its complement. This is illustrated in FIG. 4. No controls at the receiving station are effected unless a match exists between the control word and its following complement. This is accomplished in the Control Format Decoder 33 which not only examines for preamble but also compares the next 16 bits for complements of the first 8 bits in the second 8 bits. This provides error checking. Normally, control signals are expected several times so that bit errors in the control words can be ignored and the next correct transmission of control words can be detected and used.

A simplified version of a code translator or scrambler 12 is shown in FIG. 2. It comprises a plurality of shift registers 50a to 50n where n, for example, is 7. Data on lead 10 is introduced into EXCLUSIVE OR gate 51 along with the EXCLUSIVE ORed output of the a(1) and n(7) stages of the shift register 50 as combined in second EXCLUSIVE OR gate 52. The logical equation of the translator 12 is:

where

A is the incoming data;

A.sub.s is scrambled data;

A.sub.s.sub.-a is data in stage a of the shift register 50;

A.sub.s.sub.-n is data in stage n of the shift register 50; and

.sym. denotes the EXCLUSIVE OR function.

At the receiving station, data is easily recoverable employing the inverse of the circuit shown in FIG. 5. It employs a multistage shift register 60 to which scrambled data As is introduced. The output of its n(7) stage is EXCLUSIVELY ORed with the output of its a(1) stage in EXCLUSIVE OR gate 61. The output of gate 61 is EXCLUSIVELY ORed with the scrambled data As in second EXCLUSIVE OR gate 62. The output is thus unscrambled data A.

The descramble equation is simply:

As described above in connection with FIG. 1, scrambled data must be continuously monitored for the presence of control word preambles in illegal code detector 21. A suitable detector is shown in FIG. 3 for detecting control code preambles as illustrated in FIG. 4. The detector is configured to search for the sequence of a pair of (1)s followed by alternate (0)s and (1)s. It employs a 31 state ring counter 70, either clocked by each data bit or at the data rate. Scrambled data A.sub.s is introduced over lead L as one input to an OR gate 71. Clocking input C is also introduced into a one stage storage register 72.

Data in register 72 is applied to EXCLUSIVE OR gate 73 plus AND gate 75. Gate 75 detects the appearance of two (1)s appearing in a sequence and resets counter 70 via an OR gate 74. Thereafter, alternate (0)s and (1)s appearing in the scrambled data are monitored by the EXCLUSIVE OR gate 73. Deviation from the alternate pattern will reset counter 70 via an inverter 76 and OR gate 74. If it runs through the full series of alternate (0)s and (1)s after the initial pair of (1)s, it signals a full preamble on its lead END, which it introduces into the preamble at the last bit of the code words appearing in the raw data.

Where the system employing this invention requires or allows retransmission of detected errors, the probability of a second or subsequent repeat error existing is 2.sup..sup.-7. This is due to the fact that the encoding or scrambling in OR gate 71 thereby changing the last (0) to a (1) and corrupting the data to 1101010 - - 11. Any time 2 successive (1)s appear in the data, the counter is reset and the count begins again. As recognized, if a different preamble is desired, the configuration of the detector is changed accordingly.

It is to be recognized that the implementation in general may be changed in accordance with the data and control word format carrying out the method of this invention without departing from the principles disclosed in this specification. The preferred embodiment is shown but one might change the specific embodiment without departing from the spirit and scope of this invention as set forth in the claims below.

The above described embodiments of this invention are merely descriptive of its principals and are not to be considered limiting. The scope of this invention instead shall be determined from the scope of the following claims, including their equivalents.