NO843923L - PROCEDURE AND DEVICE FOR AA FOEYE IN OR DIGITAL BINED SCRAP SIGNAL - Google Patents

Description

From the journal "TELECOM REPORT", 2 (1979), special edition digital-transmission technique, pages 46 to 51, there is known a method for adding a digital narrowband binary signal to or separating it from a time-division multiplex signal. In a device DSE 64K/2 for inserting a digital signal, 64 kbit/s digital signals are added to or separated from a 2048 kbit/s digital signal. At the insertion point, the 2048 kbit/s digital signal's line code is decoded, the 64 kbit/s digital signals by the filler bit method are inserted into predetermined places in the 2048 kbit/s digital signal and the line code is again newly formed. After further transmission, the 64 kbit/s digital signals can again be separated after decoding the line code.

This method has the disadvantage that the line code must be decoded at the point of insertion and coded again, which would be rather expensive for digital signals with a higher bitrate than 34 Mbit/s. The same applies to the filling operation.

A similar method for inserting sub-signals is used with the PCM tone channel system MStD, which is described in the same journal on pages 52-58. With this system, the sub-signal is not inserted by the filling method, but as a rule synchronously, which costs less, but presupposes a synchronism between the sub-signal and the digital signal's clock frequencies.

Yet another method for inserting a digital part signal into a digital signal is used when transmitting an E&M signaling information by time multiplex device PCM30F, which is described on pages 35 - 41 of the aforementioned journal. The digital signaling information is sampled at a frequency of 500 or 1000 Hz which is considerably higher than the double bit follower frequency of the digital signaling signal and does not need to be synchronous with this. Each sample is transmitted with one bit. On the receiving side, the signaling signal is recovered in a very simple way.

This method does indeed require very little equipment, but has the disadvantage that the bitrate for the transmission of the sub-signal is considerably higher than the bit-rate of the sub-signal. This disadvantage is only insignificant when the bitrates are available

for the transmission of the partial signal in the digital signal, must not be poorly measured. Digital broadband signals are used for the transmission

coded with a previously unmentioned wiring code so that direct current components are avoided and the transmitted signal's power spectrum is affected in a suitable way. A code that has proven beneficial for digital signal transmission, particularly over glass fibers, is the 5B/6B code, which is described in the journal "TELECOM REPORT", 6 (April 1983), special edition Information Transmission by Light, p.

133 - 137, table 2 and reproduced on fig. 1.

With this line code, an uncoded digital signal is divided into equal, successive blocks, each with a length of 5 bits. These blocks make up the 5B code words. In accordance with the code table in fig. 1, the 5B code words are converted into 6 bit long 6B code words, which is associated with an increase in the bit clock frequency by 20%. Whether a 6B code word in the code table is taken from the number column for the positive group (group +) or the column for the negative group (group -) depends on which sequence group was entered with the previous 6B code word. The shift between the two groups takes place for the purpose of equalizing as far as possible the summed number of zeros and ones in the sequence of 6B code words to achieve freedom for direct current components. Among the 32 5B code words there are 18 to which the same 6B code word is assigned for both groups. For these 6B code words, the sequence group has the same sign as the group itself. These 6B code words, which are the same in both groups, can be replaced by arbitrary other 6B code words without violating the rule according to which the sequence group's sign was searched.

The task of the invention is to provide instructions for a method which enables later insertion and separate separation of a single digital sub-signal with a relatively small bitrate in a 5B/6B coded digital signal with little equipment cost.

According to the invention, this task is solved by

that in the time multiplex signal a broadband signal with synchronous words is transmitted in a binary 5B/6B line code which for some 5B codewords contains the same 6B codeword (selection codeword)

in the positive and the negative group, as well as extra selection code words as filler words,

that for the narrowband signal, use is also made of the 5B/6B line code and exclusively selection codewords,

that the narrowband signal at the insertion point is synchronously replaced by at least one filler word,

and that the narrowband signal at the separation point is ascertained on the basis of the sync word and switched off.

In this connection, particular thought has been given to the insertion and separation of a digital ISDN signal (Integrated Services Digital Network) in resp. from a digital image signal. One imagines here that the picture signal in the terminal (picture phone) is brought in 5B/6B line-coded form and in a further device at the participant location (participant connection device) is supplemented with the ISDN signal, which can contain telephone and data signals.

Furthermore, it is assumed that the 5B/6B coded video signal is transmitted in a video switching field in network nodes (subscriber switching center), but that the digital ISDN signal is separated in advance from the digital video signal, because the ISDN signal does not necessarily accompany the picture signal.

For the ISDN signal, a bitrate of 160 kbit/s is assumed here, although the invention also applies to other bitrates, for example 80 kbit/s. For the image signal, a bitrate of approximately 34 Mbit/s is further assumed, although the invention is also applicable to image signals with e.g. 140 Mbit/s.

A first variant of the method according to the invention

is characterized by

that during the formation of the time multiplex signal is output

from a 34.368 Mbit/s digital signal with CCITT pulse frames as basic frames,

that the time multiplex signal with the broadband signal is subjected to 5B/6B line coding,

that five successive base frames form an upper frame,

that the frame marker word for the first basic frame in the upper frame is located as the upper frame marker word in two 6B code words,

and that a fixed number of 6B code words serve as filler words for each superframe. For this purpose, a line coded narrowband signal with a bit frequency of 5/6 is selected. 716 kbit/s, and on the upper frame the marking word therefore directly follows 32 filler words.

Fig. 2 is a non-faithful reproduction of the structure of the pulse frame and shows in the upper half the digital signal as a sequence of 5B code words, while the lower half reproduces the digital signal as a sequence of 6B code words. The drawing in the 5B plane is chosen here for the sake of clarity, because the insertion and separation operation actually belongs to the 6B plane. Overframe marking order URKW is in itself a normal frame marking order RKW

which only differs from the others in that its 10 bits in the 5B plane lie exactly in two consecutive 5B code words. In the 6B plane, this overframe marking word URKW therefore appears in two successive 6B code words. Of these 12 bits in total, 10 bits are used directly for the determination of the upper frame marking word; in addition, the second and fifth bits not directly used must be the same. Every fifth frame marker word RKW is an overframe marker word URKW. The other frame marker words RKW contribute to more than two 6B code words and do not form an overframe marker word.

Another variant of the method according to the invention

is characterized by

that a 34.368-Mbit/s digital signal with a CCITT pulse frame as the basic frame is used to create the time multiplex signal,

that the baud rate of the time multiplex signal with the broadband signal is increased to 41.242 Mbd with 5B/6B coding,

that five successive basic frames form an upper frame, that the first basic frame's frame marker word is located as upper frame marker word in two 6B code words,

and that during a further increase of the baud rate, a fixed number of filler words are inserted between the 6B code words of the 41.242 MBd digital signal.

In this connection, it is advantageous if a line coded narrowband signal with bit sequence 6/5 is selected. 2864 kbit/s and the line coded narrowband signal replaces 128 filler words per upper frame distributed at even distances over the upper frame, so a total baud rate of 44.6784 MBd appears.

Fig. 3 is a non-faithful rendering of the structure of the pulse frame with this variant set up in the same way as in fig.

2. The baud rate is increased by that for the added filler words. This way of proceeding would make no sense if the supplemented digital signal was transmitted over CCITT lines, as already without this increase it exactly corresponds to the CCITT bit rate. However, since this method is mainly taken into account for the participant network, it may make sense here to use a higher bitrate than that recommended by the CCITT.

A favorable further design of the method according to the invention is characterized by the fact that the narrowband signal is sampled at a frequency that is more than twice its bitrate, that each sampled random sample is coded with one bit, that four and four bits forming a 4B code word are converted to a selection code word, and that the selection code words are synchronously exchanged for the filler words in the time multiplex signal.

In a first embodiment, which corresponds to fig. 2, it is advantageous if the sampling frequency for the narrowband signal amounts to 572.8 kHz and the bits in four consecutive random samples which form a 4B codeword are converted into one and one sampling codeword with a baud rate of 6/5. 716 kBd.

In another embodiment, which corresponds to fig. 3, it is advantageous if four mutually synchronous, plesiochronous or asynchronous narrowband signals are sampled simultaneously, each at 572.8 kHz, and four and four simultaneously sampled random samples forming a 4B codeword are converted into one and one selection codeword with a baud rate of 5/6 . 2864 kBd.

A conversion of 4B code words to 6B code words takes place in accordance with fig. 4, where the associated 5B code words are also indicated for the sake of clarity. The useful information is denoted by NI and the filling information by Fl. The two 6B code words FW1 and FW2 which are used as filler words have the meaning of a digital blank character as long as no 5B/6B line coded narrowband signal is inserted into the time multiplex signal. These filler words can also be added again when the narrowband signal is separated again.

The recoding of 4B code words to 6B code words is possible with a slow link that can be easily integrated. This saves the recoding of the broadband signal with a view to insertion or separation, a recoding which is costly due to the necessary quick-acting coupling technique. This means an economic advantage of the invention.

Fig. 5 gives an overview of the two mentioned variants of the invention. The letters A to I have the following meaning:

A ) means that the 4B/5B channel code and the 5B/6B line code

can be summarized into a 4B/6B code, i.e. that an ISDN entry at telegraphy speed is possible.

In the first line of fig. 5 it is seen with respect to a broadband signal DBS of approx. 34 Mbit/s that the increase of the baud rate by the factor 13/12 means that the difference between the bitrate from the line of 44.6784 MBd and the bitrate of the image signal of 34.368 Mbit/s corresponds to the factor 13/10, which can be realized as easily as the factor 6 /5 which is used in the 5B/6B line coding. A clock frequency conversion with one of these factors becomes necessary when the line signal is to be converted into a CCITT interface signal with a bitrate of 34.368 Mbit/s for the remote transmission.

In addition to an increase factor of 13/12 for the baud rate, other factors such as e.g. 55/54, 85/84, 115/114 or 25/24 in consideration, in that case with a factor between wire baud rate and CCITT interface bit rate of 5/4, 11/9, 17/14 or 23/19 respectively. With the same numerical values, it is also possible to process a 140 Mbit/s digital signal.

Devices for carrying out the method according to the invention appear from the independent claims 9-12.

In the following, the invention will be explained in more detail by way of examples. Fig. 6 shows two sections for time multiplex signals with switching on and off. Fig. 7 shows a switch for the first variant of the method. Fig. 8 shows a disconnector for the first variant of the

the way.

Fig. 9 shows a switch for another variant of the method, and

fig. 10 shows a disconnector for another variant of the method.

In fig. 6 shows a first transmission path in the direction from input 1 to output 6 for a first time multiplex signal and another transmission path in the opposite direction from input 16 to output 11 for another time multiplex signal.

The device contains regenerators 2 and 15 with clock generation, a switch 5, a switch 12, a fork 8 and a two-wire connection 7 for ISDN narrowband signals.

In the switcher 5, the first time multiplex signal with the digital broadband signal DBS1 is supplemented with the narrowband signal ISDN1, the latter being exchanged for filler words. In the disconnector 12, an ISDN narrowband signal ISDN2 is extracted from another time multiplex signal.

The fork 8 ensures that the ISDN narrowband signals ISDNl

and ISDN2 can pass over the two-wire cable with terminal 7.

Fig. 7 shows a practical embodiment of the switch 5 in fig. 6 for the first variant of the method. The switcher 5 contains a multiplexer 17, a 4B/6B encoder 18, a serial-to-parallel converter 19, a frequency divider 20 and a detection device 21 for the overframe marking word URKW.

The time multiplex signal with the broadband signal DBSl and filler word comes via the detection device 21 to the multiplexer 17. The detection device 21 derives a synchronous signal Syncl which is likewise fed to the multiplexer 17. The clock Tl of the time multiplex signal passes over the input 3 of the multiplexer 17 and to the frequency section 20. The latter derives the clock for the series-to-parallel converter 19. The ISDN signal ISDN1 that appears at the input 9 is converted in the series-to-parallel converter 19 into 4B code words which, via the 4B/6B code converter 18, also finally reaches the multiplexer 17. There, the ISDN signal ISDN1 becomes, possibly during replacement of filler words, inserted into the time multiplex signal which can be taken out at output 6. The used bitrates and frequencies can be derived from fig. 7.

Fig. 8 shows a disconnector 12 according to fig. 6 for the first variant of the method. The device contains a detection device 22 with an output 23 for indication and frame synchronization, a demultiplexer 24, a 6B/4B code converter 25, a parallel series converter 26, a phase control circuit 27, a D flip-flop 28 and a frequency divider 29.

The input 13 receives a time multiplex signal with a digital broadband signal DBS2 and a digital ISDN narrowband signal ISDN2. This signal comes via the detection device 22

for the upper frame marking word URKW to the demultiplexer 24. In addition to a signal for identification and frame synchronization at the output 23, the detection device 22 also emits a synchronous signal Sync2 to the demultiplexer 24, which emits partly the digital broadband signal DBS2 to the output 11 and partly 6B codewords for the ISDN2 signal to the 6B/4B code converter 25. This leads the 4B code words to the parallel series converter 26, whose output signal goes to the D flip-flop 28, which is controlled by the phase control circuit 27. At the output 10, the ISDN signal ISDN2 appears. The detection device 22, the demultiplexer 24 and the frequency divider 29 are controlled by the clock via the input 14. The output signal from the frequency divider 29 in turn controls the parallel series converter 26.

Fig. 2 shows a switch 5' according to fig. 6 for another variant of the method. Then at least the dimensioning of the switch in fig. 7 deviates from this variant, all references are provided with an apostrophe.

The switcher 5' contains a multiplexer 17<1>, a 4B/6B encoder 18', a register store 19', a frequency divider 20' and a detection device 21' for the superframe marking word

URKW.

In contrast to the device in fig. 7, four narrowband signals ISDNla to ISDNld are fed there which are converted into 4B code words in the register storage 19'. Apart from other bitrates, this decoupler 5' otherwise works like the decoupler 5 in fig. 7.

Fig. 10 shows a disconnector 2' according to fig. 6 for another variant of the method. Here, too, the reference designations are due to partially deviant mode of operation marked with an apostrophe f. The device contains a detection device 22' for the upper frame marking word URKW and an output 23' for indication and frame synchronization, a demultiplexer 24', a 6B/4B code converter 25', a register storage 26' , four phase control circuits 27a' to 27d', four D flip-flops 28a' to 28d' and a frequency divider 29'.

In contrast to the disconnector 12 in fig. 8, the time multiplex signal at the input 13' contains four ISDN narrowband signals ISDN2a to ISDN2d. These are converted with other bit rates into 4B code words, which in the register storage 26' are converted into four output signals which can be extracted at the outputs 10a<1> to 10d' as ISDN narrowband signals ISDN2a to ISDN2d.

Claims (13)

1. Method for inserting a digital narrowband digital signal into or separating it from a time multiplex signal, characterized in that in the time multiplex signal a wideband signal is transmitted with synchronous words in a binary 5B/6B line code which for some 5B codewords contains the same 6B codeword (selection codeword ) in the positive and the negative group, as well as extra selection code words as filler words, that for the narrowband signal, the 5B/6B line code and exclusively selection code words are also used, that the narrowband signal at the insertion point is synchronously replaced by at least one filler word, and that the narrowband signal at the separation point is determined on the basis of the synchronous word and switched off. 2. Method as stated in claim 1, characterized in that when the time multiplex signal is formed, a 34.368-Mbit/s digital signal with CCITT pulse frames as basic frames is used, that the time multiplex signal with the broadband signal is subjected to 5B/6B line coding, that there five successive basic frames form a superframe, that the frame marker word for the first basic frame in the superframe is located as a superframe marker word in two 6B codewords, and that a fixed number of 6B codewords serve as filler words for each superframe. 3. Method as stated in claim 2, characterized in that a wire-coded narrowband signal with a bit sequence of 6/5 is selected. 716 kbit/s, and that the superframe marking word is directly followed by 32 filler words. 4. Method as stated in claim 1, characterized in that for the formation of the time multiplex signal, the starting point is a 34.368-Mbit/s digital signal with CCITT pulse frame as basic frame. that the baud rate for the time multiplex signal with the broadband signal is increased to 41.242 Mbd with 5B/6B coding, that an overframe is formed from five successive basic frames, that the frame marker word of the first basic frame is located as an overframe marker word in two 6B code words, and that during a further increase in the baud rate, a fixed number of filler words between the 41.242 MBd digital signal's 6B code words. 5. Method as stated in claim 4, characterized in that a line coded narrowband signal with a bit sequence of 6/5 is selected. 2864 kbit/s, and that the wire-coded narrowband signal takes the place of one hundred and three-and-twenty filler words per upper frame distributed at even distances above this, so there is a total baud rate of 44.6784 MBd. 6. Method as stated in claim 1, characterized in that the narrowband signal is sampled at a frequency of more than twice its baud rate, that each sampled random sample is coded with one bit, that four and four bits forming a 4B code word are converted into selection code words, and that the selection code words are synchronously exchanged for the filler words in the time multiplex signal. 7. Method as stated in claims 3 and 6, characterized in that the sampling frequency for the narrowband signal amounts to 572.8 kHz, and that the bits of four directly following random samples that form a 4B code word are converted to one and one sample code word with a baud rate of 6/5 . 716 kBd. 8 Method as stated in claims 5 and 6, characterized in that four mutually synchronous, plesiochronous or asynchronous narrowband signals, each with 572.8 kHz, are sampled simultaneously, and that four and four simultaneously sampled random samples forming a 4B code word are converted into one and one selection code word with baud rate 6/5. 2864 kBd. 9. Method as stated in one of claims 1-8, characterized by inserting at least one ISDN signal into or separating an ISDN signal from a time multiplex signal containing a digital picture signal. 10. Enables the implementation of a procedure as stated in claim 2 or 3, characterized in that a series-parallel converter (19) is arranged which converts a narrowband signal into 4B codewords, that a 4B/6B code converter (18) is arranged, and that a multiplexer (17) is arranged for forming the time multiplex signal. 11. Disables the implementation of a procedure as stated in claim 2 or 3, characterized in that a demultiplexer (24) is arranged to separate the narrowband signal (ISDN2), that a 6B/4B code converter (25) is arranged, that a parallel series converter (26) is arranged for converting the 4B code words, and that a D flip-flop (28) controlled by a phase control circuit is arranged on the output side (27). 12. Enables the implementation of a procedure as stated in claim 4 or 5, characterized in that a register storage (19<1>) is arranged which converts four narrowband signals (ISDNla to ISDNld) into 4B code words, that a 4B/6B code converter (18') is arranged, and that a multiplexer is arranged ( 17') for forming the time multiplex signal. 13. Disables the implementation of a procedure as stated in claim 4 or 5, characterized in that a demultiplexer (24<1>) is arranged to separate the narrowband signals (ISDN2a to ISDN2b), that a 6B/4B code converter (24<1>) is arranged, that a register storage (26<1>) is arranged for converting 4B code words into four narrowband signals (ISDN2a to ISDN2d), and that four D flip-flops (28a' to 28d') are arranged per output (10a' to lOd'), controlled by four phase control circuits (27a' to 27d').