Transmitting Section of PCM Station

My present invention relates to a PCM station of a telecommunication system and, more particularly, to a transmitting section of such a station designed to send messages pertaining to a multiplicity of communication channels in a time-division-multiplex (TDM) mode to a remote terminal.

In commonly owned U.S. Pat. No. 4,081,611 there has been disclosed a PCM station of this type, acting as a coupling network or transit exchange between incoming and outgoing links, in which arriving messages are fed to a line unit via a receiving interface while departing messages pass through a transmitting interface. Between these two interfaces the line unit includes, inter alia, a malfunction detector which is controlled by a timing circuit and can send alarm signals to an insertion circuit enabling their interpolation with outgoing message signals. The station dialogues with a central processor which supervises its operation and sends instructions concerning the connections to be established.

In conformity with international regulations, data transmitted in such a PCM/TDM telecommunication system are organized in a succession of frames each consisting of a multiplicity of channels accommodating respective multibit words, usually 8-bit bytes; a channel may be defined as a time slot subdivided into n clock cycles or phases. For proper synchronization of clocks controlling the operations at the transmitting and at the receiving end of a signal path, certain alignment words are used in an initial channel of each frame (referred to hereinafter as the No. 0 channel). These alignment words generally differ from each other in odd-numbered or "first" and even-numbered or "second" frames; thus, they will alternately assume two different forms referred to hereinafter as "word A" and "word B". Word A, appearing in the No. 0 channel of the recurrent first frame, may have a large number of its bits (e.g. 7 out of 8) arranged in an invariable configuration; word B, present in the No. 0 channel of the immediately following second frame, need only have one particular bit in a predetermined time position.

When data to be sent out from such a station (e.g. in response to instructions from the processor) have not yet been organized into frames, bytes assigned to the No. 0 channel will have to be given bit configurations conforming to these alignment words. Bits not included in the invariable configurations can still be used, however, to carry information to the remote terminal.

An object of my present invention is to provide means in such a PCM station for enabling the use of certain data words, assigned to the No. 0 channel of an outgoing frame, for the conveyance of internal switching instructions to the transmitting section of the station before the transformation of these data words into an alignment word A or B, these instructions utilizing one or more bits which are replaced in the transformation so as not to appear in the alignment word sent to the remote terminal.

A more particular object is to provide means selectively controlled by these instructions for generating supervisory signals of various kinds, specifically a loop command for the establishment of a closed circuit within the PCM station and a remote-supervision command for the testing of conventional regenerators inserted in the line connecting the PCM station to the remote terminal.

A transmitting section of a PCM station according to my invention comprises channel-forming means for organizing multibit data words into alternating first and second outgoing frames as described above, timing means stepped by a local clock for identifying the No. 0 channel of each frame and controlling the channel-forming means to replace certain parts of a data word in that channel by predetermined bit yielding the partly invariable alignment words A and B, decoding means emitting supervisory instructions upon detecting certain bit groupings in a data word assigned to such a channel, and switching means connected to the decoding means for generating internal signals in response to the supervisory instructions thus emitted, the channel-forming means being connected to the switching means for modifying a bit in one of the alignment words (specifically word B) in the presence of these instructions.

Pursuant to a more particular feature of my invention, the channel-forming means comprises a series/parallel converter with n stages (n being preferably equal to 8) accommodating respective bits of an n-bit data word successively arriving at an input thereof, a parallel/series converter with n stages and with a series input connected to a first stage output of the series/parallel converter for sequential loading with the bits of an arriving data word, and a multiplexer with n outputs connected to respective stage inputs of the parallel/series converter and with a plurality of inputs connected to respective stage outputs of the series/parallel converter. Additional inputs of the multiplexer are connected to points of fixed potential reflecting predetermined logical values, the timing means emitting a channel-designating signal which blocks the sequential loading of the parallel/series converter in the No. 0 channel of each frame while enabling the parallel loading thereof via the n multiplexer outputs. Some of these multiplexer outputs are switched to certain of the additional inputs thereof, under the control of a frame-designating signal emitted by the timing means, during the No. 0 channel of the first frame (referred to hereinafter as frame F.sub. A) and to other of these additional inputs during the No. 0 channel of the second frame (referred to hereinafter as frame F.sub.B) while other multiplexer outputs are left connected to the associated stage outputs of the series/parallel converter; the parallel/series converter has an output for sequentially reading out the bits loaded into its stages during any channel of each frame F.sub.A, F.sub.B.

In accordance with another feature of my invention, a plurality of the data inputs of the decoding means are connected to the stage outputs of the series/parallel converter while an enabling input thereof is connected to the timing means in order that the decoding means may receive the contents of corresponding stages of that converter in the presence of the channel-designating signal, i.e. during the No. 0 channel of any frame. The switching means advantageously includes a flip-flop settable by a specific bit combination in these converter stages for emitting a loop command, as described above, and further includes logic circuitry operable in response to at least one other bit combination in these stages for emitting a remote-supervision command which modifies the operation of a message coder receiving the bits that are sequentially read out from the parallel/series converter.

The above and other features of my invention will now be described in detail with reference to the accompanying drawing in which:

FIG. 1 is a block diagram of a transmitting section of a PCM station embodying my invention; and

FIG. 2 shows details of certain components of the transmitting section illustrated in FIG. 1.

As shown in FIG. 1, a transmitting section CT of a PCM station according to my invention comprises a timer T, a channel former FC, a loop detector RL, a message coder CD and a remote-supervision circuit TS. All these components are preferably embodied in a single integrated-circuit chip.

Timer T, essentially consisting of a counter, is stepped by locally generated clock pulses CK and emits various clock signals in the course of a two-frame counting cycle, namely a channel-designation signal T.sub.0 during the No. 0 channel of each frame, a timing signal T.sub.0+1 during the No. 0 and No. 1 channels thereof, and a frame-designating signal ST discriminating between the first frame F.sub.A and the second frame F.sub.B of each cycle. The clock pulses CK also go to the channel former FC and to the message coder CD; channel-designating signal T.sub.0 is transmitted to components FC, RL and TS, the latter also receiving the timing signal T.sub.0+1. The frame-designating signal ST is fed only to the channel former FC which furthermore receives a "raw" bit stream DE as well as an alarm signal AL generated by a malfunction detector. The data of bit stream DE may come from any source not provided with its own channel former. The malfunction detector could be of the type disclosed in my concurrently filed application Ser. No. 278,065 . A loop command L emitted by detector RL, in response to certain bits D.sub.2, D.sub.3 of a data word recognized in channel former FC, controls a nonillustrated switching stage designed to establish a closed circuit between the transmitting and receiving sections of the PCM station for self-checking purposes; command L is also delivered to circuits FC and TS. The latter circuit receives, in addition to the bits D.sub.2, D.sub.3 emitted by channel former FC, a control signal CS from a nonillustrated instruction decoder connected to the associated processor; signal CS, when retransmitted by circuit TS as a signal CS', determines the mode of operation of coder CD as more fully described hereinafter. When signal CS is blocked, this mode of operation is controlled by two signals S.sub.1 and S.sub.2 selectively generated by circuit TS. Data bits DU, constituting bytes organized into frames by channel former FC, are translated by coder CD into outgoing messages OU on a lead 20.

The incoming bit stream DE may be generated under the control of the processor by a microprogram executing a test routine. During remote-supervision operations, in which the processor checks on signals sent back from the line regenerator under test in response to outgoing code words, a corresponding command generated within circuit TS gives rise to a signal Q.sub.2 delivered to channel former FC.

Reference will now be made to FIG. 2 for a more detailed description of components FC, RL and TS. Channel former FC comprises a series/parallel converter SP, a parallel/series converter PS, and a multiplexer M switchable by the frame-designating signal ST from timer T. Each converter essentially consists of an 8-stage shift register. Clock pulses CK are fed directly to the output register constituted by converter PS and via an inverter IN.sub.1 to the input register constituted by converter SP whereby, in any time slot other than the one assigned to the No. 0 channel of a frame by timer T, the two converters are alternately stepped eight times per time slot. Under these circumstances, all bits of the incoming stream DE are immediately read out via a first stage output 21 of converter SP to a corresponding stage input 31 of converter PS whose stages are thus serially loaded. During the time slot assigned to the No. 0 channel, signal T.sub.0 from timer T inhibits such serial loading but enables the eight stages of converter PS to be loaded in parallel via multiplexer M. An extension of lead 21 and a 7-lead multiple 22 carry the contents of all the stages of converter SP, i.e. bits D.sub.1 -D.sub.8, to respective inputs of multiplexer M from whose eight outputs an 8-lead multiple 32 extends to all the stage inputs of converter PS. In frame F.sub.A, however, multiplexer M is switched by the binary signal ST into a position in which seven of its outputs are connected to respective leads of another 7-lead input multiple 23 to which fixed potentials of predetermined logical values "0" and "1" are applied in an invariable pattern W.sub.A. In a specific instance, this pattern has the configuration of 0-0-1-1-0-1-1. Parallel loading, enabled by the trailing edge of signal T.sub.0 in the final phase of the No. 0 channel of frame F.sub.A, will then introduce only the bit D.sub.1 from the first stage of converter SP along with a predetermined 7-bit combination from multiple 23 into respective stages of converter PS whereby an alignment word A is formed. With the disappearance of signal T.sub.0, the eight bits of the alignment word are sequentially read out on a lead 33 as part of the sequence DU, followed by data words serially loaded in the aforedescribed manner via lead 21 into converter PS during the remaining time slots of that frame.

When the No. 0 channel of frame F.sub.B arrives, signal T.sub.0 again inhibits the serial loading of converter PS while the stages of converter SP are being filled with bits D.sub.1 -D.sub.8 of a data word then arriving. With signal ST now in its alternate state, multiplexer M connects two of its outputs--specifically the second and third of the group tied to multiple 32--to inputs other than those served by mutliple 22, namely two inputs respectively connected to a pair of leads 24 and 25. Lead 24 is connected to a source W.sub.B of fixed potential of logical value "1". Lead 25 extends from an OR gate OG.sub.1 carrying a signal ATL of logical value "0" or "1" as described below. Thus, an alignment word B loaded in parallel into the stages of converter PS in the final phase of No. 0 channel of frame F.sub.B has an invariable bit "1" in second position while the bit in its third positon depends on the signal ATL. Therefore, words A and B will have the following configuration:

The letter X in the foregoing Table indicates digits of either logical value read out from converter SP via lead 21 and multiple 22. Since bits D.sub.2 and D.sub.3 are never transferred from converter SP to converter PS via multiplexer M in the present instance, the corresponding leads may be omitted in multiple 22.

Since ATL, designed to alert the remote terminal to a special situation in the PCM station here considered, is high whenever one of the aforedescribed signals L, AL and Q.sub.2 appears on a respective input of OR gate OG.sub.1. When the presence of bit ATL=1 in alignment word B is detected at the remote terminal receiving the outgoing messages OU, e.g. by circuitry such as that described in my concurrently filed application Ser. No. 278,065, that terminal may send back alignment words of its own also containing such an alerting bit ATL. If desired, the closure of an internal loop in response to command L may be communicated to the remote terminal in a more distinctive way, as by a blanking of the outgoing message bits DU in response to a specific bit D.sub.4 read out from converter SP as described hereinafter; in that case, the loop command L need not be fed to OR gate OG.sub.1.

Loop detector RL and remote-supervision circuit TS are shown in FIG. 2 to be provided with a common decoder DC which is enabled by the channel-designating signal T.sub.0 and has two data inputs receiving the bits D.sub.2 and D.sub.3 from the corresponding stages of converter SP. Decoder DC has four output leads 41, 42, 43, 44 which are respectively energized when bits D.sub.2 and D.sub.3 have configurations "00", "11", "10" and "01". High voltage on lead 44 sets a flip-flop FF.sub.1 which is reset through an inverter IN.sub.2 when that voltage goes low. In its set state, flip-flop FF.sub.1 emits a signal Q.sub.1 constituting the loop command L. Energization of lead 41 or 42 establishes one of two possible remote-supervision routines and causes, via an OR gate OG.sub.2, the resetting of a flip-flop FF.sub.2 which is otherwise set, via an OR gate OG.sub.6, by voltage on lead 43 or 44. When reset, flip-flop FF.sub.2 emits the aforementioned signal Q.sub.2 to OR gate OG.sub.1 ; in its set state, flip-flop FF.sub.2 transmits an output signal Q.sub.2 through an OR gate OG.sub.3 to one input of an AND gate AG.sub.1 which is thereby enabled to pass the mode-control signal CS referred to in connection with FIG. 1. OR gate OG.sub.3 also has an input receiving the timing signal T.sub.0+1 during the No. 0 and No. 1 channels of the frame.

AND gate AG.sub.1, when conducting, delivers the signal CS' to message coder CD to establish one or the other mode of operation thereof, depending on the logical value of signal CS. This signal, therefore, controls the coder CD during the first two channels of any frame, irrespectively of the state of flip-flop FF.sub.2. Coder CD, stepped by the clock pulses CK, receives the outgoing bits DU as bits DU' through an OR gate OG.sub.5 and translates them into, say, a pseudo-ternary representation known as AMI (for "alternate mark inversion") when operating in its first mode; in its second mode, the representation may be of another pseudo-ternary type such as that known as HDB3 in which the rule of alternate mark inversion is violated in a well-defined manner. Reference may be made to annex 1 of CCITT Recommendation G.703 for a more detailed description of codes AMI and HDB3.

Circuit TS comprises three cascaded flip-flops FF.sub.3, FF.sub.4 and FF.sub.5. Flip-flop FF.sub.3 is of JK type with a setting input tied to lead 41 and a resetting input joined to an output lead 45 of an OR gate OG.sub.4 with three inputs respectively connected to leads 42, 43 and 44. Flip-flop FF.sub.4 is of D type with a clock input connected to lead 41, a data input connected to the set output of flip-flop FF.sub.3 carrying a signal Q.sub.3, and a resetting input connected to lead 45. Flip-flop FF.sub.5 is of the set/reset type with a switching input joined to the set output of flip-flop FF.sub.4 carrying a signal Q.sub.4. The set outputs of flip-flops FF.sub.4 and FF.sub.5 extend to respective inputs of two AND gates AG.sub.2 and AG.sub.3 which have other inputs receiving the negated timing signal T.sub.0+1 through an inverter IN.sub.3 ; these AND gates, when conducting, respectively emit signals S.sub.1 and S.sub.2 to a pair of control inputs of coder CD.

As long as neither of leads 41 and 42 is energized in the first time slot of a frame, i.e. during the No. 0 channel thereof marked by the presence of signal T.sub.0, flip-flop FF.sub.2 remains set whereby AND gate AG.sub.1 is permanently enabled to give passage to the control signal CS; that signal, accordingly, then dictates the mode of operation of coder CD also in all the later channels of the frame.

If an instruction for the initiation of remote supervision causes bits D.sub.2 and D.sub.3 to assume the configuration "00" or "11" in the No. 1 channel, the resetting of flip-flop FF.sub.2 blocks the AND gate AG.sub.1 after the end of the No. 1 channel. If lead 41 is energized in the first time slot, by a pulse coextensive with signal T.sub.0, flip-flop FF.sub.4 (assumed to have been previously reset by voltage on lead 45) is not switched since flip-flop FF.sub.3 is set only by the trailing edge of that pulse. When the same pluse on lead 41 recurs at the beginning of the next frame, signal Q.sub.3 is already in existence so that flip-flop FF.sub.4 is set to generate signal Q.sub.4. In the first frame of a remote-supervision routine commanded by a bit combination "00", therefore, AND gate AG.sub.2 is cut off whereas AND gate AG.sub.3 may or may not be conducting after the second time slot, depending on the state last reached by flip-flop FF.sub.5. With mode-changing signal S.sub.2 =0, coder CD will operate in its first mode (AMI) regardless of the value of signal S.sub.1 as long as AND gate AG.sub.1 is blocked to cut off the control signal CS; it should be noted, however, that signal S.sub.2 would be overridden by signal CS'=1 when gate AG.sub.1 conducts to establish the second mode (HDB3). In the following frames, as long as the pulse on lead 41 recurs during the No. 0 channel thereof, the presence of signal Q.sub.4 makes S.sub.1 =1 after the No. 1 channel; this establishes the second mode of operation. Since the appearance of signal Q.sub.4 has reversed the flip-flop FF.sub.5, signal S.sub.1 has changed its logical value; this signal is relevant only for the second mode by determining the polarity of the violation of the alternate-mark-inversion rule discussed above. Thus, a subsequent recurrence of the same remote-supervision routine (with intervening resetting of flip-flops FF.sub.3 and FF.sub.4 by a pulse on any of the other decoder outputs) will result in a variation of mode HDB3 compared with the previous instance. If such a variation is not required, flip-flop FF.sub.5 and gate AG.sub.3 can be omitted.

If the remote-supervision routine is commanded by the bit combination "11", the pulse on lead 42 resets (or holds reset) the flip-flop FF.sub.3 so that neither signal Q.sub.3 nor signal Q.sub.4 will appear in any frame. Mode AMI is then established in each frame, at least from the third time slot on.

Coder CD may be divided into two parts to which the outgoing bits DU' are fed in parallel for transformation into the AMI and the HDB3 representation, respectively, one or the other part being switched to its output 20 by a multiplexer responsive in the aforedescribed manner to signals CS' and S.sub.2. Signal S.sub.1 will then control only the operation of the part generating the HDB3 mode.

FIG. 2 further shows a D-type flip-flop FF.sub.6 whose clock input receives the signal T.sub.0 and whose data input is connected to the fourth stage of converter SP carrying the bit D.sub.4. The set output of flip-flop FF.sub.6 is connected to another input of OR gate OG.sub.5 giving passage to the outgoing bits DU. In the absence of a loop command L, flip-flop FF.sub.6 is inhibited by a signal Q.sub.1 fed to its resetting input from the reset input of flip-flop FF.sub.1. When a loop command is generated, flip-flop FF.sub.6 responds to the trailing edge of signal T.sub.0 to emit a signal Q.sub.6 to OR gate OG.sub.5 if at that instant the bit D.sub.4 has the logical value "1". Signal Q.sub.6 then blanks the bits DU since the output lead of OR gate OG.sub.5 will feed a continuous high voltage to the data input of coder CD until command L disappears and/or D.sub.4 =0 in the No. 0 channel of a subsequent frame. It will be apparent that similar blanking effects can be obtained by replacing OR gate OG.sub.5 with an AND gate having an input connected to the reset output of flip-flop FF.sub.6, for example.

Coder CD could be part of a transmitting interface of a line unit, such as that shown at IT in FIG. 6 of the aforementioned U.S. Pat. No. 4,081,611, if the bit stream DU' is interleaved with message frames arriving from another remote terminal at the associated receiving section and passing through a switching matrix (such as that shown at RS in the patent referred to) to an insertion circuit (CI) disposed between gate OG.sub.5 and the coder.