WO1988003738A1

WO1988003738A1 - Method and digital switching networks for fault supervision

Info

Publication number: WO1988003738A1
Application number: PCT/SE1987/000511
Authority: WO
Inventors: Sture Gösta ROOS
Original assignee: Telefonaktiebolaget L M Ericsson
Priority date: 1986-11-06
Filing date: 1987-10-30
Publication date: 1988-05-19
Also published as: AU8233987A; SE8604763L; SE8604763D0; SE455459B

Abstract

Method and means for controlling digital message information through a switching network, which has at least partially duplicated equipment (DDM-A, DDM-B) enabling the establishment of at least partially duplicated transmission paths and transmission of message information with fault control via each of the duplicated transmission paths. In a method in accordance with the invention the read addresses for a speech store (SS) are also controlled for each transmission path and in a similar way as control of the message information. In detection of an error in a read address the read-out information is transmitted together with fault control result information indicating that a fault affecting this message information has been detected. A switching network in accordance with the invention has control means (CH) and information selection means (PS) for this purpose.

Description

METHOD AND DIGITAL SWITCHING NETWORKS FOR FAULT SUPERVISION

TECHNICAL FIELD

The present invention relates to a supervision method in transmitting digital message information via a digital switching network, where at least partially duplicated information transmission paths are established in the switching network with the message information being transmitted under fault supervision over each of the duplicated information transmission paths.

The present invention also relates to a digital switching network with fault control means and transmitting digital message information, this network having at least partially duplicated equipment enabling establishing at least partially duplicated information transmission paths in the switching network and transmission of the message information over each of the duplicated information transmission paths.

BACKGROUND ART

For increasing the reliability of a switching network it is well known to at least partially duplicate equipment for enabling establishing at least partially duplicated information transmission paths in the switching network and at least periodically transmitting certain information over each of the duplicated information transmission paths. In such a case it i3 well known to select with the aid of error supervision control means which of the duplicated transmission paths or information transmitted over duplicated transmission paths shall be used. Amongst the patent literature relating to such switching networks and methods for information transmission can be mentioned such as US 3851105, US 3920914 and US 3974340.

US 4543651 teaches that on detecting faults affecting the message information transmitted on a duplicated transmission path, the information transmitted on the path can be formed so that the affected message information is transmitted together with fault control result information, which indicates that faults affecting the message information in question have been detected. A switching network according to US 4543651 also includes information selection means adapted to select, with the aid of the fault control result information transmitted via the respective transmission path, between the information transmitted via one of the duplicated transmission paths and the corresponding information transmitted via the other duplicated transmission path, regarding which information that shall be expedited via the switching network. According to this patent, such information transmitted via a duplicated transmission path may be selected as is not affected by faults, according to the switching network fault supervision.

In the transmission of digital message information according to the above- mentioned publications, the message information is written into a speech store for the respective transmission path on memory positions given by write addresses supplied to the respective speech store. Further information for error supervision and/or control of the information selection means and pertaining to the digital message information is simultaneously written into the speech store. For transmission over the respective duplicated transmission path there is further read from the respective speech store information from memory positions which are given by read addresses which have been supplied to the speech store from a control store associated with the respective transmission path.

US 4393490 teaches, in connection with error control and plane selection, providing the read addresses with, inter alia, a parity bit on writing read addresses into a control store. On reading the addresses from the control store errors in them can be detected with the aid of the parity bit associated with the respective read address. On detecing an error in a read address that has been read, a special code word is expedited with intentional erroneous parity in the respective plane, instead of a word read out from a memory position in the speech store. The code word is unique for the respective control store and has intentionally erroneous parity, whereby both plane selection and identification of the faulty control store can take place.

DISCLOSURE OF INVENTION

The method of transmitting digital information and a switching network with at least partially duplicated equipment and fault supervision according to the above-mentioned publications can have, depending on the special detail embodiment in the individual case, more or less varying essential advantages apparent from these publications. A disadvantage with some known methods and switching networks is, however, that in spite of the duplication of parts of the equipment such as speech store and control store, they may have deficient capacity for detecting and remedying both faults originating from the control stores or corresponding means for generating and supplying read addresses to the respective speech store and faults occurring through other equipment pertaining to the respective transmission path.

More particularly, some known methods and switching networks with duplicated transmission paths and equipment of that special kind have the drawback that essential parts of the duplicated equipment are tied to a certain one of the duplicated transmission paths. Accordingly, and by way of example, the task of a malfunctioning control store of delivering read addresses to the speech store for one transmission path cannot be simply taken over by a properly functioning control store associated to the other transmission path.

More specifically, some known methods and switching networks of the kind in question also have the drawback that they are not self-remedying for occasional errors in read addresses stored in the control stores, whereby the same error in a read address supplied to a speech store may occur a large number of times after each other under the duration of a telephone conversation or a data connection, for example.

Some known methods and switching networks with duplicated transmission paths and fault control of the kind in question further have the disadvantage from the maintenance aspect that they require a special memory and/ or a processor for updating the control store with up-to-date read addresses after replacing the store.

The present invention has the object of solving the problem with the deficient ability of some known methods and networks to detect and remedy both errors arising from faults in the control store or corresponding means generating and supplying read addresses to the speech store for the respective duplicated information transmission path and faults occurring in other equipment pertaining to the respective transmission part.

It is also an object of the present invention to eliminate at least certain of the more specifically mentioned drawbacks above with the aid of preferred embodiments of the method and switching network.

It is further an object of the present invention to provide a method and a switching network where errors in the message information and faults in the control stores/read addresses can be detected and remedied in a similar way and with similar means, whereby fault supervision will be simpler and cheaper.

Somewhat simplified and summarily expressed, a method and a switching network in accordance with the invention could be said to be built on the idea of supervising the read addresses in a similar way as with the message information, and on detecting errors in a supplied read address to inform and act on the information selection means via the respective transmission path in a similar way as with the detection of errors in the message information on the transmission path corresponding to the read address. In the detection of errors in a read address supplied to a speech store, fault control result information is then transmitted, in accordance with the invention, together with the information read from the speech store with the aid of the erroneous read address, the former information indicating that an error pertaining to the information read out has been detected. A switching network in accordance with the invention has address control means for carrying this out.

Alternatively, also somewhat simplified and summarily expressed, at least certain embodiments of the method and switching network in accordance with the invention could be said to be based on the idea that the read addresses or the control stores shall not be reserved exclusively for the speech store for a given transmission path, but could be utilized for the speech store for either of the duplicated transmission paths. In such embodiments pairs of mutually corresponding read addresses are read from the control store for the respective transmission path, these addresses then being subjected to error supervision, a choice being made for each transmission path between corresponding read addresses in the respective pair. In appropriate cases, an error-free read address is thus selected from respective pairs, this read address being supplied to the respective speech store. A switching network in accordance with the invention may thus have, inter alia, read address selection means for carrying out this selection between corresponding read addresses.

According to a further alternative way of looking at at least certain embodiments of the method and switching network in accordance with the invention, they could be said to be based on the idea of utilizing the control store or the read addresses pertaining to the speech store of one of the duplicated transmission paths for maintenance and updating of a control store or read addresses for a speech store of the other of the duplicated transmission paths. In such embodiments, there is written into a control store for one transmission path a read address which is error- free according to the fault control, this address being read from a control store for the other transmission path. An inventive switching network can have, inter alia, read address replacement means for carrying this out.

What is more correctly distinguishing for a method and a switching network in accordance with the invention and preferred embodiments of such is apparent from the independent claims and the dependent claims.

Forming a method and a switching network in accordance with the invention, and particularly in accordance with preferred embodiments thereof, results in several special advantages, of which the following may be mentioned: At least single errors which do not occur simultaneously in the read addresses of both transmission paths can be detected and remedied. In the case of errors in a read address or a control store pertaining to one of the duplicated transmission paths, a corresponding read address or corresponding control store for the other of the duplicated transmission paths can replace the faulty one.

After the repair/exchange of a faulty control store for one transmission path, no extra memory or processor is required for updating the control store, and this can take place autonomously without central tables or the like with the aid of the control store for the other transmission path. The respective transmission path and existing equipment for transmitting fault control result information relating to message information can be used also for fault control result information relating to read addresses. Further to the above-mentioned special advantages, a method and a switching network in accordance with the invention can furthermore have the majority of the advantages of known methods and switching networks with duplicated transmission paths and duplicated transmission of the kind in question. Selection of transmission path and selection between transmitted information can take place individually for each sample or digital word of the message information. The method and the switching network in accordance with the invention are not limited to a given method for detecting faults or to fault control merely in connection with the speech stores and read addresses, but It is possible for every device included in the switching network to be controlled, as required by an adequate method for the participating means, e.g. parity, coding, level sensing etc. It is not necessary to block certain faulty units such as the control stores with limited or occasional faults. Since a faulty unit does not usually need to be blocked, the nonfunctioning parts of the network will be much smaller and reliability will thus be better. A method and a switching network in accordance with the Invention enables direct Indication of a faulty unit without the faults spreading through the network and thus causing a number of alarms in different units. After a first fault affecting given message information a transmission path is detected, further faults occurring later in the transmission path affecting the same message information can also be detected.

Further possible advantages of a method or a switching network in accordance with the Invention will be understood by one skilled in the art after studying the described embodiments.

BRIEF DESCRIPTION OF DRAWINGS

Figure 1 is a heavily simplified block diagram of a first embodiment of a method and switching network in accordance with the invention. Figure 2 illustrates an embodiment of the format of digital words which can be transmitted with a method and a switching network according to Figure 1.

Figure 3 is a simplified block diagram of a first embodiment of a fault control means for a method and a switching network in accordance to Figure 1 in transmitting digital words according to Figure 2.

Figure 4 is a simplified block diagram illustrating the use of an embodiment of the fault control means for controlling read and write addresses in a switching network according to Figure 1 or 6 in transmitting digital words according to Figure 2.

Figure 5 is a simplified block diagram of an embodiment of an information selection means for a method and a switching circuit according to Figure 1 in transmitting words according to Figure 2 and having fault control with means according to Figure 3. Figure 6 is a greatly simplified block diagram of a second embodiment of a method and a switching network in accordance with the invention

BEST MODES FOR CARRYING OUT THE INVENTION

Figure 1 is a greatly simplified block diagram of how equipment in a digital switching network is duplicated to enable establishing of duplicate information transmission paths between two information processing units denoted DEVU and transmission of information under fault supervision via each of the duplicated information transmission paths from left to right in the figure.

There is a transmission module DDM-A and DDM-B for each of the duplicated information transmission paths. Each transmission module includes a multiplexing unit MU, a speech store unit SSU, a demultiplexing unit DMU, a write address unit WAU, a control store unit CSU, a read address unit RAU and a control unit SNC.

The information processing unit DEVU, from which the message information is to be transmitted, is connected via information transmission links, e.g. electrical lines or optical links, to a multiplexing unit MU in each transmission module. For fault supervision of the information which is sent from the information processing unit on the respective transmission link to the respective multiplexing unit the information processing unit DEVU is provided with the fault control means CH. The multiplexing unit for the respective transmission path has fault control means CH for fault supervision of the information which is received from the information processing unit via the transmission link. In addition, the respective multiplexing unit Is formed for connection to a further plurality of transmission links for receiving and fault supervision of information from further information processing units not illustrated, e.g. seven such units. For the sake of space, only one further connection with fault control means CH of the respective multiplexing unit is illustrated in Figure 1, however.

The multiplexing units are connected via transmission links to the respective speech store unit SSU for the same transmission path, and are arranged to transmit in time multiplex to the respective speech store unit information supplied to the respective multiplexing unit from connected information processing means via the transmission links. For fault supervision of the information sent from the respective multiplexing unit on the respective transmission link the multiplexing units have the fault control means CH.

The speech store unit SSU for the respective transmission path has the fault control means CH for fault supervision of the information received from the respective multiplexing unit via the transmission links. The speech store units are connected via transmission links to the respective demultiplexing unit DMU for the same transmission path and are adapted to transmit to the respective demultiplexing unit via the transmission links information which was supplied to the respective speech store unit from the multiplexing unit for the same transmission path. The speech store units have the fault control means CH for fault control of the information which is sent from the respective speech store unit on the respective transmission link to the respective demultiplexing unit DMU.

The demultiplexing unit DMU for the respective transmission path also has the fault control means CH for fault supervision of the information received from the respective speech store unit via the transmission links. Via transmission links, the demultiplexing units for both transmission paths are connected to the righthand information processing unit DEVU, to which message information is to be transmitted via each transmission path from the information processing unit DEVU to the left in Figure 1. The demultiplexing units have the fault control unit CH for fault supervision of the information from these units which is sent on the transmission links to the information processing unit. The respective demultiplexing unit DMU is further formed for connecting a further plurality of transmission links for transmitting and fault supervision of information to further information processing units not illustrated in Figure 1, e.g. seven such units. For the sake of space only one further connection to fault control means CH of the respective multiplexing unit is illustrated in Figure 1, however.

The information processing unit DEVU connected to the demultiplexing units has the fault control means CH for supervising the information received via the transmission links from the respective demultiplexing unit. This information processing unit also includes the information selection means PS for selecting between the information transmitted via one transmission path through DDM-A and the corresponding information transmitted over the other information path through DDM-B.

Apart from error supervision, the transmission module for each transmission path may generally have the same tasks, and generally function according to the same principles as a conventional time switch module in a telephone exchange or a switching network, e.g. for circuit switched data connections. The respective transmission module thus includes a write address unit WAU adapted to supply the speech store unit SSU with write addresses for the speech store, a control store unit CSU adapted for supplying the speech store unit with read addresses for the speech store, and a control unit SNC for information exchange with other units in the respective transmission module and at least controlling writing the read addresses for the speech store in a control store CS in the control store unit. There is a read address unit RAU for reading addresses written into the control store.

The speech store unit includes, as previously mentioned, a speech store for receiving and storing digital information word by word on memory positions given by write addresses from the write address unit. The speech store unit is also adapted for reading stored words from memory positions in the speech store given by read addresses from the control store unit. The units illustrated in Figure 1 together with units functioning according to known principles for the supply of message information to the lefthand DEVU in Figure 1 and for further expedition of message information from the righthand DEVU, possibly together with further information processing means connected to the respective transmission module, may constitute a greater or less part of a digital switching network for transmitting digital message information, e.g. digitally coded sound or data. In such a case the respective transmission module DDM may function as a time switch stage in establishing time multiplex connections through the switching network, e.g. telephone calls between subscribers connected to the switching network. The function of time switch stages and the establishment of time multiplex connections through switching networks with such stages is well known and is therefore not described here.

The fault supervision in a method and switching network according to Figure 1 is based on fault control result information FIB being sent via each of the duplicated transmission paths by the switching network together wtih the message information. FIB can indicate that some kind of fault affecting the message information has been detected or has not been detected. The respective fault supervision means therefore has the task, in detecting faults affecting the message information transmitted on the transmission path, of forming the information transmitted on the transmission path so that the affected message information is transmitted together with the FIB, indicating that a fault affecting this information has been detected. For this fault supervision, and for controlling respective information selection means PS the information is transmitted via the respective transmission path in a switching network according to Figure 1, preferably in the form of words which Include the message information and FIB, and with a format according to Figure 2.

Since fault supervision according to Figure 1 occurs in several places, i.e. six along the transmission path within the respective transmission module, it may be that two or more different errors affecting the same message information occur during the transmission along a transmission path. Already before the error supervision at a given place the message information can therefore be accompanied by FIB indicating that a fault has been detected. If a fault control means detects a new fault affecting the same message information, this means does not need to alter the FIB, which can be expedited further without being changed.

The digital information transmitted via the respective transmission path preferably has the form of digital words, each of which includes a part D with the message information, a part FIB with the fault control result information, a part S with control or signalling and a part P with information facilitating fault detection. Each word preferably consists of twelve binary digits of which eight, DO, Dl, ... D7 are used for the message information, one binary digit for FIB, one binary digit for parity check P and the two remaining binary digits SO, SI, which are used for other purposes.

A first embodiment of the fault control means CH for fault control of words with twelve binary symbols and a format according to Figure 2 is illustrated in Figure 3. This means is intended for controlling signals which, in a multiplexing unit, demultiplexing unit or information processing unit, are transmitted in parallel or are received in parallel from a transmission link. The fault control means in Figure 3 may also be used for fault control of signals in a speech store unit or a control store unit. The fault control means CH according to Figure 3 has twelve inputs IN for receiving twelve incoming signals DO, Dl, ... D7, SO, SI, FIB and P which together represent a binary word with the format according to Figure 2. In addition, the means has a further input EIN for optional reception of an incoming fault signal or fault control result information from an optional further fault control means.

The fault control means according to Figure 3 has thirteen outputs OUT for an outgoing fault or alarm signal E as well as twelve outgoing signals DO, Dl, ... D7, SO, SI, FIB and P, the outgoing signals together representing a word with the format according to Figure 2.

The fault control means includes a control device PCH connected for receiving the incoming signals representing a word and adapted to carry out a parity check and possibly further supervision according to some adquate method, e.g. the level of the signals. The control device PCH has a fault signal output for sending a fault signal on detecting that one of the incoming signals individually and/ or the relationship between the signals, e.g. parity, is of a kind such as to be regarded as faulty according to the check method. The fault signal output of the control device is connected to the fault control means fault or alarm output EOUT and also to the input of a setting device FIBR for the fault control result information FIB. The setting device has two further Inputs, of which one is connected to an input FIBIN for receiving the signal FIB in a received word, this signal FIB representing the fault control result information. The third input of the device FIBR is connected for optional reception of an optional fault signal from another fault control means via the input EIN. The setting device FIBR has an output for sending fault control result information FIB and is adapted to send an output signal on this output indicating that a fault has been detected If at least one signal on one of its three inputs indicates that a fault has been detected.

The fault control means according to Figure 3 includes a parity regenerating device PR with eleven inputs, of which ten are coupled for receiving the incoming singals representing DO, Dl, ... D7, SO and SI in an incoming word. The eleventh input of the regenerator PR is connected for reception of the signal on the output of the setting device FIBR. The regenerator has an output and is adapted to give a signal on this output representing a binary value such that the parity of the word sent on the twelve outputs D0OUT ... POUT after fault supervision will be correct irrespective of whether FIB has been changed or whether a fault has been detected by the fault control means.

Logical circuits are commercially available which generate on one output a parity signal, the binary logical value of which depends on the parity of a plurality of logical signals on their respective inputs of the circuit. Such circuits could be used both for parity checking and for parity regeneration in a fault control means according to Figure 3.

With the fault supervision described so far, certain such errors in the transmitted information can be detected which display themselves in the form of some measurable fault in the words transmitted on the respective transmission path. What faults which are measurable naturally depends on the configuration of the fault control means used. Parity faults or signal level fault can be detected, for example. However, faults in the respective transmission module DDM-A or DDM-B may occur which do not display themselves in the form of some fault measurable in a single word or the words transmitted by the respective transmission paths. For example, if the respective speech store unit is supplied with erroneous read addresses giving existing but unintended memory positions in the respective speech store, words can be read from the respective speech store which words have parity, signal level etc which are individually correct but are stored in some other memory position than the intended one. A similar fault can occur if the respective speech store unit is instead supplied with erroneous write addresses giving existing but not intended memory positions in the respective speech store, whereby words, the parity., signal level etc of which are individually correct are written into a memory position other than the intended one in the respective speech store. Such words stored in wrong memory positions will naturally not be read out later at the right occasion if the right read address is then supplied to the speech store unit. Instead, other information stored in the right memory position will be read out, this information having the right parity signal level etc. In a method and a switching network in accordance with the present invention the read addresses and possibly also the write addresses supplied to the speech store unit are therefore subjected to supervision.

In the form of a simplified block diagram, Figure 4 illustrates how control of write and read addresses supplied to a speech store SS can take place in connection with writing and reading words into and from the speech store. The speech store is connected for receiving words coming from a transmission path via a fault control means CHIN, which in turn is connected to receive on twelve separate inputs IN the incoming signals DO, Dl, ... D7, SO, SI, FIB and P, representing together an incoming word with the format according to Figure 2. For receiving write addresses giving the memory positions in the speech store, in which positions the incoming words are to be written, the speech store is connected to a fault control means CHWA for write addresses. The fault control means CHWA is connected to receive on its ten inputs write address signals A0, Al, ... A7, FIB and P, together representing write addresses with associated fault control result information and parity. The means CHWA is adapted to carry out a fault check of the received write address signals according to an adequate control method, e.g. for parity and signal level, and in detecting what can be regarded as fault, according to the control mehtod, to give an alarm or fault signal WAE indicating that a fault affecting the write address has been detected by CHWA. The means CHWA further expedites the signals A0, ... A7 to the speech store but not the signals FIB and P. Instead, the means CHWA has an output connected to the means CHIN for transmitting a signal FIBWA representing the fault control result information FIB of the write address.

For fault supervision of outgoing words which after reading from the speech store are to be further expedited, the outputs of the speech store are connected to a fault control means CHOUT, which in turn has twelve outputs OUT for twelve signals DO, Dl, ... D7, SO, SI, FIB and P representing together outgoing words which are to be expedited further. For receiving read addresses giving the memory positions in the speech store from which words are to be read the speech store is coupled to a fault control means CHRA for read addresses. The means CHRA is circuited to receive on its ten inputs read address signals A0, Al, ... A7, FIB and P representing together read addresses with associated fault control result information and parity. It is adapted to carry out a fault check of the received read address signals according to an adequate control method, e.g. for parity and signal level, and to send, on detecting what is to be regarded as a fault according to the control method, an alarm or fault signal RAE indicating that a fault affecting a read address has been discovered by it. It further expedites the signals A0, ... A7 to the speech store but not the signals FIB and P- Instead it has an output connected to CHOUT for transmitting a signal FIBRA representing the read address fault control result information FIB.

The fault control means CHIN and CHOUT are adapted to carry out a check on the incoming signals received on the inputs IN and from the speech store individually and/or together according to an adequate control method, e.g. parity check and signal level check, and on detection of a fault, according to the fault control method, to send respectively an alarm or error output signal INE or SSE indicating that CHIN has detected a fault in the incoming signals or that CHOUT has detected a fault in the signals from the speech store. The means CHIN is furthermore adapted to send to the speech store SS the fault control result information FIB indicating that a fault affecting a word which is to be written into the speech store has been discovered if either the signal applied to the input IN on the CHIN representing the fault control result information FIB of the received word indicates that a fault affecting the word has been discovered earlier, or the means CHIN itself detects a fault in the supplied incoming signals representing the word, or a fault is detected by

CHWA in the signals representing the address for the memory position in which the word is to be written into the speech store, CHWA then sending a signal

FIBWA to CHIN. In a corresponding manner the means CHOUT is adapted such that for the word which is further expedited after reading from the speech store it sends the fault control result information FIB indicating that an error affecting the respective word has been detected, if either the signal from the speech store representing FIB indicates that a fault has been detected earlier, or if CHOUT itself detects a fault in the signals which are received from the speech store, or if a fault is detected by CHRA in the signals representing the address for the memory position in the speech store from which the word is read out, CHRA then sending a signal FIBRA to CHOUT.

Depending on the type of transmission link between MU, SSU and DMU, CHIN and/ or CHOUT can be connected to the respective transmission link directly or via conventional means such as signal converters, buffer stores, amplifiers etc.

Each of the fault control means CHIN, CHOUT, CHWA and CHRA can, but do not necessarily need to be implemented in a similar way as the fault control means according to Figure 3, and include setting devices FIBR, regeneration devices PR and control devices PCH. In addition to the devices illustrated in Figure 3 CHIN, CHWA and CHRA also preferably include means for ensuring that all signals representing a word, which is to be written and its associated write address is supplied to the memory in the right time relationship as well as that all signals representing a read address are supplied to the memory in the right time relationship. Since such means and writing and reading into speech stores are well known in similar connections, this is not described further here.

The fault control means CHIN, CHOUT, CHWA and CHRA together achieve that the fault control information FIB in a word which is further expedited after writing in and reading out from a speech store will indicate that a fault affecting the word is detected, irrespective of whether a fault has already been detected earlier and/or whether a fault is detected first during writing into or reading from the speech store. In fault supervision with means according to Figure 4 and a word format according to Figure 2, it is not apparent from the FIB in a word where the fault has been detected. On the other hand, it is apparent from the alarm or fault signals INE, SSE, WAE and RAE if any, and in such a case which means CHIN, CHOUT, CHWA or CHRA has detected a fault. If a fault has been detected earlier before the word was supplied to CHIN and no further fault is detected by any of CHIN, CHOUT, CHWA or CHRA no fault signal INE, SSE, WAE or RAE will indicate that any furhter fault has been detected. The fault control means in Figure 4 functions in this respect as corresponding means in Figure 3. Irrespective of whether a fault has been detected earlier and/or detected by CHIN, CHWA, CHOUT or CHRA the means CHIN and CHOUT achieve that words supplied to the speech store SS for writing in and words which after reading out from the speech store are further expedited have the right parity. In this way, a fault in a transmitted word does not automatically give rise to fault signals in further means along a transmission path.

In a method and a switching network according to Figures 3 and 4, fault control result information indicating whether some kind of fault affecting the message information has been detected can be sent together with the message information via each of the duplicated transmission paths with the aid of fault supervision and fault controlling means according to the figures. An information selection means PS can therefore readily select, with the aid of the fault control result information FIB in appropriate cases, the information which according to the fault control result information FIB is not effected by any detected fault. An embodiment of such an information selection means for words with a format according to Figure 2 is illustrated in Figure 5.

The information selection means according to Figure 5 includes eleven switches SWO, SW1, ... SW10 each with two signal inputs, a control input and a signal output, there also being included a switch controller SWC with two signal inputs, a control output and a signal output. Each switch and switch controller is adapted to receive a binary signal ADO, AD1, ...AD7, ASO, ASH, AFIB or AP on one signal input, representing a binary digit in a word transmitted via one of the duplicated transmission paths. Each switch and switch controller is also adapted to receive on its other signal input a signal BDO, BD1, ... BD7, BSO, BS1, BFIB or BP representing the corresponding binary digit in a corresponding word transmitted via the second transmission path. On its first signal input the switch controller is adapted to receive the signal AFIB representing the fault control result information FIB for the word transmitted via the first transmission path, and on its second signal input to receive the signal BFIB representing the fault control result information FIB for the word transmitted via the second transmission path. The controller is adapted, in response to received fault control result information on its control output, to send a binary control signal to the control input on all eleven switches giving these switches equal control, so that either all eleven further expedite to the output OUT the signal from the first transmission path, or all eleven further expedite to the output OUT the signal from the second transmission path. The switch controller simultaneously further expedites to the output OUT the fault control result signal from the corresponding transmission path. The signals DO, Dl, ... D7, SO, SI, FIB and P occurring on the twelve outputs OUT of the information selection means thus together represent either the whole word transmitted via the first transmission path or the whole word transmitted via the second transmission path.

The switch controller is adapted to control the switches SW such that signals representing a word where its FIB indicates that a fault has not been detected are further expedited to the outputs of the information selection means when the FIB for the word transmitted via one of the transmission path indicates that a fault has been detected simultaneously as the FIB for the word transmitted via the other transmission path indicates that a fault has not been detected. Several alternatives are conceivable for selection in the case where the FIB of both transmitted words indicates that a fault has been discovered or that the FIB of none of the transmitted words indicates that a fault has been detected. For example, the controller can be adapted to make a random selection, or select words alternately from the first path and the second path. Alternately, the word which has been transmitted via a given one of the transmission paths can always be selected when the FIB of both words or neither words indicates that a fault has been detected. Examples of corresponding control are to be found in the above-mentioned patent specification US 4543651.

Figure 6 illustrates an embodiment of a method and a switching network in accordance with the invention, where a control store unit CSU for either transmission path can be used for supplying read addresses to each of the speech store units SSU of the paths. For each path there is an information selection means PS connected for receiving read addresses not only from the control store unit for the same path but also from the control store unit for the other path. The read addresses are supplied to the speech store unit for a transmission path from the information selection means for this path.

In a method and switching network according to Figure 6, pairs of mutually corresponding read addresses are read, from a control store CS in the control store unit for the first path and from a control store in the control store unit for the second transmission path, with the aid of corresponding addresses from the respective RAU. Addresses from the respective RAU which are supplied to the respective control store are controlled with the aid of the fault control means CH. The read-out pairs of corresponding read addresses ^" are fault controlled with the aid of fault control means CH in the respective control store unit. The fault control means CH in CSU can be of the kind previously described for the speech store SS in Figure 4. With the aid of the fault supervision and the fault control result information FIB thus generated, a selection is made by the respective information selection means PS associated with the respective transmission path between corresponding read addresses in respective pairs, which read address that is to be supplied to the respective speech store unit. If an error is detected in fault supervision in only one of the read addresses, the error free read address is selected. If no error is detected in any of the read addresses in a pair, the respective information selection means can, for example, select the read address coming from the control store unit in the same transmission module. If errors are detected in both read addresses in a pair, the selection of read address can, for example, take place ramdomly or a read address can be selected alternatingly from the first control store unit and the second control store unit. Other variations are conceivable. The information selection means can here be of the kind previously described in connection with Figure 5.

In a method and switching network according to Figure 6, repair to and exchange of such as the control store in a control store unit for one of the duplicated transmission paths can take place without the path needing to be blocked if, in connection with repair and exchange, the read addresses from the control store unit in question contain fault control result information indicating that errors affecting these addresses have been detected. This can be achieved by the transmission of signals to the fault control means of the control store unit being interrupted, whereby at least the signal level and sometimes the parity also are detected as being faulty by the fault control means. The fault control means CH will then send read addresses the FIB of which indicate that a fault has been detected.

For a control store in a transmission path to be taken into operation after repair or replacement it must contain up-to-date switching information. Such information is available in the control store for the other transmission path. The switching information appears at the output of the information selection means for the transmission path in the form of a train of successive read addresses, at least when the FIB associated with these addresses does not indicate that a fault has been detected simultaneously as the FIB associated with the read addresses from the repaired or replaced control store indicate that a fault has been detected. By feeding back the read addresses successively occurring on the output of the information selection means and successively writing these read addresses into the repaired or replaced control store, it may be updated without aid of special tables or processors. For this purpose a transmission module DDM for a transmission path according to Figure 6 includes a read address feedback means CDU, which is connected a) to the read address selection means PS for the transmission path for receiving selected read addresses, b) to the control store for the transmission path and c) to the control unit SNC for the transmission path for replacing in the control store, by writing in selected read addresses without detected faults, existing corresponding read addresses in which faults have been detected.

Writing in such selected read addresses is performed at memory positions in the control store given by the write addresses from the respective control unit

SNC and preferably in a similar way as writing in new read addresses in conjunction with establishing new connections in the switching network. Such writing in of read addresses is well known and is therefore not described here.

To avoid that out-of-date switching information, in the form of earlier correct read addresses, which were previously written into a control store before repairs or replacement, after the repair or replacement has been carried out, are selected by an information selection means instead of correct up-to-date read addresses from the other control store, all the read addresses written into the control store in conjunction with a repair or exchange having been started or terminated are provided with an FIB indicating that faults affecting each of these read addresses have been discovered. After repairs or replacement these read addresses can be successively and automatically replaced by corresponding up-to-date read addresses with the aid of the read address feed-back means CDU.

A method and a switching network according to Figures 1-6 can be modified in many ways within the scope of the invention. For example, there may be fault supervision of read addresses but not of write addresses. Alternatively, with the aid of further fault control means CH in the respective write address unit WAU and further information selection means PS connected between the write address units and the speech store units SSU, a selection can be made between pairs of corresponding write addresses from the write address units regarding which write address in the respective pair shall be supplied to the respective speech store unit. In addition, it is also conceivable to carry out updating of a write address unit with up-to-date switching information after repairs or replacement, using here a write address feedback means CDU for the respective write address unit, the means CDU also being connected to the respective information selection means for the write address.

If the transmission of read addresses or write addresses between respective control store unit or write address unit and the respective information selection means for read addresses or write addresses is done via a link of a given type, it is conceivable that in connection with the respective information selection means for read addresses or write addresses to arrange fault supervision with the aid of fault control means.

It is of course not absolutely necessary to apply fault supervision to the input and output side of every unit along each transmission path according to Figures 1 and 6. Depending on the type of transmission links between the units and the implementation of the units fault control and fault control means may possibly be eliminated at some places, or in one or more units. For example, control of the respective write addresses and read addresses may be carried out only by the respective CHWA and CHRA in the SSU. If the information selection means. PS in the information processing unit DEVU is not associated with fault control means CH according to Figure 1 the information selction means PS should preferably also take the parity of the received signals into account when deciding which path and transmitted information to select.

Information transmission over the duplicated transmission paths does not necessarily need to take place in the form of words with the format according to Figure 2. The fault control result information and/or the fault detection coding can naturally be more sophisticated than having each consisting of a separate binary signal/digit.

Claims

C L A I M S

1 A method of fault supervision in transmission of digital meassage information (D) by a digital switching network, where at least partially duplicated Information transmission paths are established in the switching network and the message information is transmitted with fault supervision via each of the duplicated information transmission paths, where information transmitted via a transmission path is written into memory positions in a speech store (SS) for this transmission path, these memory positions being given by write addresses supplied to the speech store, where information written into the speech store for transmission via the transmission path is read out from memory positions given by read addresses supplied to the speech store, said fault supervision being implemented such that on detecting faults affecting the message information transmitted on a duplicated transmission path, the information transmitted on the transmission path is formed such that the affected message information is transmitted together with fault supervision result information (FIB) indicating that a fault affecting the message information in question has been detected, where a selection then takes place, with the aid of the fault supervision result information transmitted via the respective duplicated transmission path, between the information (ADO, AD1, ... AD7, ASO, AS1, AFIB, AP) transmitted via one of the duplicated transmission paths and the corresponding information (BDO, LBD1, ... BD7, BSO, BS1, BFIB, BP) transmitted via the other of the duplicated transmission paths which information shall be expedited further through the switching network, where in appropriate cases such information transmitted via a duplicated transmission path is selected as not being affected by faults, according to the fault supervision, characterized in that the read addresses (AO, Al, ... A7, FIB, P) are also subjected to fault supervision, and in that on detection of a fault in a read address supplied to a speech store, the digital information which is transmitted after reading out from the memory position given by the read address, which read address was faulty according to the fault supervision, is transmitted together with fault supervision result information (FIB) indicating that a fault affecting this information has been detected. 2 A method as claimed in claim 1, characterized in that pairs of mutually corresponding read addresses are read out from a control store in a control store unit (CSU) for the first transmission path and a control store in a control store unit (CSU) for the second transmission path, that the pairs of read addresses read out are subjected to fault supervision, in that a selection with the aid of the fault supervision is made for each of the duplicated transmission paths between corresponding read addresses in the respective pair, of which read address is to be supplied to the speech store (SS) for the respective transmission path, and in that in appropriate cases a fault-free read address, according to the fault supervision, of the two corresponding read addresses in the respective pair is selected and supplied to the speech store.

3 A method as claimed in claim 2, characterized in that if a read address read out from a control store (CS) is faulty according to the fault supervision, this faulty read address is replaced in this control store by the corresponding read address in the pair, provided that the corresponding read address in the pair is fault-free according to the fault supervision, by writing the selected fault-free read address into the control store in the memory position previously occupied by the faulty read address.

4 A digital switching network for transmission αf digital message information (D), the network having at least partially duplicated equipment (DDM-A, DDM-B) enabling establishment in the network of at least partially duplicated information transmission paths and transmission of message information over each of the duplicated paths, the network including a speech store unit (SSU) with a speech store (SS) and fault control means (CH) for the duplicated information transmission paths, the speech store unit for a transmission path being adapted, for transmission over the transmission path, to receive and store in the speech store digital information in memory positions given by write addresses supplied to the speech store unit, said speech store unit being further adapted, for transmission over the transmission path, tα read from the speech store information jwrittot in

memory positions given by read addresses supplied to the speech store, said fault control means (CH) for a transmission path being adapted such that, on detection of a fault affecting the message information (D) transmitted on the transmission path to form the information transmitted on the transmission path so that the affected message information is transmitted together with fault control result information (FIB) indicating that a fault affecting this information has been detected, said switching network including information selection means (PS) adapted to select between the Information (ADO, AD1, ... AD7, ASO, AS1, AFIB, AP) transmitted via the first of the duplicated transmission paths and the corresponding information (BDO, BD1, ... BD7, BSO, BS1, BFIB, BP) transmitted via the second duplicated transmission path, which information shall be further expedited through the switching network, said information selection means (PS) being adapted to select in appropriate cases such information transmitted via a duplicated transmission path which is not affected by faults, according to the fault control, characterized by address control means (CHRA) for controlling the read addresses (AD, Al, ... A7, FIB, P), and in that the fault control means are adapted to form, on detecting an error in a read address supplied to the speech store unit (SSU), to form the digital information transmitted after reading out from the memory position given by the incorrect read address so that affected message information (D) is transmitted together with fault control result information (FIB) indicating that a fault affecting this information has been detected.

5 A switching network as claimed in claim 4, characterized in that there is a control store unit (CSU) for the first transmission path and a control store unit (CSU) for the second transmission path adapted for reading out pairs of corresponding read addresses for the speech store unit (SSU) for the respective duplicated transmission path, in that address control means (CH) are adapted for controlling the pairs of read-out read addresses, in that there are read address selection means (PS) adapted, with the aid of the address control means, to select for each of the duplicated transmission paths between the read addresses in the respective pair what read address shall be supplied to the speech store for the respective transmission path, and In that in appropriate cases the read address selection means are adapted to select one of the addresses, fault-free according to the control, in the respective pair and supply the speech store with this address.

6 A switching network as claimed in claim 5, characterized by read address replacement means (CDU, SNC) for replacing in appropriate cases the incorrect read address in a pair, this address having been read from a control store, and replacing this address with the corresponding fault-free address in the pair in the respective memory position in this control store by writing into it the selected read address, said read address replacement means being connected to the read address selection means (PS) and the respective control store unit for supplying selected read addresses to the control store.

7 A digitial switching network for transmitting digital message information and having at least partially duplicated equipment enabling establishment in the network of at least partially duplicated information transmission paths and transmission of message information over each of the duplicated paths, the equipment comprising for each transmission path a speech store unit and a control store unit and fault control means, the speech store unit for a transmission path comprising a speech store and being adapted, for transmission over the transmission path, to receive and store in the speech store digital information in memory positions given by write addresses supplied to the speech store unit, the speech store unit for a transmission path being further adapted, for transmission over the transmission path, to read out from the speech store information in memory positions given by read addresses supplied to the speech store unit, the control store unit for a transmission path comprising a control store and being adapted to reading out read addresses for the speech store unit, the fault control means for a transmission path being connected to receive a succession of read addresses from the control store unit for one transmission path and a succession of corresponding read addresses from the control store unit for the other transmission path, the fault control means for a transmission path comprising address control means for controlling the received read-out read addresses, the fault control means for a transmission path comprising read address selection means for selection between each pair of corresponding read addresses what read address shall be supplied to the speech store unit for the transmission path, the read address selection means being adapted to select from each pair of corresponding read addresses a read address which is fault- free according to the address control means, the read address selection means for a transmission path being connected to supply the speech store unit for the transmission path with a succession of addresses selected from the pairs, the fault control means for a transmission path being adapted such that, on detection of a fault affecting the message information transmitted on the transmission path, to form the information transmitted via the transmission path so that the affected message information is transmitted together with fault control result information, which fault control result information Indicates that a fault affecting the message information in question has been detected, the network comprising information selection means connected to receive information transmitted via one of the duplicated transmission paths and the corresponding information transmitted via the other of the duplicated transmission paths, the information selection means being adapted to select, with the aid of the fault control result information transmitted via the respective transmission path, between the information transmitted via one of the duplicated transmission paths and the corresponding information transmitted via the other of the duplicated transmission paths, which information shall be further expedited through the switching network, the information selection means being adapted to select, if possible, such information transmitted via a duplicated transmission path which is not affected by faults according to the fault control result information.

8 A network as claimed in claim 7 comprising read address replacement means for each of the duplicated transmission paths, the read address replacement means for a transmission path being connected to the read address selection means for the transmission path as well as to the control store unit for the transmission path, the read address replacement means for a transmission path being adapted to supply, upon detection of a fault affecting the read address supplied to the address selection means for the transmission path from the control store unit for the transmission path, to the control store unit for the transmission path the selected fault-free read address for replacing the faulty read address in the control store unit for the transmisson path.