MXPA97007846A - Transmission of voice frequency signals in a movi telephone system - Google Patents

Abstract

According to the invention when the same MFTD signal is detected at the transmitting end, successive MFTD cycles are established related to this signal, these signs are modified DES groups and individual link information is added (Order, Start, Duration) to each cycle in addition to the information on MFTD signal frequency pairs, this link information indicates the duration of the MFTD signal from the first detection period of the first modified cycle of DES to the cycle involved, including said individual information The interlinking of successive cycles of MFTD (cycles 1-4) related to the same MFTD signal, at the receiving end, in the MFTD receiver cycles, calculates in each error-free cycle how much time has been received or should have been received. the MFTD signal, only when the individual link information of a received cycle indicates that the duration of the MFTD signal is longer than the m-value Minimum, for example 40 ms the signal of MFTD indicated by the information on pairs of frequencies of the received cycles is generated to a subscriber station, during the waiting period, noise is generated

Description

TRANSFER OF VOICE FREQUENCY SIGNALS IN A MOBILE TELEPHONE SYSTEM DESCRIPTIVE MEMORY The present invention relates to the transmission of oz frequency signals, especially MFTD signals, over the radio interface of a digital mobile telephone system. In a digital telephone system, a speech signal is encoded in some way before it is encoded by channel and transmitted to the radio path. In speech coding, digitized speech is processed cycle by cycle in periods of approximately 20 rns using different methods in such a way that the re-echoed is a group of parameters representing speech for each cycle. This information, ie the group of parameters, is coded per channel and transmitted to the transmission path. In coding by channel, the information is protected by different error correction codes. The speech coding method used in the GSM cellular system is the so-called ERP-PLP (Regular Pulse Excitation LPC with Long Term Prediction). The basic parts of the algorithm are the LPC filter of linear predictive coding and the coding of residual signals as the last stage of the pulse sequence. The operation is completed by LP estimation of sound tone. The encoder thus produces short-term filter parameters, long-term PLP prediction parameters and ERP parameters. In a decoder, the ERP parameters act as a filter Excitation signal and the short and long term received parameters act as filter parameters. The speech coding algorithm employed by the digital cellular system in the United States belongs to the code encoding category excited PLEC (Code Excited Linear Prediction) and the encoder is referred to by the + errn? No Excited Linear Predictive Coding of Vector Sum (PLESV). The result of the speech coding is a group of parameters, by means of which and also by means of the codebooks having a predetermined structure, a speech signal is synthesized in the decoder of the receiver. The speech signal residue is not transmitted at all, as it is done in an RPG-PLP encoder. The coders of both systems have in common the fact that the coder produces speech cycles whose duration is 20 s and that a speech cycle consists of subcycles of 5 s, each of which contains a group of speech parameters. In addition to the current coding, the following functions are also formed in digital speech processing: a) on the transmitter side, the voice activity + ection (DAV), by means of which the transmitter can be activated only when there is talk to transmit (Discontinuous Transmission, TD), b) on the transmitter side, valuation of background noise and generation of parameters corresponding to noise, and on the receiver side, generation of comfort noises in the decoder from the parameters , making this comfort noise an interruption in the connection sound more comfortable than absolute silence, and c) cancellation of the acoustic echo. As an example of speech processing, the speech processing arrangement that is used in the GSM mobile telephone system is described with reference to Figure 1 showing one side of the transmitter. The input of the speech coder 1 is either a PCB signal of 13 b ts arriving from the network, obtained by taking a sample of audio signal at a frequency of 8,000 samples per second, or 13-bit PCM converted to A / D that arrive from the audio part of the mobile station. The duration of the speech cycle obtained from the encoder output is 20 rns and comprises 260 bits. They are generated by coding 160 speech samples encoded by PCM. The speech coder 1 produces the parameters mentioned above for each speech cycle of 20 rns and the speech activity detector (DAV) 2 determines based on these parameters whether the cycle contains speech or not. According to the content of the cycle information, the DAV detector determines an appropriate flag that controls the operation of a TD control and the operation block 4. Its value can be DAV = 1, cycles applied to an encoder per channel 5 and from there forward to the path radio as the so-called traffic cycles are thus produced the speech cycles by the speech coder. TD control and operation block 4 determines an SP flag that controls the coding per channel for each cycle applied to the encoder of channel 5. In speech transmission, background noise is included in the speech, background noise that would be interrupted also in the use of discontinuous TD transmission, which would cause disturbing interruptions at the receiving end. Therefore, the cycles of DES (Silent Descriptor) containing noise parameters are transmitted after a profusion of speech and in certain intervals also during pauses of speech indicated by the DAV 2. Thus the receiver is capable of generating similar noise to the original noise of these parameters also during breaks. The duration of such a cycle and the number of bits in the cycle are the same as those of a speech cycle. The noise parameters are determined by a noise transmission function block 3 based on the parameters obtained by the speech coder 1. According to Figure 2 which shows the fields of a DES cycle, only a part of the 260 bits of the DES cycle are necessary in the coding of noise meters. The information on the background noise spectrum is encoded in field B and the level of background noise is encoded in field C. As for the other bits, 95 bits are used for the DES coding word, field A , and all the bits have the value of 0 in the word. The rest of the bits of the DES cycle have the value of 0, field I. When there is a pause in speech, ie the flag of DAV is 0, it causes the fact that the cycles transmitted from the control of TD and the operation block 4 to the encoder per channel and in addition to the radio path as the so-called traffic cycles are cycles of DES containing noise meters. The value of the SP flag adjusts the coding per channel to be suitable for these cycles. Figure 3 shows a known receiver arrangement used in the GSM mobile phone system. Channel decoding and detection are performed on the radio signal received in a block 35. The cycle of traffic detected with errors committed in the decoding per channel is provided with a MIC flag (Poor Cycle Indicator), which indicates whether the cycle of summary traffic is erroneous or error-free. As for the traffic cycle, it is verified in a block of detection of cycles of DES 36 if a cycle of DES containing noise information is in question. This is done by carrying out the coding word of the traffic cycle received bit by bit with the coding word stored in the receiver. Depending on how many bits are deviated from the correct place, a DES flag is provided for one of the three possible values. In addition, information on synchronization of traffic cycles is provided by means of a flag of J3AT (Flag of Time Alignment). The inputs of a TD control and an operation block 34 are thus the bits of information about the traffic cycles, erroneous information / 1 BER errors error concerning the cycle, and notification of whether the cycle is a cycle of DES that contains noise parameters. If the traffic cycle is an error-free speech cycle, it is applied to the input of a decoder * of speech 31, which generates the original speech based on the parameters. If the traffic cycle is classified as a speech or cycle of bad or lost DES based on the BFI flag, a certain procedure of replacing the bad speech cycles in a block 32 is carried out by applying, for example, the last good values of the parameters as attenuated to the speech decoder. If the traffic cycle is an error-free DES cycle, is applied to a noise RX function block 33, which adjusts the speech decoder 31 to produce noise similar to the original noise for as long as the speech cycles are received again. A basic characteristic of digital networks is that they do not pass through signals like a telephone network. They do not pass through MFTD signals properly, do not say a signal from the V.29 rnodern used in the fax machines. In a telephone network, the signaling of MFTD (Multi-Frequency Dual Tone), contrary to dial-up uses, penetrates the entire connection entirely to subscriber B and this is why they are especially useful for use in the remote control apparatus, for example in remote interrogation telephone answering machines or in coded voice data transmission. In the MFTD signaling, the simultaneous voice frequencies are used to indicate a specific * character. All digits 0/9 and the characters * and # are indicated as a combination of two different frequencies selected from the frequencies 697 Hz, 770 Hz, 852 Hz, 941 H, 1209 Hz, 1336 Hz, and 1477 Hz. twelve combinations allowed. Using the frequency of 1633 Hz, the literal symbols A, B, C and D are also obtained. The number of allowed combinations of frequencies is thus 16. For telefax machines, a special adaptation function is specified in the networks of GSM, adapting the analog signal of a machine to a digital radio channel by means of this function. The transmission of MFTD signals from a mobile station to the network, ie in the uplink direction, has also been specified. According to the specification, the MFTD voices are not generated by the mobile station but by a mobile station, whereby the voice signals do not have to be applied through a speech coder. By pressing number buttons of the station during the speech connection, the mobile station transmits a message and the mobile station generates the message after having obtained a corresponding MFTD signal.

The problem with the current networks is thus the transmission of MFTD signals in the downlink direction. This is not specified in any way in the current mobile networks. It is true that the MFTD signals that move from the network to a mobile station reach the mobile station, but in a distorted way, since they have to travel, on the network side, through a speech coder and then through a speech decoder in the mobile station. Due to the distortion, they do not meet the conditions established by the MFTD detectors of a fixed network to detect MFTD signals. The transmission of the signals in the uplink direction also presents problems despite the aforesaid specification: when the user uses the MFTD installation of the mobile station, the station transmits both the initial message and the final message of a MFTD signal, the central office confirms both messages with confirmation messages. Therefore, the transmission of a number comprising, for example, 10 characters requires a total of 40 messages, this loads the network. The problem is exacerbated especially in telephone systems in which a fixed connection between a central station and the subscriber stations in a fixed telephone network with a radio connection is replaced. The solution is referred to as a telephone system that applies a wireless subscriber connection, that is, the CLI system (System of Local Wireless Circuits). In the CLI system, a fixed wireless terminal equipment comprises a radio unit provided with an antenna and a telephone adapter, which connects a normal subscrip- tor station to the terminal equipment. The subscriber station may be a conventional telephone apparatus to which a telephone answering machine is connected. The user uses the s? Bscpptora station in the same manner as a conventional fixed network, even if the subscriber line connection consists of a radio connection between the terminal equipment and a base station. The base station is connected to a special subscriber network element, which is connected to a normal telephone exchange. The CLI system can be constructed by applying, for example, the GSM digital system components. The signaling of the CLI system is in accordance with the related system. In the CLI system the transmission of MFTD signals from the network over a radio path to a subscriber station would be extremely desirable. A proposed solution to solve the problems presented is described in the European patent application 534 852. According to it, an MFTD detector and an MFTD encoder are provided at the transmission end, in the transcoder of a base station, in addition to a speech coder. The minimum period established for the detector to detect an MFTD signal is short, only 5 rns. When the detector detects an MFTD signal coming from LO the network gives a control signal to the MFTD encoder associated with the speech encoder and the transmitter. The MFTD encoder thus establishes a cycle, which resembles a DES cycle and contains information about the detected MFTD. The transmitter, as controlled by a controller, selects this MFTD cycle similar to a DES cycle instead of a speech cycle. The fields of such MFTD cycle are shown in Figure 4. The first three fields A, B, and C correspond to the fields of the DES cycle of Figure 2, comprising the A field 95 bits thus containing the cycle identifier of DES, containing field B information on the quality of background noises and field C on the level of background noises. Being distinct from the cycle of Figure 2, a cycle of MFTD according to the European Patent Application contains the additional fields D, E and F. The D field contains a cycle identifier of MFTD, which comprises 8 bits each being in state 1. Field E contains a frequency pair code of MFTD, which comprises 4 bits, whereby there may be 16 pairs of frequencies. The 4-bit F field indicates the duration of the MFTD voices as multiples of 5 ns. At the receiving end, the MFTD cycle is identified by means of a DES cycle identifier (Field A) and the MFTD cycle identifier (Field D). The "codes" parameter indicated in the E field defines the MFTD frequency pair in question and the "duration" parameter indicated in the F field indicates which periods of the 20 rns cycle divided into periods (subcycles) of 5 rns contain the signal of MFTD. At the reception, the MFTD signal according to the code of the E field is generated during those periods of 5 rns which contain the MFTD signal according to the duration of the MFTD signal. For other periods of the cycle, the background noise defined by the DES parameters is generated. The disadvantage of this known solution is that they are not taken into account errors caused by the radio path. If erroneous MFTD cycles are occasionally received at the receiving end, the regeneration of an MFTD signal, performed at the receiver, can become problematic, since the successive cycles are not interlaced in any way. There is no way to know for sure how long the MFTD signal has been received or received and a new MFTD signal has already been started. The code of an MFTD signal may be the same in successive MFTD cycles even if two separate MFTD signals were involved. According to CEPT recommendation T / CS 46-02, the conditions of a reliable detection of MFTD are that an MFTD signal lasts for 40 ms and that it is preceded by a state, which lasts for more than 40 rns and does not contain any voice frequency signal, oq? e is a state of detection of a different voice frequency signal. Since the detector in the transmitter end transcoder uses at least 5 rns for the detection of MFTD signals, there is no time to detect the signal if it starts during the last period of 5 rns of the cycle. Accordingly, the MFTD signal may occur at the end of the cycle for a period of less than 5 rns, this signal traveling through the coding-decoding string of the speech and being distorted. Immediately afterwards a pure voice frequency signal occurs based on the cycles received from MFTD. If a pause of 40 rns is not maintained between the distorted MFTD signal and the pure MFTD signal, the detection of the MP TD signal at the subscriber end may fail completely. Also, in the solution according to the prior art, it is not verified at any stage if the MFTD signal arrives from the network to the transmitting end to last 40 ms. In the transcoder of the transmitting end, it is thus possible to detect for example a MFTD signal which lasts less than 20 ms and transmit it further on the radio path in full to a subscriber station. In the subscriber station, the voice with a MFTD signal is not recognized, which is why it is uncomfortably audible in speech. The object of the present invention is such a method for transmitting MFTD signals over a digital radio path that exhibits no disadvantages associated with the arrangements according to the prior art and that is appropriate for transmitting voice frequency signals reliably in the uplink direction. as descending The object defined with the method described in claim 1 is achieved. According to the method of the invention, upon detection of the same MFTD signal at the transmitting end, the successive MFTD cycles related to this signal are established, these cycles being modified of DES, and individual link information is added to each cycle in addition to information on frequency pairs of said MFTD signal, this individual link information indicating the duration of the MFTD signal of the first detection time period of the first modified cycle of DES to the cycle involved, including said cycle. The individual link information, which is a group of parameters, intertwines successive MFTD cycles related to the same MFTD signal. At the receiving end, upon receiving the MFTD cycles, it is calculated in each error-free cycle how long the MFTD signal has been received or received. Only for the individual link information of a received cycle indicates that the duration of the MFTD signal is longer than the minimum value, for example 40 ms, the MFTD signal indicated by the information on frequency pairs of the received cycles is generated to a subscriber station. During the waiting period, background noise is generated. Next, the invention will be described in more detail by means of a preferred embodiment of the invention with reference to the accompanying drawings, in which Figure 1 shows speech processing at the transmitting end in a known system, Figure 2 shows A known pattern of a DES cycle, Figure 3 shows speech processing at the receiving end in a known system, Figure 4 shows an DES cycle adapted to transmit the MFTD information, Figure 5 shows the processing of the speech at the transmitting end using the circuit arrangement of the invention, FIG. 6 shows a DES cycle adapted in accordance with the invention for transmitting MFTD information. The conditions of the detection of MFTD is that the MFTD signal lasts for more than 40 rns and q? E is preceded either by a state that lasts more than 40 ms, that does not contain a voice frequency signal, or by a detection status of a different voice frequency signal. In Figure 5, in which the same reference numbers are used as in Figure 1 when applicable, an incoming audio signal is monitored by a controller 51, which contains an MFTD detector. When the detector detects an MFTD signal, it identifies the pair of frequencies in question at the same time. The controller 51 notifies the TD control of the operation block 4 by means of an MFTD flag that an MFTD signal is arriving. In response to the flag, block 4 establishes an SP flag to such a state that the coding block by channel 5 knows that the traffic transmitted cycles are cycles adapted from DES that contain MFTD information and then the b Loque 5 does not perform coding per channel in the same way as a pure DES cycle. These types of cycles will be referred to later as MFTD cycles. Based on the DAV and MFTD flags, TD control and operation block 4 must be able to apply correct bits of information to the encoder per channel 5 and establish that the SP flag is such that the MTFD cycles and not the DES cycles are transmitted as traffic cycles during a speech pause. Block 4 must thus perform the logical deduction according to the chart in Figure 6. In the example, the value of 0 of the DAV flag indicates that the voice activity detector 4 has detected a pause and the value of 1 of the flag indicate that the detector 4 a detected speaks. The value of 1 of the MFTD flag has detected an MFTD signal and correspondingly, the value of 0 indicates that the mentioned voice frequency signal has not been detected. The value of 0 of the SP flag indicates that the traffic transmitted cycle is a DES cycle. Such arrangements are applied in block 4 that setting the flags in the manner shown in Table 6 reduces the situation that the DES cycles are transmitted only when no speech or MFTD signal is detected and cycles are not transmitted. of MFTD without considering the DAV flag whenever a MFTD signal is detected.

When a detection section of the controller 51 has detected the MFTD signal, the controller notifies yet an MFTD encoder 52 of the frequency pair, q? E is a notification of the MFTD character and the duration of the MFTD signal. If the signal ends too early, for example in less than 40 rns, the controller 51 gives the MFTD encoder an order to transmit a DES cycle without MFTD information. The notification of the MFTD detected in cycles similar to a DES cycle, which are referred to as MFTD cycles, is transmitted. The form of a MFTD cycle is shown in Figure 7. The content of the information of the AE is essentially the same as in the cycle according to the prior art shown in Figure 4. In this way, in addition to the identifier of the DES cycle included in the A field and the background noise parameters indicated in fields B and C, the cycle contains an identifier of MFTD cycles (13 b ts) in the D field and? n MFTD signal code? (4 bits) in the E field, by means of which the cycle is identified as a MFTD cycle and which MFTD signal is detected in question. According to the invention, individual link information is added to a MFTD cycle, this link information indicating, in the manner described above, the duration of the MFTD signal of the first detection time of the first MFTD cycle to the cycle involved. , including said cycle. The individual link information, which is a group of parameters, thus interleaves successive MFTD cycles related to the same MFTD signal. The fields F, G and H are defined in one cycle. The link parameters are arranged in the same way: the "Order" parameter (3 bits) in the F field, the "Start" parameter (2 bits) in the G field and the "Duration" parameter (2 bits) in the H. Field The "Order" parameter determines which of the cycles related to the same MFTD signal is in question. The "Start" parameter determines the start time of the MFTD signal, ie how many periods of 5 ins the first cycle belonging to the same MFTD signal is included. The parameter "Duration" determines correspondingly, with an accuracy of 5 rns, that is so long contains the MFTD voice of the cycle involved. All 24 bits related to the MFTD signal, ie the bits of the D-H fields are arranged between the class bits, which have the best protection against errors. The rest of the bits of the cycle are set to zero. The link is illustrated in Figure 8, which shows the values of the parameters established by the MFTD encoder 52 in a MFTD cycle when the MFTD detector has detected the MFTD signal that appears in the Figure. During the first cycle, the MFTD detector, the detection frequency which is 5 ms, has been detected in the third subcycle that a MFTD signal is arriving. The duration of the signal extends to the third subcycle of the fourth cycle. The subclasses containing the MFTD signal are indicated by oblique lines and the subcycles containing background speech / noise are not indicated or oblique lines. Each cycle is established as one MFTD cycle and consecutively numbered by the "Order" parameter. The "Start" parameter provided in each cycle is the same, in this case 2, which indicates that the first cycle includes 2 subcycles that contains the MFTD signal, that is, the duration of the signal in the first cycle is LO rns. The "Duration" parameter indicates the number of MFTD subcycles in each MFTD cycle. In the second and third cycle the "Duration" -4, thus reaching the MFTD signal during the entire cycle. In the fourth cycle, the "Duration" = 3, thus this cycle contains 15 rns of the MFDT signal. The sum of the values of "Duration" is 13, thus the total duration of the MFTD signal of 13 * 5 rns = 75 ms. In this way, the MFTD cycles related to the same MFTD signal are interleaved by means of the "Order", "Start" and "Duration" parameters contained by the MFTD cycles. By means of these parameters, it is then possible to know in any error-free cycle how long the MFTD signal should have been received at a time if erroneous MFTD cycles were occasionally received. The length of the MFTD signal in a specific cycle is then unequivocally: TDTMF = Start "5 rns + (Order 2)" 20 rns + Duration "5 ms, if the 0rden >; 2, and TDTMF = Start "5 ms = Duration" 5 ms, if the Order < 2. If the MFTD detector 51, Figure 5, notices at the transmitting end that the MFTD signal is interrupted before it has lasted for more than 40 rns, the MFTD encoder 52 stops the establishment of the MFTD cycles and stops its transmission. Next, the MFTD detector only reacts to such a new MFTD signal that only starts when a pause of 40 rns, which does not contain MFTD, pauses after the interrupted signal. Due to this procedure, the contents of the MFTD cycles are unambiguous at the receiving end. As soon as at least one error-free MFTD cycle is received, it is known that all subsequent cycles, whether erroneous or not, contain the MFTD signal for as long as it has not already been received for more than 40 years. rns. Figure 9 shows the speech processing arrangement of the receiving end. A detector section 91 of MFTD cycles must be incorporated into the processing block of the received cycles. If the detector section 36 of the DES cycles is noted based on an identification word of the descending cycle cycles that is arriving is a DES cycle, the MFTD cycle detector section 91 examines whether the cycle also contains a cycle. MFTD identifier, block D Figure 7. 2 bits can be distributed for the flag that indicates it. If the MFTD flag is 0, an MFTD cycle has not been received, thus progressing the speech processing according to the prior art, ie the speech coder 31 generates background noise according to the noise parameters of the MFTD. DES cycle that is arriving at the function block 33 of noise RX. If the MFTD flag is 1 or 2, an MFTD cycle has been received. In this way, the control of TD and the operation block 34 transmits the MFTD cycle and 2 control bits of MFTD to a controller 92. The controller calculates from the cycle parameters by when the same time signal is received. MFTD and controls a multiplexer 93 that acts accordingly as a switch. The controller waits first 40 rns before connecting the MFTD voice generated by an MFTD 94 decoder to the output. The time supervision ensures the bed from which a pause of more than 40 rns is obtained between a distorted signal and a pure MFTD signal, which is a reliable detection condition of MFTD signals according to the CEPT recommendation. When it is expected that a specific MFTD signal will be received for more than 40 rns before starting to generate it at the output and at the line, this ensures at the same time that it signals the duration which is less than 40 rns, does not reach the subscribing station. During the waiting period, the controller uses the MUX 93 to multiplex the noise signal produced by the speech decoder 31 as an output. The speech decoder has received an order from the controller 92 to generate noise by means of the welfare noise parameters. 2L Reference is still made to Figure 5 showing the transmitting end. The welfare noise parameters, included in the MFTD cycles, must have been generated at some more anticipated point. The previous acceptable parameters of welfare noise- must be stored in the memory, for example in the MFTD 52 encoder. If this is not done, it may happen that welfare noise parameters are generated during a MFTD signal, which does not It is desirable, of course, since it would sound unpleasant at the receiving end. Such a procedure can also be followed so that the welfare noise parameters are not generated during an MFTD signal, since the DAV 4 detector does not interpret an MFTD signal as silence. Also in this case, usable parameters of welfare noise must be stored somewhere in the memory. According to a preferred embodiment, it is possible to follow such a procedure that if an MFTD signal lasts more than 40 rns at the transmitting end, the last MFTD cycle can be transmitted again twice more after the signal has ended, whereas the Two subsequent periods of 20 rns do not contain a new MFTD signal. This does not cause any additional loss on the speech information, since at least two speech / noise cycles are lost at the receiving end in any case after the generation of the MFTD signal is completed, when the delay is removed the volatile memory. 0 0 When the MFTD signal has been received at the receiver end of 40 rns, it is necessary to check the received MFTD cycles below to find out if they belong to the same MFTD signal or to the next one. The verification can be carried out by means of the parameters "Order" and "Start" of the invention. If in the same MFTD signal it is still in question, the "Start" parameter may have the same value as in the previous received cycle and the "Order" parameter must be greater than or equal to the previous received cycle. The equality condition is possible if the last MFTD cycle has been retransmitted twice more at the transmitting end after the MFTD signal has ended. The disclosed invention is suitable for use in the transmission of MFTD signals over the radio path in a number of addresses in a digital mobile telephone network and especially in the CLI wireless subscriber line system utilizing the characteristic of such a network. The foregoing description and figures related thereto have the primary purpose of illustrating the present invention. Different variations and modifications of the invention will be apparent to one skilled in the art, deviating from the scope and spirit of the invention described in the appended claims.

Claims (13)

y. NOVELTY OF THE INVENTION CLAIMS

1. - A method for transmitting MFTD signals over a radio path in a digital mobile telephone system, in which the following steps are taken at the transmitting end: establishing, from an incoming signal, speech cycles containing parameters of the speaks during active cycles of sound and speech noises that contain background noise parameters during pauses; detect, in periods of detection, an MFTD signal possibly included in the incoming signal and establish modified welfare noise cycles, each of which contains both the identifier and the information on MFTD frequency pairs in addition to the noise information of welfare, further characterized by adding individual link information to each modified cycle of welfare noise related to the same MFTD signal, indicating to said individual link information the duration of the MFTD signal from the first detection period of first modified cycle of welfare noise to the cycle involved, including said cycle, transmitting the established modified cycloe of welfare noise as an encoded channel to the radio path.

2. - A method according to claim 1, further characterized in that the individual link information is a group of par meters and because each link parameter is located in a dedicated field in a modified welfare noise cycle.

3. A method according to claim 2, further characterized in that the first link parameter "Order" indicates which of the cycles of the successive modified cycles of welfare noises related to the same MFTD signal is in question.

4. A method according to claim 2, further characterized in that the second link parameter "Start" indicates how many detection periods contains the first modified cycle of welfare noises related to the same MFTD signal.

5. A method according to claim 2, further characterized in that the third link parameter "Duration" indicates how many detection periods the cycle involved contains.

6. A method according to claim 1, further characterized in that if the duration of a detected MFTD signal is shorter than the first established reference period-preferably 40 rns-the establishment of the modified cycles of noises of well-being and its trancenisión to the trajectory of radio.

7. A method according to claim 1, further characterized in that the modified cycles of welfare noises related to the new MFTD signal are established and transmitted to the radio path only when a second established reference period has elapsed. preferably 40 nm - from the termination of the previous MFTD signal.

8. A method for transmitting MFTD signals over a radio path in a digital mobile telephony system, further characterized by the following steps being taken at the receiving end: verifying whether a cycle received from welfare noises contains an identifier of MFTD and classify the welfare noise cycle containing said identifier as a modified cycle of welfare noises; detecting information on MFTD frequency pairs in the modified welfare noise cycle and generating a voice frequency signal of MFTD q? e corresponds to said information of frequency pairs, further characterized because it is analyzed in each modified cycle received from welfare noises the individual link information including the cycle, indicating said individual link information of the transmitted signal of MFTD from the first modified cycle received from welfare noises to the cycle involved, including said cycle; generating a voice frequency signal of MFTD only when the individual analyzed link information indicates that said duration exceeds a set minimum value, preferably 40 rns.

9. - A method according to claim 8, further characterized in that the individual link information is a group of link parameters, in which each link parameter is located in a dedicated field in a modified cycle of welfare noises.

10. A method in accordance with the claim 9, further characterized in that in each cycle received modified: the first link parameter "Order" indicates which of the cycles of successive modified cycles received from welfare noises related to the signal with the same MFTD signal is in question; the second link parameter "Start" indicates how many detection periods the first modified cycle of welfare noises contains at the same MFTD signal; the third link parameter "Duration" indicates how many detection periods the cycle involved contains.

11. A method in accordance with the claim 10, further characterized in that the length TDTMF of a signal received from MFTD is as follows a modified cycle of welfare noises: TDTMF =? N? C? O-5 rns + (order-2) -20 ms + duration -5 rns, if the Order > 2 and TDTMF = start -5 rne = duration- 5 rne, if the Order < 2, where 5 ms is a detection period of the DMFTD detector at the transmitting end and 20 rns is the duration of a modified welfare noise cycle.

12. A method according to claim 8, further characterized in that when the analyzed link information indicates that said duration is shorter than the minimum value established, background noise is generated.

13. A method according to claim 8, further characterized in that when the individual link information q? E is analyzed from a received modified cycle it indicates that said duration exceeds the minimum established value, an MFTD signal is generated at Despite possible erroneous modified cycles that are occasionally received.