EP0961264B1 - Emitting/receiving device for the selection of a source coder and methods used therein - Google Patents

Description

• des moyens de codage de source comprenant une pluralité de codeurs de source pour compresser ledit signal de parole et délivrer un signal compressé à un débit de sortie donné, ledit signal compressé ayant une distorsion mesurable et
• des moyens de sélection d'un codeur de source parmi la pluralité de codeurs.

The invention as claimed relates to an apparatus for transmitting / receiving digital signals comprising a transmitting part and a receiving part, the transmitting part receiving as input a speech signal, called the original signal, and comprising:

• source encoding means comprising a plurality of source encoders for compressing said speech signal and outputting a compressed signal at a given output rate, said compressed signal having a measurable distortion and
• means for selecting a source coder from the plurality of coders.

It also relates to telephone equipment and a communication system. digital telecommunications.

• une étape de réception d'un signal de parole, dit signal original,
• une étape de codage de source pour compresser ledit signal de parole et délivrer un signal compressé à un débit de sortie donné, ledit signal compressé ayant une distorsion mesurable et
• une étape de sélection d'un codeur de source parmi une pluralité de codeurs pour réaliser l'étape de codage de source.

The invention also relates to a method for transmitting digital signals comprising the following steps:

• a step of receiving a speech signal, called the original signal,
• a source coding step for compressing said speech signal and delivering a compressed signal at a given output rate, said compressed signal having a measurable distortion and
• a step of selecting a source coder from a plurality of coders to carry out the source coding step.

Finally, it relates to a method for receiving digital signals comprising a source decoding step.

The invention applies in particular to any cellular terminal operating according to a digital telecommunications standard of GSM type (from the English Global System for Mobile communications), PCS1900 (Personal Communications System), PHS (Personal Handyphone System), TDMA (Time Division Multiple Access), CDMA (Code Division Multiple Access), WBCDMA (Wide Band CDMA), UMTS (Universal Mobile Telecommunications System), etc.

The Japanese abstract published under the number 08237711 A describes a device transmission / reception of the kind mentioned in the introductory paragraph, to increase the quality of an audio signal to be transmitted. For this, the device includes means for test directly on the useful signal, the performance of source encoder families successively selected from a list of available families and means for compare each output signal to the original signal in order to transmit only the signal coded closest to the original signal. EPO 417,739 also discloses such a device.

The invention proposes a more economical method for optimizing the compromise quality of transmission of the speech signal / capacity of the communication medium, within digital telecommunications apparatus involving source coding.

• des moyens de calcul pour effectuer une estimation de la distorsion du signal compressé,
• des moyens de contrôle pour comparer cette estimation à des valeurs de consigne et sélectionner un codeur de source en fonction du résultat de ladite comparaison.

For this, there is provided an apparatus of the kind mentioned in the introductory paragraph remarkable in that said selection means include:

• calculation means for estimating the distortion of the compressed signal,
• control means for comparing this estimate with reference values and selecting a source encoder according to the result of said comparison.

Thus, each original signal undergoes only one source coding test before the selection of the optimum encoder, while remaining in the same family of encoders.

According to an important characteristic of the invention, the reception part comprises a plurality of decoders compatible with said source coders and means for command cooperating with said control means to automatically select, among the plurality of decoders, a decoder compatible with the source encoder selected by said control means.

According to a particular embodiment of the invention, the source coder delivers an internal residual error signal and said calculation means use said error signal to estimate said distortion.

• des moyens de décodage de source inverse coopérant avec les moyens de codage de source pour fournir un signal décodé à partir dudit signal compressé,
• des moyens de comparaison du signal décodé au signal original pour fournir un signal d'erreur résiduelle et
• des moyens de traitement du signal d'erreur résiduelle pour en déduire ladite estimation.

According to another embodiment, the calculation means comprise:

• reverse source decoding means cooperating with the source coding means to supply a decoded signal from said compressed signal,
• means for comparing the decoded signal with the original signal to provide a residual error signal and
• means for processing the residual error signal to deduce said estimate.

• une sous-étape de calcul pour effectuer une estimation de la distorsion du signal compressé,
• une sous-étape de contrôle pour comparer cette estimation à des valeurs de consigne et sélectionner un codeur de source en fonction du résultat de ladite comparaison.

There is provided an emission method of the kind mentioned in the introductory paragraph, remarkable in that said selection step comprises:

• a calculation sub-step for estimating the distortion of the compressed signal,
• a control sub-step for comparing this estimate with set values and selecting a source encoder according to the result of said comparison.

The invention finally provides a reception method of the kind mentioned in the introductory paragraph for decoding a signal compressed by a transmission method of the type above, remarkable in that it includes a step of automatic selection of a decoder from among a plurality of decoders available as a function of said source coder selected.

The following description, made with reference to the attached drawings, all given to by way of nonlimiting example will make it clear how the invention can be implemented.

Figure 1 is a block diagram of a conventional digital transmitter.

FIG. 2 is a block diagram of the transmission part of a transmission / reception apparatus according to the invention.

FIG. 3 illustrates a first embodiment of the transmission part shown schematically in Figure 2.

FIG. 4 illustrates a second embodiment of the transmission part shown schematically in Figure 2.

FIG. 5 is a flow diagram illustrating a transmission method according to the invention.

Figure 6 shows an example of a digital telecommunications system according to the invention.

The general structure of a digital transmitter for a GSM type system, that is to say without spectrum spreading, is given in FIG. 1. It includes a source 10, a source encoder 11 (COD), a channel encoder 12, a multiplexer 13, a modulator 14 (MOD), a device 15 for transposing the signal to the radio frequency fr and an antenna 16.

The signal S emitted by the source 10 is either analog, such as for example the speech, and must therefore be digitized, either directly digital like the signals of signaling. We are interested here in the speech signal, the quality of which we seek to improve. program. This signal undergoes coding, called source coding, by the source coder 11 for minimize the amount of information to be sent.

There are many methods of source coding such as modulation coded pulse (MIC) or analysis analysis coding. The first performs a 8 kHz sampling (slightly more than twice the highest frequency present, depending on the Nyquist criterion) of the band called "telephone band" located between 300 and 3400 Hz. Each sample is coded on eight bits. We obtain a bit rate of 8 * 8 = 64kbits. The second method currently used in the GSM standard uses a production model of the speech using a Linear Predictive Coding (LPC) analysis of the speech signal word. By using this type of encoder and these variants, data rates are obtained which are much lower than those obtained by MIC coding, such as 13 kbps for the full GSM coder debit.

The signal thus coded is processed by a second coder 12, called the channel coder, whose purpose is to add redundancy to the sequence of symbols to be transmitted, to reduce the risk of transmission errors. The coded data is then brought into form and multiplexed between the different available logical channels, by the multiplexer 13, depending on the type of multiple access used for transmission. For example, for a time multiplexing system, a physical channel occupies only one time slot limited. It is therefore necessary to share the information blocks coded in sub-blocks to be inserted in the time slots on the radio channel.

Once multiplexing has been carried out, the symbol sequence is modulated by the modulator 14. This operation consists in transforming the sequence of numerical symbols into signal to be transmitted through the channel. The signal is then transformed into a waveform according to the type of modulation chosen. The modulation produces an overflow on the neighboring channels to be filtered. The appropriate device 15 then transposes the signal to the radio frequency fr, i.e. the carrier frequency of the channel, before it is transmitted on radio waves through the antenna 16.

Since the receiver is also of generally conventional shape, a person skilled in the art can easily deduce its structure from that of the transmitter which has just been described.

A transmission device according to the invention is illustrated in FIG. 2. The blocks common in Figure 1 have the same references. A calculation block 21 as well as a block 23 control have been added to the conventional broadcast chain. The calculation block 21 includes measurement means for estimating the distortion of the signal source encoder output 11. The source encoding block 11 includes several encoders likely to be selected by the control block 23. For this, the control block 23 compare the estimate of the distortion to threshold values stored in a table in device memory and selects a source encoder from the encoders available in the source coding block 11, as a function of the result of the comparison.

The different encoders can for example be referenced in order of precision increasing, that is to say in order of increasing output flow. In this case, when a threshold of high reference (respectively low) is reached for the estimation of the distortion, the block of control automatically selects the next higher available encoder (respectively lower) in terms of precision.

The control block 23 must then make a request to the network to ask it to change the encoder and must wait for its agreement before actually selecting the new coder better suited. In case of network refusal, the old encoder is kept. In Indeed, it is imperative that the two communicating parties, here the radiotelephone and the network, use compatible encoders and decoders.

Some networks transmit digital messages from mobile to mobile without decode the information. In this case, the receivers of mobile devices must use a decoder capable of decoding the messages transmitted by the transmitters of their correspondents. For this, the invention provides that the selection of a source encoder in the transmitter of the transmitting / receiving device automatically triggers the selection a compatible decoder in the receiver.

FIG. 3 illustrates a first embodiment of the invention for estimating the distortion of the signal at the output of the source encoder 11. According to this embodiment, performs a subtraction 31 between the original speech signal S and the signal coded by the source encoder 11 then decoded by a reverse decoder 33 to obtain an error signal e representing the error between the transmitted signal and the coded signal. This error e is then filtered by a perceptual filter 34 and its energy is calculated by an energy calculation block 35. We obtains at the output of the calculation device 31 + 33 + 34 + 35 an estimate E of the distortion due to the encoder 11 which will be processed by the control block 23.

FIG. 4 illustrates a second embodiment of the invention for estimating the signal distortion at the source encoder output. According to this embodiment, the encoder from source 11 has an internal ER residual error which can be accessed to get information. This residual error is then filtered by a perceptual filter 41 (this filter is already present in the emission chain of radiotelephones conforming to the standard GSM EFR 06.60). The energy of the filtered error is then calculated by a calculation device of energy 43 to provide the control unit 23 with an estimate E of the signal distortion due to source encoder 11.

A method according to the invention is illustrated in FIG. 5. It includes steps K0 to K7. Step K0 represents the reception of the speech signal S by the source coder. The signal S then undergoes a double coding step K1: a source coding intended for compress the signal to be transmitted and a channel coding intended to protect it against transmission errors. In step K2, an estimate E is made of the distortion provided by the speech coder to the original signal. This estimate E can be made, by example, according to one of the methods described in relation to FIGS. 3 and 4. The method is continues with a control step which includes a sub-step K3 for comparing the estimated distortion E at setpoints, followed by a decision substep K4 for choose a coder according to this estimate, among the coders present in the transmitter. The choice of the source encoder being made, the control block must validate this choice by the radiotelephony network. Step K5 consists in making a request to the network to ask it to replace the old encoder with the selected encoder and wait for the network response. If the answer is positive (K6), the coder beforehand selected replaces the old one in the transmitter, otherwise (K7), the old encoder is kept.

This process is preferably carried out once per data frame. But if the network or the correspondent's mobile receiver does not allow the choice of decoder (for example example it has only one decoder or else the bit rate is unacceptable), the process will only be done once per connection or per communication.

Figure 6 shows a cellular radio system, for example GSM type. However, the invention can be implemented in all systems of digital communications for which an audio signal source coding is made.

The system shown in Figure 6 includes two base radio stations 60, 61 connected to the GSM 62 network, as well as two mobile phones 63 and 64 capable of communicate with each other by radio within a coverage area 65, representing the radio coverage area of the GSM network, through base stations 60, 61 and of the network 62. The stations 60, 61 provide the radio interface between the GSM 62 network and the mobiles 63, 64.

The current GSM standard provides for network decoding of coded messages and transmitted by the transmitting mobiles, before transmitting them to the receiving mobiles. The source coding means located in mobiles must therefore be compatible with decoding means used by the network.

According to a particular embodiment, corresponding in particular to the standard Current GSM, the transmitting part of mobile 63, after having selected a source coder as previously described, must send a request to the network 60, 61, 62 to ask him to adapt his decoding means to the source encoder which has just been selected on transmission. Upon receipt of a confirmation message, the sending party of the mobile 63 can then effectively change the encoder.

However, if the network does not decode the messages coded by mobiles but would simply transmit them, there is another mode of production.

We distinguish in a connection between two mobiles, a mobile transmitter of a speech message, for example mobile 63 and a receiving mobile, for example mobile 64. According to the invention, the mobile 63 selects a source encoder, but instead of ask the network for authorization to use this source encoder, it sends a request to mobile receiver 64 in the form of a signaling message, via the network. Because if the network does not decode the coded messages in transmission, the task falls to the mobile receiver 64. For this, the receiving part of the receiving mobile 64 must be provided with means for selecting a decoder compatible with the source coder used in transmission. These means comprise a plurality of decoders compatible with the coders of source of the transmission part as well as control means for selecting automatically, at the request of the sending mobile, a compatible decoder among the plurality of decoders.

It is also planned to make the communication symmetrical between the two that each one uses the same encoder and the same decoder, but it is not mandatory. Indeed, the communication can use a type of coding / decoding in a sense and another in the other direction. In order to standardize the system, the invention provides that the reception part of a mobile automatically selects a decoder compatible with the encoder previously selected for transmission. For this, the reception part includes a plurality of decoders compatible with said source coders and means for command cooperating with the control means of the transmission part to select automatically, among the plurality of decoders, a decoder compatible with the coder source selected during transmission.

Claims (11)

1. A digital signal transmitting device for transmitting compressed output signals from input speech signals, characterized in that it comprises:

   source coding means including a plurality of source coders with variable output rates and of which a single coder, called current coder, from the plurality of coders is used both for compressing the input speech signal and for delivering a compressed output signal at a given output rate, the compressed output signal having a distortion relative to the input speech signal,

   calculation means for calculating an estimate of said distortion of the compressed output signal with the aid of the current coder,

   check means for comparing said estimate to reference values and for selecting a source coder, called new current coder, from the plurality of coders so as to deliver the compressed output signal.
2. A device as claimed in claim 1, in which the source coders deliver an internal residual error signal which is used by the calculation means for calculating said distortion estimate.
3. A device as claimed in claim 1, in which the calculation means comprise:

   inverse source decoding means co-operating with the source coding means to produce a decoded signal from the compressed output signal,

   means for comparing the decoded signal with the input speech signal to produce a residual error signal, and

   means for processing the residual error to derive said estimate therefrom.
4. A device as claimed in claim 1, in which the check means comprise a memory for storing said reference values and in which each reference value is associated with a source coder.
5. A device as claimed in claim 1, in which the source coding means supply a unique residual error signal to the calculation means, which signal corresponds to the distortion of the compressed output signal delivered by a unique current coder from the plurality of coders and in which said calculation means use said unique residual error signal to estimate the distortion of the compressed output signal.
6. A device as claimed in claim 1, in which the check means select the new current coder, called optimal coder, after a single comparative test of the distortion estimate calculated for the previous current coder with the reference values.
7. A device as claimed in claim 1, in which the new selected current coder is the current coder previously used if the distortion estimate is in a predetermined range of acceptable values and, if not, is another coder associated with a reference value corresponding to the calculated estimate.
8. A digital signal receiving device for receiving compressed speech signals delivered by a device as claimed in claim 1, characterized in that the receiving device comprises:

   a plurality of source decoders which are compatible with said source coders,

   control means cooperating with said control means at the transmitting end to select a decoder from the plurality compatible with the selected coder at the transmitting end.
9. Telephony equipment comprising a device as claimed in any one of the preceding claims.
10. A telecommunications system comprising at least a device as claimed in any one of the claims 1 to 7 and at least a device as claimed in claim 8, for exchanging speech signals.
11. A digital signal transmission method comprising the following steps:

    a receiving step of receiving an input speech signal,

    a source coding step of compressing the input speech signal by means of a simple source coder, called current coder, from a plurality of source coders having variable output rates and for delivering a compressed output signal at a given output rate, the compressed output signal comprising a distortion relative to an input speech signal,

    characterized in that selection step comprises:

    a calculation step of calculating an estimate of said distortion of the compressed output signal by means of the current coder,

    a selection step of selecting said current source coder, comprising the following substeps:

    a calculation sub-step of calculating an estimate of the distortion relative to the input signal of the current coder,

    a check sub-step of comparing said estimate with reference values and of selecting a source coder, called new current coder, from the plurality of coders with a view to delivering the compressed output signal.

Images